Corporate Decision-Making in Canada
T
HE FINDINGS AND RECOMMENDATIONS assembled in this volume are
mainly the result of work undertaken by outside independent
researchers. As such, the studies ultimately remain the sole responsibility of
each author and do not necessarily reflect the policies or opinions of
Industry Canada or the Government of Canada.
GENERAL EDITORS:
RONALD ]. DANIELS & RANDALL MORCK
Corporate Decision-Making
in Canada
The Industry Canada Research Series
University of Calgary Press
© Minister of Supply and Services Canada 1995
ISBN 1-895176-76-X
ISSN 1188-0988
University of Calgary Press
2500 University Dr. N.W.
Calgary, Alberta, Canada T2N 1N4
Canadian Cataloguing in Publication Data
Main entry under title:
Corporate decision-making in Canada
(Industry Canada research series, ISSN 1188-0988 ; V)
Issued also in French under title: La prise de decision dans les entreprises au Canada.
Conference held in Toronto on March 20-21, 1995
Includes bibliographical references.
ISBN 1-895176-76-X
Cat. no. Id53-11/5-1995E
1. Decision-making—Canada—Congresses.
2. Corporate governance—Canada—Congresses.
I. Daniels, Ronald J. II. Morck, Randall
III. Canada. Industry Canada. IV. Series.
HD30.23C67 1995 658.4'03 C95-980250-9
All rights reserved. No part of this work covered by the copyrights hereon may be
reproduced or used in any form or by any means - graphic, electronic or mechanical -
without the prior permission of the publisher. Any request for photocopying, recording,
taping or reproducing in information storage and retrieval systems of any part of this
book shall be directed in writing to the Canadian Reprography Collective, 379 Adelaide
Street West, Suite Ml, Toronto, Ontario M5V 1S5.
The University of Calgary Press appreciates the assistance of the Alberta Foundation for
the Arts (a beneficiary of Alberta Lotteries) for its 1995 publishing program.
EDITORIAL & TYPESETTING SERVICES: Ampersand Communications Inc.
COVER & INTERIOR DESIGN: Brant Cowie/ArtPlus Limited
Printed and bound in Canada. 1996 printing
This book is printed on acid-free paper.
Table of Contents
PREFACE xiii
PART I THE IMPORTANCE OF
CORPORATE GOVERNANCE
1. CANADIAN CORPORATE GOVERNANCE:
THE CHALLENGE 3
RONALD J. DANIELS & RANDALL MORCK
Globalisation: The Economics Behind the Scenes 3
Corporate Governance and the Public Interest 6
Corporate Governance Problems in Canada 11
The Studies in this Volume 20
PART II THE GOVERNANCE OF
CANADIAN CORPORATIONS
2. GOVERNANCE STRUCTURE, CORPORATE
DECISION-MAKING AND FIRM PERFORMANCE
IN NORTH AMERICA 43
P. SOMESHWAR RAO & CLlFTON R. LEE-SlNG
Introduction 43
The Governance Systems in Canada and the United States 45
Analytical Framework 56
Empirical Results 64
Conclusions 72
Appendix I Sources and Characteristics of the Database 76
Appendix 2 Governance Environment in North America 81
Appendix 3 Corporate Governance Structure: Detailed Tabulations 86
Appendix 4 Concentration of Ownership Tabulations 99
3. CONTROL AND PERFORMANCE:
EVIDENCE FROM THE TSE 300 105
VIJAY JOG & AGiT TULPULE
Introduction 105
Previous Research 106
Terms of Reference 110
Methodology 114
Results 115
Summary, Limitations and Conclusions 135
DISCUSSANTS' COMMENTS ON PART II 141
DAVID A. STANGELAND
Commentary on Part II
Governance Structure, Corporate Decision-Making and Firm
Performance in North America 141
Control and Performance: Evidence from the TSE 300 143
GIOVANNI BARONE-ADESI
Commentary on Part II 145
PART III THE BOARD AND BEYOND
4. ALTERNATIVE MECHANISMS FOR CORPORATE
GOVERNANCE AND BOARD COMPOSITION 149
JEAN-MARIE GAGNON & JOSEE ST-PIERRE
Mechanisms for Corporate Governance 149
Board Composition 159
Data 166
Board Composition and Monitoring 172
Board Composition and Performance 177
Public Policy 182
Conclusions 182
5. EXECUTIVE COMPENSATION AND FIRM VALUE 189
RAMY ELITZUR & PAUL HALPERN
Introduction 189
The Effects of Compensation Plans on Managers' Incentives 192
The Theory of Executive Incentive Compensation Design 195
Linking the Theory of Executive Compensation Design
to Testable Hypotheses 197
Executive Compensation in Practice 199
Relationship of Compensation and Firm-Specific Variables 204
Conclusions 214
Appendix The Theory of Executive Compensation Schemes 220
6. IN HIGH GEAR: A CASE STUDY OF THE
HEES-EDPER CORPORATE GROUP 223
DAVID A. STANGELAND, RONALD J. DANIELS & RANDALL MORCK
Introduction 223
The Economics of Conglomerates 223
The Failure of the Conglomerate Form 226
Data and Methodology 231
Findings 231
Conclusions 234
7. BELL CANADA ENTERPRISES:
WEALTH CREATION OR DESTRUCTION? 241
PAUL HALPERN & VIJAY JOG
Introduction 241
Why BCE? 242
Some Theoretical Background and BCE-Specific Data 243
BCE: Formation and Growth 249
Evaluation of the Operating Strategies 261
Valuation of BCE 268
Summary and Conclusions 277
DISCUSSANTS' COMMENTS ON PART III 282
CLIFFORD G. HOLDERNESS
Commentary on Part III
Large-Block Shareholders and Conglomerates: Sense or Nonsense? 282
VIKAS C. MEHROTRA
Commentary on Part III 289
LEE GILL
Commentary on Part III
Alternative Mechanisms for Corporate Governance
and Board Composition 296
Executive Compensation and Firm Value 298
PART IV INSTITUTIONAL INVESTORS
8. DO INSTITUTIONAL AND CONTROLLING
SHAREHOLDERS INCREASE CORPORATE VALUE? 303
JEFFREY G. MACINTOSH & LAWRENCE P. SCHWARTZ
Shareholders' Incentives and Firm Value 303
Do Controlling Shareholders Increase or Decrease Firm Value? 308
Do Institutional Shareholders Increase or Decrease Corporate Value? 310
Data 314
Methodology 318
Results 319
Window Dressing? 327
The Price-to-Book Ratio and Untraded Equity 350
Summary and Conclusions 330
9. WHY INSTITUTIONAL INVESTORS
ARE NOT BETTER SHAREHOLDERS 341
MICHEL PATRY & MICHEL POiTEVIN
Introduction 341
The Need for Corporate Governance 342
Corporate Governance: Exit and Voice Strategies 348
Institutional Investors in Canada 351
The Governance of Mutual Funds and Pension Funds 354
The Role of Institutional Investors in
Resolving Market Imperfections 363
Policy Analysis and Conclusions 368
10. INSTITUTIONAL ACTIVISM BY PUBLIC PENSION FUNDS
THE CALPERS MODEL IN CANADA? 379
STEPHEN FOERSTER
Introduction 379
Background 38]
The CalPERS Model 384
Survey Methodology 390
Summary and Conclusions 397
11. OUTSIDE FINANCIAL DIRECTORS
AND CORPORATE GOVERNANCE 401
BEN AMOAKO-ADU & BRIAN SMITH
Introduction 401
Independent and Affiliated Outside Directors 403
Methodology 403
Data 406
Board Composition and Sample Characteristics 407
Analysis of Regression Results 410
Conclusions and Policy Implications 413
DISCUSSANTS' COMMENTS ON PART IV
MARK HuSON
Commentary on Part IV 421
MICHAEL S. WEISBACH
Commentary on Part IV
Institutional Investors and the Governance of Canadian
Corporations: An American Perspective 428
PART V INTERNATIONAL ASPECTS OF
CORPORATE GOVERNANCE
12. THE CORPORATE GOVERNANCE
OF MULTINATIONALS 433
RANDALL MORCK & BERNARD YEUNG
Introduction 433
Our Theoretical and Empirical Understanding of
Multinational Firms 435
The Changing Environment for Multinationals 441
Multinationals and Canadian Corporate Governance 444
Conclusions 453
13. CORPORATE GOVERNANCE AND SUPERVISION
OF THE FINANCIAL SYSTEM 457
LEWIS JOHNSON & TED NEAVE
Introduction 457
Deals and Governance 460
Financial Firms 463
Governing Specialized Portfolios 464
Incentives and Governance 467
Change 468
Principles of Supervision 470
White Paper Recommendations 471
Summary 476
Appendix I Recommendations 478
14. BANKS AND CORPORATE GOVERNANCE
IN CANADA 481
RANDALL MORCK & MASAO NAKAMuRA
Introduction 481
The History of Universal Banking 482
Banks and Corporate Governance in Canada 491
Should Public Policy Push Canadian Banks Toward a
Corporate Governance Role? 496
Conclusions and Public Policy Options 498
DISCUSSANTS' COMMENTS ON PART V 503
ROBERTA ROMANO
Commentary on Part V 503
ADRIAN E. TSCHOEGL
Commentary on Part V 512
PART VI CORPORATE GOVERNANCE
AND SOCIAL RESPONSIBILITY
15. REWARDING WHISTLEBLOWERS:
THE COSTS AND BENEFITS OF AN INCENTIVE-BASED
COMPLIANCE STRATEGY 525
ROBERT HOWSE & RONALD J. DANIELS
Introduction 525
Do We Need More Incentives? 527
The Case for Whistleblowing 531
Issues in Calculating Incentives to Whistleblowers 534
Internal and External Whistleblowing:
Conflicts or Complements? 536
Whistleblower Complicity in Corporate Wrongdoing 538
The Risk of Fabricated Claims 540
Guarding the Guardians: The Case for a Private Right of Action 542
Conclusions 544
16. CORPORATE GOVERNANCE AND WORKER EDUCATION:
AN ALTERNATIVE VIEW 551
ALICE NAKAMURA, JOHN CRAGG & KATHLEEN SAYERS
Employer, Worker, Government and Pubh'c Stakes 551
Mechanisms for Inducing Employers to Invest More
in Worker Education 564
Conclusions 572
17. PATIENT CAPITAL?
R&D INVESTMENT IN CANADA 575
RONALD M. GIAMMARINO
Introduction 575
How Patient is the Capital Market? 578
How Patient are Managers? 582
Conclusions and an Agenda for Future Research 587
18. THE GOVERNANCE OF NONPROFITS 591
RON HlRSHHORN
Nonprofits in the Canadian Economy 592
Governance Issues 593
Nonprofitsvas Producers of Different Types of Output 598
Other Factors Affecting the Governance of Nonprofits 604
Transaction Costs and the Problem of Organizational Choice:
An Examination of Community Health Centres and Airports 607
Universities and the Problems of Nonprofit Governance 611
Improving the Accountability of Nonprofits 614
Conclusions 615
DISCUSSANTS' COMMENTS ON PART VI 624
MICHAEL TREBILCOCK
Commentary on Part VI
Corporate Governance and Worker Education:
An Alternative View 624
BRUCE CHAPMAN
Commentary on Part VI
The Governance of Nonprofits 628
JENNIFER ARLEN
Commentary on Part VI
The Effect of Corporate Liability on Whistleblowing Bounties:
A Comment on Howse and Daniels 635
ROGER HEATH
Commentary on Part VI
Patient Capital? R&D Investment in Canada:
An Interim Report 648
Other Avenues 567
19. CORPORATE GOVERNANCE IN THE YEAR 2000 651
JEAN-CLAUDE DELORME
The Growing Power of Institutional Investors 651
What Principles of Corporate Governance will
Apply in the Year 2000? 652
The Adaptability of Companies and their Officers 654
Interest Groups and Corporate Governance 657
Conclusions 658
20. CANADIAN CORPORATE GOVERNANCE:
POLICY OPTIONS 661
RONALD J. DANIELS & RANDALL MORCK
Introduction 661
Options for Improving Canadian Corporate Governance 662
A Public-Policy Philosophy on Corporate Governance 693
ABOUT THE CONTRIBUTORS 697
Preface
I
N AN ERA OF GLOBALIZATION and rapid technological advancement,
investments, particularly those in technology and human capital accumulation,
are key determinants of corporate and national economic success. The role of
corporate governance and the behaviour of corporate institutions are critical
in this setting because they are the primary factors influencing investment
decisions. This view is exemplified by the economic difficulties of many large
and well-known global companies, whose experiences over the course of the
1980s and 1990s highlight the importance of corporate decision-making in
effectively managing the challenges of structural change and economic
adjustment.
One of the striking features of the research on corporate decision-
making in Canada is the extent to which it has remained largely uninformed
by both quantitative and qualitative empirical data. The extent to which past
research has concentrated narrowly on the role of the board of directors and
the board's ability to ensure that stakeholders' interests do not supersede
those of shareholders is a further problem. This narrow focus has neglected
other equally important market and legal constraints that affect corporate
behaviour. As a result, the policy debate surrounding corporate governance
has been too narrowly conceived, and has not been sufficiently attentive to
the distinctive features of the Canadian economic landscape.
The Micro-Economic Policy Analysis Branch of Industry Canada, in col-
laboration with the Financial Research Foundation of Canada, commissioned
18 research studies to broaden and deepen our understanding of several key
issues related to corporate governance and decision-making: ownership and
concentration; the Board; director liability; institutional investors; executive
compensation; minority shareholder protection; and long-term investment
decisions. These issues are addressed from both economic and legal perspec-
tives. The studies were presented and their findings discussed at a conference
on "Corporate Decision-Making in Canada", held in Toronto in March 1995.
The final version of these studies appear in this research volume.
xiii
PREFACE
The research assembled here will contribute to government policy-
making on two broad fronts: by contributing to efforts to improve the growth
and productivity performance of the Canadian economy by bettering the
understanding of the concepts and linkages outlined above; and by deepening
the knowledge-base for ongoing major revisions to federal business legislation,
notably the Phase II amendments to the Canadian Business Corporations Act.
Academic and private-sector organizations actively participate in the
preparation of some of the Department's research documents. Occasionally,
other organizations also contribute resources toward Industry Canada's
research programs. For this volume, I would like to acknowledge the Financial
Research Foundation of Canada which contributed both the time of its principals
and other support toward the successful completion of this project.
Dean Ronald J. Daniels and Professor Randall Morck have overseen this
project and served as General Editors for the research volume. I would like to
thank them both, as well as all of the authors and discussants for their fine
work. I know that this volume will be of interest to the business and policy-
making community, as well as to the wider public interested in economic
issues here in Canada.
JOHN MANLEY
MINISTER OF INDUSTRY
xiv
Part I The Importance of
Corporate Governance
This page intentionally left blank 
Ronald J. Daniels
Dean of Law
University of Toronto
Randall Morck 1
Faculty of Business 1
University of Alberta
Canadian Corporate Governance:
The Challenge
N
O CRYSTAL BALL IS REQUIRED TO PREDICT that, in the coming decades, the
Canadian economy will increasingly be subjected to the phenomenon
the popular press has labelled "globalization". Some key effects of globalization
on Canada are already evident. One of them is that consumers now have
greater choice because products from all over the world are available in
Canada at affordable prices. Another is that globalization constrains govern-
ment in new ways. Investors, entrepreneurs and businesses who do not like
Canadian government policy are free, as never before, to take their business
elsewhere. Globalization has also opened world markets to Canadian businesses
while at the same time subjecting many Canadian companies to competition
from parts of the world almost unheard of a decade ago. In Canada, the
combined effects of globalization have forced a rapid rationalization of the
economy, which has disrupted the status quo. People of different ideological
persuasions may view these effects in different lights, but there is no longer
any doubt that the effects are real.
The purpose of this volume is to examine corporate decision-making in
Canada and to clarify the factors that, in the past, have sometimes led to less
than optimal corporate governance. In this context, poor corporate gover-
nance practices that might have been tolerable even recently are now untenable.
Our ultimate goal is to clarify government policy options that are realistic in
the new global economic environment and also likely to improve Canadian
corporate governance.
GLOBALIZATION: THE ECONOMICS BEHIND THE SCENES
I
N 1930 THE AUSTRIAN ECONOMIST, JOSEPH SCHUMPETER, proposed that a
process he termed "creative destruction" underlies the success of capitalism.
Capitalism hugely, some would say obscenely, rewards people who create inno-
vations that improve efficiency or better meet consumer demand. Capitalism
also destroys firms, sometimes brutally, that fail in these dimensions. Creative
3
1
DANIELS & MORCK
destruction, Schumpeter argues, leads to unmatched improvements in both
production efficiency and living standards. Increasingly, mainstream
economists are accepting Schumpeter's ideas, and now widely agree that giving
free reign to capitalist creativity is more important than avoiding transitory
monopoly pricing or other economic distortions.
Over the last several decades, the role of markets has increased steadily
in both the industrialized and developing world. In large part, this growth of
market importance, and the consequent premium on competitiveness, is related
to the global integration of product, capital, and labour markets. The source of
this integration has been thoroughly canvassed elsewhere and is mainly the
result of reductions in domestic trade-protection barriers, technological inno-
vation, and the liberalization of the command-based economies. The premium
on international competitiveness has been felt more acutely in Canada than
in other countries, owing to this country's relative openness to foreign compe-
tition. Compared to other OECD countries, the Canadian economy exhibits
high levels of export dependency and import penetration. Canada's export
sector, for instance, constitutes 25.2 percent of the domestic economy - second
only to Germany's among the G7 countries in terms of the importance of
export trade to the overall economy.
1
 Likewise, in 1970 Canada's import
penetration rate was more than five times that of the United States, and was
three times the U.S. rate in 1985.
2
 Another indication of Canada's dependence
on external markets is the high level of foreign direct investment. In 1990 for
instance, Canada received 5 percent of the total foreign direct investment
inflows to larger industrialized countries, whereas the United States, with an
economy roughly 10 times as large, received only 29 percent.
3
The increasing openness of industrialized economies to the pressures of
external markets has spawned a number of different effects. One of the most
important is a sharp increase in the pace of innovation. In 1992, 187,200
patent applications were filed in the United States, up from 105,300 in 1972
and 68,384 in 1952. There were also 3,107 new product introductions in the
United States in 1992, up from 1,762 in 1982.
4
 When less tangible innovations
are included in areas such as human resources management, marketing strategy,
etc., the rate of creativity may well be even greater. Continual innovation is
expensive, so innovative firms need to be able to reach large numbers of
customers quickly in order to earn maximum returns on their creativity. Access
to global markets is therefore essential for Canada, and that means granting
foreign firms reciprocal access to Canadian markets.
This stepped-up pace of innovation means firms that lag behind can be
pushed into obsolescence, and their work forces left high and dry. An innovative
new competitor from a remote corner of the world can grab market share with
little warning. The bankruptcy rate tracks the downside of this creativity
explosion. In 1993 in the United States, 85,982 businesses failed; in 1952
there were 8,862 business failures.
5
 Of course, bankruptcy laws and practices
have changed over the years, as has the distribution of company failures
4
CANADIAN CORPORATE GOVERNANCE: THE CHALLENGE
across industries. Also, the employees of failed companies do not always lose
their jobs. Often in bankruptcy, the creditors sell off the assets of the defunct
firm as a unit and the buyer retains many or most of the predecessor firm's
employees. Nonetheless, this increased pace of bankruptcy has substantial
social costs.
In Canada there appears to be a clear national interest in fostering
innovation, in encouraging Canadian firms to get ahead and stay ahead, and
in cushioning firms that fall behind. Yet, the globalization of the economy
constrains government in new ways too.
Traditional government policies based on taxes and subsidies are in
disrepute. Government subsidies and tax credits for R&D are perhaps more
likely to foster innovative "mining" of the government than true innovation.
Government industrial policies that have sought to pick winners and subsidize
their growth have seldom succeeded. Even the one previously notable exception,
Japan, is now known to be no exception at all. Reason & Weinstein (1994)
collected hard data on how much money the Japanese industrial policy directed
at whom; and show convincingly that subsidies in that country were directed
mainly at losers. The recipients of the biggest subsidies in Japan were weak
firms whose collective performance actually declined subsequently. Indeed,
taxing winners to subsidize losers, or even potential winners, is particularl y
unwise in a global economy where individual nations must compete for mobile
capital and, especially, for information (i.e., people with expertise). Both capital
and people can go elsewhere if they are too heavily taxed. In this new environ-
ment, taxing winners heavily to cushion losers is likely to lead in short order
to a country of losers.
In short, government itself has become a competitive business in the
new global economy. In the past, governments were monopolies. Businesses
and people who did not like the government of the day could work to change
it, but rarely could they simply take their business elsewhere. Today they can,
and do. Governments are therefore under pressure themselves to become
"competitive". Competitive government is not necessarily small government;
rather, it is government that provides services most people and businesses
want at tax rates they are willing to pay. Understandably, selective subsidies
financed by taxes levied on everyone are seldom seen to fit these categories.
How then is government, robbed of its traditional policy tools, to promote
the public interest in this new economic reality? We devote the final study in
this volume to a list of viable options. A central thrust of our analysis is that
governments should focus on framework policy. That is to say, the state should
focus on providing the legal and institutional environment in which markets
and firms are able to thrive. As Michael Porter has observed
6
•••[g]overnment's proper role is as a pusher and challenger. There is a vital role
for pressure even adversity in the process of creating national competitive
advantage ... Sound government policy seeks to provide the tools necessary to
5
DANIELS & MORCK
compete, through active efforts to bolster factor creation, while ensuring a
certain discomfort and strong competitive pressure.
In our view, a core feature of an effective framework for competition is
the nature and quality of the corporate governance system that obtains in a
given country. Here, we refer to the legal and market institutions that make up
a country's corporate governance system. Nevertheless, before we begin to
think about the precise nature of an optimal corporate governance system,
there is great need to sort out where exactly the public interest lies in issues of
corporate governance.
CORPORATE GOVERNANCE AND THE PUBLIC INTEREST
T
HE STANDARD OF LIVING ENJOYED BY CANADIAN S depends critically on the
success of Canadian business, which in turn depends on the decisions
made by its top managers. Those decisions are heavily influenced by the legal
and institutional settings in which directors and corporate officers function.
How should corporations be run? If we are to propose ways in which govern-
ment might improve their management, we must consider how corporate
decision-making can go awry in the first place. This depends critically on the
nature of the firm in question.
WHY CARE ABOUT MAXIMIZING SHAREHOLDER VALUE?
A CORPORATION IS A LEGAL FICTION. It has the rights and responsibilities of a
"legal person", yet it is owned by shareholders, and has complex contractual
links to its employees, creditors, customers, suppliers and the community —
collectively termed its "stakeholders". Often, the interests of a firm's share-
holders and various stakeholders conflict with each other and with others'
perceptions of the greater good. This begs the question of whose interests
should be paramount?
It is conventional wisdom, as well as orthodox economic theory, that
those who bear the costs should have the decision-making authority.
Theoretically, at least, this avoids problems analogous to out-of-control
medical bills that arise from authorizing physicians to order tests and obliging
taxpayers to pay for them.
A normal, healthy corporation has well-defined, legally enforceable
contractual commitments to its employees, creditors, customers, suppliers and
community. It may not, in the normal course of business, default on wages,
interest payments, promised shipments, promised payments, or taxes. However,
the shareholders have no such contractual rights. Rather, they are residual
claimants. Whatever money the firm has left over after paying off its contractual
obligations can either be paid out to shareholders as dividends or reinvested to
generate capital gains for shareholders. The firm can freely alter its dividend
6
CANADIAN CORPORATE GOVERNANCE: THE CHALLENGE
payments and investment policies with few legal consequences. Thus, when
unwise business decisions are made in the boardroom, it is the shareholders
who pay the price. For this reason, economic theory dictates, corporations
should be controlled by their shareholders.
Economic theory, as with any theory, is a simplification of reality. When
a firm does poorly, employees may be laid off without the firm actually going
bankrupt. Some stakeholders may thus bear more of the costs of bad manage-
ment than will the typical shareholder. But most stakeholders do not. Senior
workers are usually well protected from layoffs. Of course, if a corporation is
run extremely badly, it may default on its wages, interest payments, deliveries
and bill payments too. But this will happen only when shareholder value sinks
to zero and the firm is bankrupt. Under such circumstances, bankruptcy
trustees must run the firm in the interests of the creditors and other former
contractual claimants who have been made residual claimants. In any event,
from a policy perspective it is important to focus on those stakeholders who
suffer from certain contracting disabilities, and are thus unlikely to have been
able to anticipate and so negotiate for effective protection from the firm (in
the form of ex ante compensation or ex post severance benefits) against the
risks of dislocation. For these stakeholders, strong policy arguments exist for
some type of public intervention, although not through modifications to the
traditional apparatus of corporate governance.
7
Ultimately, the reason good corporate governance is important is that its
absence would erode public confidence in Canada's financial markets and
therefore depress share prices. Such a lack of confidence would make raising
equity capital very difficult for Canadian companies, limiting their potential
for growth. This, in turn, would slow economic growth and thus exacerbate
problems such as unemployment and government deficits. Good corporate
governance, therefore, is unquestionably in the public interest.
DIRECTORS ARE SHAREHOLDERS' REPRESENTATIVES
IN PRACTICE, IN FIRMS WITH MANY SHAREHOLDERS, it is difficult for all share-
holders to be consulted on all business decisions. The solution is the board of
directors. Directors are elected by the shareholders and paid to represent the
interests of the shareholders in corporate decision-making. To emphasize the
ultimate purpose of the board, the law makes directors personally liable to law-
suits by shareholders if they fail in this duty. Officers of the corporation, top
managers such as the CEO, president and senior vice-presidents, are assigned a
similar legal duty and liability. In the jargon of economics, this is a type of
principal-agent relationship: the shareholders are the principals and the officers
and directors are their agents. Corporate directors and officers are required to
act in the best interests of the corporation, and that means the best interests of
its legal owners — the shareholders.
7
DANIELS &. MORCK
SHAREHOLDERS AND STAKEHOLDERS: A PRACTICAL COMPROMISE
TO SOME, THIS DOCTRINE MAY SEEM TO BE EXCESSIVELY NARROW. There are,
after all, others besides shareholders whose fates are interwoven with that of
the firm: its employees, creditors, managers, customers, suppliers, and the
communities that depend on it. To reiterate, these parties are the firm's stake-
holders.
8
' Would it not be better if top managers ran the firm in the best interests
of society, or the community, or at a bare minimum the workers and the share-
holders together?
The economist's response is based on two considerations. First, the legal
system collapses all of these into the interests of shareholders alone. If a firm
passes over top (well-qualified) job applicants because of racial or gender
prejudice, firm performance is suboptimal and the shareholders lose. If a firm
pollutes the environment and is sued, the shareholders lose. If it mistreats its
workers and is subjected to strikes or other labour unrest, share prices and
dividends fall and, again, the shareholders lose. Second, even when managers
make patently foolish decisions, they can usually point to some resulting social
good - for example, benefits to some group of workers. Clearly, assigning a
manager such a multi-dimensional responsibility effectively erases all respon-
sibility. In this context, responsibility to all means responsibility to none.
The law has evolved a workable compromise. Managers owe principal
duties to shareholders, but the legislatures and the courts have developed a
range of overriding duties (and corresponding sanctions) to ensure fidelity to
broader social goals. As a consequence, a corporation cannot claim devotion
to shareholder interests to justify its neglect of explicit occupational health
and safety, environmental, or human rights obligations. Not only will a failure
by the corporation's agents to meet those obligations subject them to individual
sanctions, the law also imposes financial sanctions on the firm's shareholders
in the form of penalties levied on the corporation. In this way, shareholders
have powerful incentives to monitor and discipline corporate misconduct.
This compromise has strong efficiency properties; but it is also overlaid with a
thick layer of democratic theory. Instead of vesting an unelected and unrepre-
sentative cadre of senior corporate managers with the task of determining
how corporate resources should serve the public good, this model relies on
accountable and elected legislatures to make those decisions. This means
that the decisions as to when and how corporate externalities should be
internalized are fully transparent and subject to full and proper public deliber-
ation and accountability.
MYOPIA
ANOTHER WIDELY REPEATED CONCERN with focusing on shareholders' interests
is that shareholders themselves are said to be myopic. It is alleged that they are
concerned mainly with short-term performance, so excessive catering to the
8
CANADIAN CORPORATE GOVERNANCE: THE CHALLENGE
wishes of shareholders means forsaking long-term investments. As the study by
Giammarino explains very convincingly, there is absolutely no credible evidence
that this concern has any basis in fact. Statistical analyses of large numbers of
U.S. firms show that firms' share prices rise when they announce long-term
investment projects or large R&D programs. The apparent conclusion is that
long-term investments please shareholders. This is supported by other studies
that find a strong and sustained positive correlation between R&.D spending
and share value. If firms do have a short-term bias, there is strong evidence to
suggest that the average shareholder would be pleased to see this change.
PRINCIPAL-AGENT DUTIES
A MORE LEGITIMATE CONCERN IS THAT under some conditions managers can
ignore small, poorly informed shareholders. Thus, a board with a single large
active shareholder may toady to that shareholder while small investors, who
collectively own most of the firm, are effectively disenfranchised. In firms that
do not have a large shareholder, such as the large chartered banks, there is a
danger that managers may ignore shareholders entirely and run their firms as
personal fiefdoms. Shareholder rights activists allege that managers can then
pursue pet projects, adopt biased hiring policies, and otherwise waste the
shareholders' money. Such breakdowns of the principal-agent relationship are
termed agency problems. Mainstream economics recognizes that various sorts of
agency problems are pervasive throughout both the public and private sectors.
Indeed, some economists even go so far as to allege that agency problems are
the chief cause of economic inefficiency in modern capitalist economies.
Although agency-cost nomenclature is relatively new, the concept of
accountability which underlies it is not. Since the early part of the century,
corporate scholars have worried about the accountability problems set in train
by the delegation(s) of authority required to realize gains from specialization in
the modern corporation. Berle & Means' (1932) seminal study of the American
corporation was focused precisely on this issue. It was these scholars who coined
the phrase "separation of ownership and control" to describe the American
system of corporate governance. Berle & Means conceived corporate America as
riven by pervasive accountability problems emanating from scattered, small-
stakes shareholdings. With so many shareholders, there was no incentive on the
part of any shareholder to assume responsibility for controlling the affairs of the
corporation. The consequence was virtually unchecked power for American
corporate managers and the resultant suppression of the profit motivation.
In retrospect, it is clear that the Berle & Means' account was overly bleak.
While it is undoubtedly true that small-stakes shareholders exert very little, if
any, direct control over directors and managers in large public corporations, it
does not necessarily follow that managers are entirely unbridled and free to
frolic on their own. As a number of law and economics scholars have demon-
strated, a variety of legal and market devices work to align managerial and
9
DANIELS & MORCK
shareholder interests. Legal instruments (such as shareholders' rights to sue
directors and officers) ensure managerial accountability by imposing ex post
costs on self-dealing managers. Market instruments (such as the takeover or
the "corporate control" market, the market for managers, and the capital market)
typically focus on less malign sources of managerial misconduct, and operate
either directly (through the threat of displacement or debased reputation) or
indirectly (through provision of information on managerial misconduct to
parties capable of taking direct action). The existence of these various legal
and market instruments does not mean, however, that the problem of account-
ability is trivial in the modern corporation; some residual agency problems
remain. Rather, the claim is that the legal and market arrangements that
comprise the system of corporate governance are fairly robust and, thus, to the
extent that improvements to that system can be made through institutional
reconfiguration, these gains are on the margin.
THE ROLE FOR LAW
IF FIRMS ARE BASED ON VOLUNTARY ACTIVITY among well-informed deliberative
stakeholders, and if markets play an important supporting role in disciplining
managerial misconduct, what role is there for law and legal institutions? As
mentioned earlier, if it can be demonstrated that certain stakeholders are
being denied access to adequate information, that their bargaining with the
corporation is beset by severe asymmetries in power, or that they are being
coerced into certain commitments with the corporation, then a plausible case
for some sort of government intervention can probably be made. Nonetheless,
most commentators agree that, save for employees, claims of this sort are
unpersuasive; and even in those cases where the claims are clearly legitimate,
it is not altogether clear that the best form of state intervention is through
corporate law — because corporate law is usually viewed as being devoted to
shareholder and, to a lesser extent, creditor interests. Policy makers must
therefore be very careful about overloading a single regulatory instrument
with multiple and often conflicting goals. This would be the result if the interests
of employees and other constituencies deserving of protection were to be
protected through corporate law.
If corporate law is, indeed, primarily about shareholder interests, what
form should it take? Early corporate statutes contained several mandatory
elements that were clearly in support of a highly interventionist role for the
state in ordering private arrangements. Today, however, the clear trend in
corporate law is toward an enabling regime, which confers considerable latitude
on parties to pick and choose among various background terms. This enabling
role for corporate law is consistent with the belief that the interaction
between a shareholder and the corporation is largely voluntary in nature, and
the law should, as much as possible, defer to the wishes of contracting parties.
Viewed in these terms, the role for corporate law is clear: lawmakers should
10
CANADIAN CORPORATE GOVERNANCE: THE CHALLENGE
develop and maintain a corporate law regime that facilitates contracting by
private parties. One way to accomplish this is to supply background legal terms
that economize on the costs of repeated negotiation for private parties.
Another way to facilitate private contracts is through the supply of certain
terms that private parties are unable to generate on their own because of high
investment costs and risks of appropriation (the public goods problem). The
elaborate system of fiduciary duties developed under corporate law is an example
of such a public good.
CORPORATE GOVERNANCE PROBLEMS IN CANADA
T
I
HE ECONOMICS UNDERLYING THESE AGENCY PROBLEMS varies across types of
firms. We examine each of the most common types of corporation in turn.
WIDELY HELD FIRMS: OTHER PEOPLE'S MONEY ?
A FIRM IS WIDELY HELD WHEN IT IS OWNED by a large number of small share-
holders, each of whom has no effective control over management decisions.
Some of Canada's largest firms, and almost all large U.S. firms, fall into this
category. All the major Canadian chartered banks are widely held. So are Bell
Canada and Air Canada. Although these and a number of other prominent
Canadian firms are widely held, this genre of ownership structure is not
common in Canada. Morck & Stangeland place only 16 percent of the 550
largest Canadian corporations in this category in 1989.
It is commonly alleged that in widely held firms managers too easily forget
their duties as shareholders' agents and govern their firms to benefit them-
selves. This agency problem impoverishes shareholders and undermines the
economic logic that links optimal corporate policy to the common good. For
example, suppose a manager gains status and social influence from an
unprofitable film-making subsidiary. Closing it would benefit the firm by
$5,000,000 but would cost him (personally) intangible losses he values at
$50,000. If he owns one-half of one percent of the firm's outstanding shares (a
situation not uncommon in many large widely-held firms) he will forego
$25,000 of share value but keep $50,000 in intangible benefits. He thus comes
out $25,000 ahead. The other shareholders lose the remaining $4,975,000,
perhaps without ever knowing they might have had it. Private admissions
by corporate insiders, as retold by Mace (1971) reveal such instances to be
disturbingly common in large U.S. firms. We doubt that large Canadian firms
are entirely innocent either.
Of course, most self-serving behaviour by managers is less transparent. It
might involve a phenomenon economists have dubbed "managerialism":
corporate empire building through unprofitable takeover binges that enhance
only the top managers' egos. Another possibility is ethnicity or gender-biased
hiring or promotion policies that keep things comfortable for the managers but
u
DANIELS & MORCK
cost shareholders the value the best candidates would have added to the firm.
Yet another example of managers' self-serving behaviour is funnelling share-
holders' money into economically questionable pet projects such as unviable
subsidiaries in exotic places. Some managers actually find it so wrenching to
pay out cash windfalls to shareholders through increased or extraordinary
dividends that they invest in almost any project, no matter how unprofitable,
to keep the money inside the firm and under their control. Unnecessary Lear
jets and palatial head office buildings are almost a caricature of self-serving
managerial behaviour.
Because widely held firms are characteristic of corporate America, both
the mass media and the academic research literature have dealt extensively
with instances of self-serving management in widely held firms. The hit movie
"Other People's Money" and the high-profile attention newspapers now give
to poison pills, greenmail, and other instances of managerial misbehaviour
testify to the extensive public awareness (if not always understanding) of
corporate governance issues in widely held firms.
CLOSELY HELD FIRMS: ENTRENCHED INSIDERS?
As THE STUDY BY RAO AND LEE-SlNG SHOWS, most large Canadian firms are
not widely held. In more than three-quarters of the Canadian corporations
they examine, at least one large blockholder controls 20 percent or more of
the voting shares, and in over half of the firms a single blockholder controls
more than 50 percent of the voting shares. Large shareholdings by manage-
ment often enable them to dominate shareholder meetings since most small
shareholders do not attend. This allows management to control director
appointments and thus indirectly control corporate decisions. Under these
circumstances, it is unlikely that senior managers would ever forget about
dominant shareholders' interests for long. Given this, one might think
Canadians would rejoice that most of our large firms are free of American-
style agency problems. Unfortunately, the ownership structure of Canadian
firms does not entirely eliminate these problems, and it brings with it another
set of agency problems.
In closely held firms, the fear is that directors and officers will toady
excessively to the dominant shareholder and ignore smaller investors.
Consequently, their fiduciary duty to act in the interests of the corporation is
interpreted to mean acting in the interests of all the shareholders. The agency
problem here is the possible conflict of interest between the dominant share-
holder (supported by the officers and directors who are under the dominant
shareholder's control) and the other shareholders.
There is considerable evidence from the United States that blockholders
do extract private benefits from firms. Barclay & Holderness (1989, 1992)
show that large blocks of stock are generally transferred at prices higher than
those prevailing on the open market for the same shares. Presumably this is
12
CANADIAN CORPORATE GOVERNANCE: THE CHALLENGE
because large blocks of shares confer more benefits than small stakes. Barclay,
Holderness & Pontiff (1993) make the further case that the prices of many
closed end funds in the United States are depressed because controlling
blockholders extract private benefits. There is no reason to assume Canadian
blockholders are more altruistic than their American peers.
Dominant shareholders are perhaps less likely deliberately to push the firm
toward non-value-maximizing activities of the sort described in connection with
widely held firms. After all, the dominant shareholder pays a high percentage of
the cost himself. However, it is not reasonable to rule out such behaviour entirely.
Large blockholders often dominate shareholder meetings with 20 percent of
the stock or less. A decision that costs the firm $5 million is clearly not in the
interest of a 20 percent dominant blockholder unless it also generates private
benefits she values at more than 20 percent of $5 million, or $1 million.
Certainly, such situations are not impossible.
An additional set of potential problems in closely-held firms involves
what financial economists call "entrenchment". Dominant blockholders who
exert a detrimental influence over corporate policy are almost impossible to
remove; they are largely immune to takeovers, proxy challenges and board
rebellion. Unfortunately, some dominant shareholders who originally brought
value to their companies may continue to exercise control long after they
should have retired.
There is substantial evidence that managerial entrenchment is also
common. Morck et al. (1988) show that in the United States firm performance
rises with insider ownership for widely held firms, but then falls as ownership
levels rise above a threshold that permits entrenchment. Johnson et al. (1985)
show that sudden deaths of CEOs over the age of 70 cause their firms' share
prices to rise on average. Often, the death of a firm's dominant blockholder leads
to it becoming widely held as the heirs cash out. However, the inheritance of
dominant blocks of stock can also put less competent heirs into positions of
power they have not earned. Morck & Stangeland (1994) find that Canadian
firms whose dominant shareholders are their founders' heirs perform significantly
worse than other firms of the same age and size in the same industries.
FIRMS WITH DUAL-CLASS SHARES: THE WORST OF BOTH WORLDS?
CANADIAN LAW AND PRACTICE ALLOW COMPANIES free reign to issue multiple
classes of shares with different voting rights. This, in theory, allows closely
held firms to grow without the dominant blockholder losing control. In practice,
many fear that it also opens Canadian firms to the worst of both worlds. By
issuing themselves stock with many votes per share, while others hold shares
with few or no votes, dominant shareholders can entrench themselves
although they own only a tiny fraction of the firm.
Dual-class recapitalizations (i.e., transformations of one-vote-per-share
firms into firms with different classes of voting stock) can be coercive. For
13
DaNELS 7 MORCKhh
example, suppose small shareholders in a one-vote-per-share firm are given
two weeks either to convert their common stock into class B common stock
that will have no votes but will pay an extraordinary dividend, or to commit
to retaining their existing common shares (renamed class A common) that do
have votes. On the one hand, each small shareholder knows that if all the
others convert and she retains her class A stock, she will miss out on the
extraordinary dividend and be left with a vote that is essentially useless. Thus,
she should convert. On the other hand, if all the other shareholders retain
their class A shares, if she chooses to convert her share to class B, her action
will not, by itself, allow management to become entrenched; so she might as
well have the dividend. Thus, again, she should convert. In essence, each
small shareholder is enticed to convert her stock to non-voting common,
despite the fact that this course entrenches management and reduces the value
of the firm. Jarrell & Poulson (1988) show empirically that dual-class recapital-
izations tend to lead to entrenchment and depress firm values.
At present, there are two specific ways in which corporate and securities
law constrain the scope for opportunistic recapitalizations. First, it is open to
shareholders to undertake a derivation action or seek an oppression remedy on
the grounds that such conduct is motivated by an improper purpose. Such a
claim would be salient in the context of a share recapitalization effected in the
context of a hostile takeover bid. Second, both corporate law (provisions
respecting fundamental changes) and securities law (e.g., Ontario Securities
Policy 1.3) require special shareholder votes when dual-class share structures
are created. These votes enable dissident shareholders to object to opportunistic
dual-class recapitalizations.
FIRMS WITH TAKEOVER DEFENCES: PROTECTING SHAREHOLDER S
FROM THE TEMPTATIONS OF WEALTH?
HOSTILE CORPORATE TAKEOVERS ARE EVENTS that often pit incumbent managers
and workers against shareholders. Takeover bids are always good for share-
holders because tender offers to buy control are generally made at premia of
more than 30 percent above previous stock market prices, and can be much
higher. It is difficult to see why shareholders need to be protected from selling
their stock on such favourable terms. Indeed, shareholder rights activists argue
that takeover defences exist primarily to entrench top managers who have
established comfortable positions for themselves.
This view may be excessive. In some circumstances it is in the interests
of the shareholders to have a takeover delayed so alternate buyers can be
found. If a bidding war can be started, the ultimate takeover price might be
increased even more. With this justification, many large Canadian firms have
constructed defences against hostile takeovers.
14
CANADIAN CORPORATE GOVERNANCE: THE CHALLENGE
Poison Pills
These are amendments to corporate charters that penalize shareholders who
acquire more than a certain amount of stock. For example, a flip-in poison pill
might declare anyone who buys more than 15 percent of outstanding voting
stock to be an "acquiring person", and then go on to say that in the event any-
one becomes an acquiring person, all other shareholders except the acquiring
person shall receive 10 free shares for each share held. This reduces both the
value of the acquirer's position and its voting strength by 90 percent; the
acquirer is virtually back where she started.
In Canada, shareholders must vote on (i.e., approve) poison pills.
However, in some cases the vote is tied to other issues, such as increased dividends,
which casts doubt on the extent to which shareholder approval is truly
voluntary. Nevertheless, the Canadian strain of poison pills is much less virulen t
than its American counterpart, which suggests that the requirement to obtain
shareholder approval has limited somewhat the scope for opportunism. Even
more significantly, there have been several setbacks for poison pills in Canadian
courts and securities commissions.
9
 The general thrust of these decisions is that
poison pills may buy time for managers to conjure up another offer for share-
holders, but ultimately shareholders must be given the opportunity to decide
whether or not they want to tender their shares to an offer.
Voting Caps
Many corporations that have been established by Acts of parliament, such as
chartered banks and privatized Crown Corporations like Air Canada and
PWA Corp., have legislative voting caps. These conditions, set out in the
statutes that created the firms, make it illegal for any shareholder to own more
than minimal amounts of the firms' shares. In the case of Air Canada, the
limit is 4 percent. For the banks, the limit is 10 percent. Voting caps are merely
extreme forms of poison pills.
Takeover Rules
Under Canadian securities laws, a takeover bid is defined as any offer to
acquire an issuer's equity that would confer more than 20 percent ownership of
a single class of shares on the offerer. Once a takeover is deemed to have
occurred, the acquirer must comply with certain rules, including pro rata take-
up of shares, minimum bid periods, information disclosure obligations, and so
on. For purchases of control from dispersed shareholders, the rules of Canadian
securities law do not operate much differently from those of the United States.
The crucial difference is in the context of sale of control by an existing control
holder. Whereas these transactions are subject to only selective ex post review
for substantive fairness in the United States, in Canada, the entire takeover
15
DANIELS & MORCK
regime applies to these transactions, thereby entitling all shareholders to
participate pro rata in the transaction. The effect of this rule is to raise the
costs of a change in control transaction for an interested acquirer. Because an
existing control blockholder is unlikely to want to part with only a portion of
her holdings (minority status is an unattractive prospect for a controlling
shareholder), the acquirer is forced to bid for 100 percent of the company's
shares. Many financial analysts regard this as a thinly disguised anti-takeover
rule. By striving to make takeovers utterly fair, we may have made many of
them prohibitively expensive.
Do takeover defences ultimately benefit shareholders? The preliminary
answer appears to be "no". Empirical evidence suggests that on average
takeover defences do have an entrenchment component. Stangeland (1994)
finds that firms with poison pills record performance levels below those of
industry rivals without such anti-takeover defences. Other recent studies also
find that the adoption of poison pills and other takeover defences is correlated
with reduced share value. A recent study by Comment & Schwert (1995)
appears to contradict this, however.
FIRMS WITH FREE CASH FLOW
HARVARD BUSINESS SCHOOL PROFESSOR MICHAEL JENSEN suggests that corporate
financial policy is closely related to corporate governance issues. He theorizes
that in mature industries, a firm's existing operations produce substantially
more cash flow than is needed for profitable capital investments. This excess
he calls "free cash flow". Firms should use cash flows they cannot profitably
use internally to pay increased dividends. In firms with inadequate corporate
governance, managers may seek to retain control over their firm's free cash
flow by retaining it for suboptimal investments. According to Jensen (1986), a
low dividend rate in a mature industry is strong evidence of poor corporate
governance. He also suggests that, in order to prevent managers from mis-
investing funds, firms in cash-rich, mature industries should be more highly
levered. Thus, there is a high probability that cash-rich firms with low debt are
also subject to poor corporate governance.
The study in this volume by Gagnon and St. Pierre takes a preliminary cut
at Canadian data and finds no evidence of a systematic link between leverage or
dividend policy and performance. More specific empirical tests, analogous to
those undertaken in the United States, have not yet been performed for Canada.
CONGLOMERATES: A SHELL GAME ?
IN CANADA, AS IN CONTINENTAL EUROPE, KOREA AND JAPAN, much corporate
activity is undertaken by conglomerates that consist of numerous related firms
that collectively own controlling blocks of each others' stock. Public share-
holders own the remaining shares at each level.
16
CANADIAN CORPORATE GOVERNANCE : THE CHALLENGE
There are many valid reasons in economic theory to explain the existence
of conglomerates. It is costly for firms to raise external capital. Financing invest-
ment projects is simpler and cheaper if it can be done using internal funds.
Conglomerates can serve as a sort of internal capital market for member firms.
Excess cash from one firm can be invested in another if the return there is higher.
If conglomerates are run by managers who understand and can control all its
diverse parts, they should make considerable economic sense.
However, the performance of conglomerates in general has not lived up
to such expectations. Lang & Stulz (1992) show that the performance of
conglomerates lags behind that of focused firms. Also, the collapses of conglom-
erates like Argus, Olympia & York and the Hees-Edper group of firms have
added to investors' doubts about the real economic value of conglomerates. In
the United States, conglomerates constitute a disproportionate share of hostile
takeover targets. Raiders there have found the share prices of some conglomerates
to be so depressed that money can be made by buying the whole conglomerate,
breaking it up, then selling all its parts separately. In these cases, at least, the
parts are worth considerably more than the whole.
The underlying problem with conglomerates is widely perceived to be
that they are much more difficult to manage than focused corporations. It is
difficult, if not impossible, for the head-office managers in a conglomerate to
understand each component business thoroughly enough to formulate strategies
that are as effective as those of their more focused rivals. This undercuts the
main advantage of a conglomerate - the alleged allocation of the group's
capital to where it earns the highest return. But more than that, conglomerates
open up a whole new type of agency problem.
By controlling interfirm dividends, by having companies within the
conglomerate group lend to each other at non-market interest rates, by organizing
intercorporate billing for goods or services at artificial prices, or by transferring
assets at synthetic prices, conglomerate managers can reduce profits in one
firm and increase them in another. The fear is that profits in firms where insiders
own relatively less stock might be diverted to firms where they own most or all
of the stock — a kind of corporate shell game. In such a case, the agency problem
is the plural version of that in a closely held firm: that the insider shareholder
who controls the conglomerate might enrich herself at the expense of the
public shareholders in all its firms.
An analogous problem arises for tax authorities in other countries when
money flows from profitable, and therefore taxable, firms to loss-making firms
within a conglomerate. This is not a problem here in Canada because tax-free
payment of dividends within a corporate group is entirely legal.
In fact, as the study by Daniels, Morck & Stangeland points out, there
are numerous other features of the Canadian legal and institutional environ-
ment that also facilitate conglomerate formation. Canada's current Investment
Companies Act is a less effective barrier to establishing conglomerates with
large numbers of partially owned subsidiaries than is the Investment Company
17
DANIELS & MORCK
Act of 1940, and its requirements can easily be avoided through provincial
reincorporation. More liberal interest deductions in Canada subsidize debt,
which provides favourable financing for acquisitions. The lack of a vigorous,
privately enforced securities disclosure regime in Canada reduces the trans-
parency of internal corporate transactions, and heightens the attractiveness of
the conglomerate form of organization to opportunistic corporate insiders.
Similarly, the lack of a clearly articulated corporate law fiduciary duty from
majority to minority shareholders in Canada helps explain, at least historically,
the attraction of conglomerates to opportunistic controlling shareholders. The
absence of such fiduciary duties allows controlling shareholders greater scope
for unfair self-dealing than would be possible in the United States.
Daniels et al. also argue that the mercantilistic industrial policies of
successive Canadian governments encouraged conglomerate formation.
Restrictions on foreign investment by Canadians, such as the foreign property
rule in the Income Tax Act, reduce Canadian shareholders' investment oppor-
tunities. When they may disagree with the policies of corporate managers,
shareholders here have fewer alternative places to put their money than would
be the case if they could freely invest abroad. This may have allowed inefficient
conglomerate holding structures to survive, and may thereby have prolonged
wealth-reducing redistribution from investors to Canadian corporate insiders.
Trade protectionism and favourable tax treatment of certain types of domestic
equity investments also contribute to an inward looking industrial economy.
Canadian corporations have focused on producing a broad range of goods and
services for the protected Canadian market rather than on a narrow range of
competitive products for the global market. In this setting, the diversified
conglomerate serves as a natural vehicle for achieving corporate growth.
Further supporting the formation of the conglomerate was, in sharp contrast to
the United States, a more congenial political environment for the concentration
of economic power. Whereas American political traditions embody a deep and
abiding mistrust of concentrated economic power, the Canadian political
environment is more sanguine. Here, the development and preservation of a
fragile national identity easily overcame concerns about concentrated power.
So, to the extent that economic concentration may be the inexorable result of
nationalism, Canadian political leaders have regarded it as a price worth paying
to promote collectivist goals.
MULTINATIONALS: A GLOBAL SHELL GAME?
MULTINATIONAL CORPORATIONS ARE MULTI-FIRM ORGANIZATIONS akin to
conglomerates, but with a more convincing economic rationale. All the
subsidiaries of a multinational are usually in the same line of business so the
overall organization is easier for head-office management to run than is a
cross-industry conglomerate. Moreover, multinationals have immediate access
to markets in many countries. This can be critical in earning a quick high
18
CANADIAN CORPORATE GOVERNANCE: THE CHALLENGE
return on expensive investments like R&D. For investment in innovation,
production and marketing costs are often minimal compared to upfront R&D
costs. Thus, the larger the firm's market for its new product, the higher the
return on its original R&D. For firms in R&D-intensive industries like pharma-
ceuticals, computers, telecommunications equipment, home electronics, etc., a
multinational structure is almost essential. Foreign partners are often avoided in
these industries because of a fear of reverse engineering or the theft of propri-
etary information. The same considerations apply in other industries with high
up-front fixed promotion costs like music recording or films, although foreign
partners are a more practical alternative there. Morck & Yeung (1991, 1992)
present empirical evidence that foreign subsidiaries do, in fact, add value only
for firms with high spending on R&D or advertising.
There is, however, another reason for a multinational structure: tax
avoidance. By shifting profits between subsidiaries (employing the same methods
used by conglomerates) multinationals can control which subsidiaries are the
most profitable and hence the most taxable. Harris et al. (1993) provide
empirical evidence that U.S. multinationals commonly shift income from
highly taxed to less-taxed subsidiaries. Canada has higher taxes than many of
the other countries in which multinationals operate. Given higher domestic
tax rates, multinationals operating in the Canadian environment have strong
incentives systematically to shift profits out of Canada through manipulation
of transfer pricing schemes. Not only does such behaviour reduce the revenues
flowing to the Canadian branch, it also reduces the wealth of Canadian
investors who hold minority stakes in the multinationals' subsidiaries. This
phenomenon illustrates poignantly the law of unintended consequences; the
creation of partially owned foreign subsidiaries was encouraged by Canadian
tax and foreign investment policy.
10
COOPERATIVES: THE MEMBERS' MONEY?
A NUMBER OF INDUSTRIES THAT ARE MADE UP OF corporations in other countries
contain cooperatives in Canada. These organizations are owned by their
members but controlled by professional managers. Thus, in theory, they might
share many of the problems of lack of managerial accountability that afflict
widely held firms and firms with entrenched management.
CROWN CORPORATIONS: TAXPAYERS' MONEY?
DESPITE A SERIES OF PRIVATIZATIONS DURING THE 1980s, Crown Corporations are
still very much a part of the Canadian business scene. Corporations like the
CBC, Alberta Treasury Branches, Ontario Hydro, and BC Tel. are unlikely to
face privatization any time soon. Universities and hospitals are likely to remain
in the public sector too. Given the agency problems that pervert decisions in the
private sector, is not public-sector ownership an attractive alternative ?
19
DANIELS &. MORCK
The answer is an emphatic "no". Megginson, Nash & Van Randenborgh
(1994) show that the performance of state-controlled enterprises, including
those only partially owned by the state, is unambiguously worse than that of
similar private sector firms. This begs the question, "why?".
The reason seems to be that state-owned enterprises have their own set
of agency problems that are, in many ways, more intractable than those of
private-sector firms. In principle, Crown Corporations are supposed to be run
in the public interest. In practice, this often means they are run in the interests
of politicians and political appointees who pay none of the substantial costs of
poor investments, empire building, etc., compared to the small fraction of such
costs incurred by the managers and dominant shareholders of private-sector
firms. The overriding agency problem in public-sector firms is that politicians
and political appointees tend to lose sight of their duty to the public.
Moreover, dysfunctional corporate governance in private-sector companies is
ultimately constrained by the firm's bottom line and the bankruptcy that its
violation triggers. State-owned enterprises have what economists call "soft"
budget constraints - their deficits are picked up by the taxpayers. State-owned
enterprises can thus tolerate worse governance than can their private-sector
counterparts. Furthermore, those mechanisms that limit agency problems in
private firms, such as shareholder votes, takeovers, project-based capital market
scrutiny etc., are not features of the governance of state-owned enterprises.
The only restraining lever the public holds is the threat of electing politicians
who will privatize, and this is being exercised increasingly often.
OTHER NONPROFIT ENTERPRISES : DONORS' MONEY?
THE LARGEST CHARITABLE ORGANIZATIONS CAN BE as big and complex as large
corporations. Their top executives have responsibilities on a par with those of
corporate executives, and make decisions involving as much money. Yet
charitable organizations have nothing analogous to shareholder votes, annual
reports, etc. To provide accountability, director liability rules do extend to
charities, even small local organizations. Is this the best way of making sure
the managers of the charity act as their donors expect?
THE STUDIES IN THIS VOLUME
I
N THIS VOLUME, INDUSTRY CANADA and the Financial Research Foundation
of Canada have gathered together the thoughts of leading Canadian business
and legal academics on topics related to corporate governance in this country.
The studies were presented at a two-day conference sponsored by Industry
Canada and the Financial Research Foundation of Canada, held in Toronto in
early 1995. A number of authorities were invited to comment on the presen-
tations; their comments are also included in this volume.
20
CANADIAN CORPORATE GOVERNANCE: THE CHALLENG E
PROBLEMS AND POTENTIAL IN THE GOVERNANCE OF
CANADIAN CORPORATIONS
IN THE STUDY TITLED "Governance Structure, Corporate Decision-Making and
Firm Performance in North America", P. Someshwar Rao and Clifton Lee-
Sing, both of Industry Canada, present a thorough and exhaustive statistical
analysis of this topic using several hundred firms from both Canada and the
United States. They essentially correlate various indicators of corporate strategy,
such as leverage, capital intensity, R&D spending, and foreign market pene-
tration, with indicators of corporate governance such as whether a company is
widely held or closely held and how its board is structured. All of these
variables are measured relative to benchmarks for firms in a given size range
and in a specific industry. They then perform a similar analysis correlating
standard accounting performance measures such as return on equity, return on
assets, and various growth and productivity measures, again measured relative
to size and industry benchmarks.
Their results are quite interesting. They find no consistent effect of insti-
tutional (i.e., pension fund, etc.) ownership on either corporate strategy or
performance in Canada, but do find positive effects on both in the United
States. This begs questions as to why Canadian institutional investors are more
reticent about pushing for better corporate governance than their American
peers. Later studies in the volume try to answer these questions. They find little
difference in either strategy or performance between widely held and closely
held Canadian firms, but find positive effects of heightened insider ownership
and negative effects of highly concentrated ownership for U.S. firms.
These findings are consistent with, and add to, other studies of both
countries. Morck & Stangeland (1995) find that the critical difference in
Canada is between subclasses of closely held firms. Closely held firms controlled
by entrepreneurs outperform widely held firms, while closely held firms
controlled by heirs lag them. Several studies of U.S. data, including Morck et
al. (1988), McConnell & Servaes (1993), and others, find that increased
insider ownership improves corporate performance up to a point but beyond
that, highly concentrated ownership is associated with poorer performance.
Rao and Lee-Sing's results for their board structure variables are also
interesting. They find that in Canada large boards are associated with less
R&D, and poorer overall performance and productivity, whereas in the United
States big boards seem to have little effect, either positive or negative. They
find that foreign directors have weak positive effects on Canadian firms, as
does having a CEO who is also chairman of the board. They find no real effect
in the United States on performance for the average firm. This is consistent with
other U.S. studies, e.g., Weisbach (1988) and Morck et al. (1989), and
Hermalin & Weisbach (1990), that find that board structure seems unimportant
for typical firms, but matters when performance is poor. The presence of out-
siders on the board is correlated with poorer performance in both countries.
21
DANIELS & MORCK
It is important to emphasize the limitations of statistical evidence. First,
statistical correlation does not usually imply causation. For example, a corre-
lation between insiders on the board and good performance could be due
either to outside directors causing poor management or to poor management
causing shareholders to demand more outside directors. Statistical evidence of
the sort in this volume can be consistent or inconsistent with a given conclusion;
it cannot provide definitive proof. Second, the results in this study are from a
type of statistical analysis called multiple regression, which looks for effects of
one variable above and beyond the effects of the other variables. Thus, the
fact that the presence of inside directors is positively correlated with perfor-
mance while inside ownership is not, means that inside ownership is not related
to performance among firms that have the same proportion of inside directors.
If having many inside directors is the result of having large blocks of insider
ownership, obviously insider ownership is still important.
In the study titled "Control and Performance: Evidence from the TSE
300", Vijay Jog, a distinguished business scholar at Carleton University, and
Ajit Tulpule of Corporate Renaissance Group (Ottawa), show that the
percentage returns provided to shareholders between 1977 and 1991 by closely
held and widely held Canadian firms are similar. Jog and Tulpule's results do
not prove that ownership structure has no effect on share prices. Their result is
consistent with the share prices of, say, widely held firms being depressed relative
to the shares of closely held firms by the same amount throughout the time period
they examine. However, their results do show that intensified competition due
to freer international trade has neither helped nor harmed either closely held
or widely held Canadian firms disproportionately. Jog and Tulpule also present
comparisons of various accounting performance measures for closely held and
widely held firms, and report that there is again no difference. This analysis is
not strictly comparable to that of Rao and Lee-Sing, since they use perfor-
mance measures relative to size and industry benchmarks, while the accounting
performance measures used here by Jog and Tulpule are unadjusted.
Giovanni Barone-Adesi, the Peter Pocklington Professor of Free
Enterprise at the University of Alberta, comments that he is not surprised that
results found for the United States do not hold up in Canada. There are
numerous institutional differences between the two countries. Canadian
managers are free of class-action suits by shareholders; they can use dual-class
shares to retain control despite issuing large amounts of equity; they are less at
risk from hostile takeovers because of coattail provisions; etc. Because of these
differences, Barone-Adesi argues that most Canadian managers are well
protected from shareholders, regardless of the structure of their boards or the
distribution of their companies' shares. He further argues that making it easier
for shareholders to bring class-action suits would help remedy this. He also
points out that France assigns special legal duties to dominant shareholders,
and argues that this might be appropriate for Canada as well.
22
CANADIAN CORPORATE GOVERNANCE: THE CHALLENGE
David Stangeland of the Drake School of Management at the University
of Manitoba comments that Jog and Tulpule make an important contribution
by emphasizing that relationships between corporate governance characteristics
of firms and their performance may change over time as institutions evolve
and as competitive pressures change. He also points out that ownership
structure and other corporate-governance-related firm characteristics are very
different in different industries, and he argues that studies in this area must
measure these features relative to industry norms.
THE BOARD AND BEYON D
As MENTIONED EARLIER, THE PRINCIPAL GOAL of any corporate governance system
is to solve the problem of delegated power from shareholders to directors and
managers. One of the striking features of the recent debate over corporate
governance is the extent to which it has focussed narrowly on the board of
directors at the expense of other instruments. While the board of directors is
admittedly the legal command centre of the corporation, it is clear that there
are other organizational and market mechanisms that can attenuate agency
problems in the modern corporation. For instance, commencing with Henry
Manne, there has been a growing recognition of the capacity of the market for
corporate control to monitor and discipline managers in widely held corpora-
tions. More recently, corporate scholars have looked to nuanced executive
compensation arrangements and selective intervention by institutional
investors as means for aligning shareholder and managerial interests.
The study "Alternative Mechanisms for Corporate Governance and
Board Composition" by Jean-Marie Gagnon of the Universite Laval, and
Josee St-Pierre of the Universite du Quebec, adopts a holistic view of the
system of corporate control. The authors invoke a cost-benefit analysis to
evaluate the efficacy of alternative mechanisms for ensuring accountability,
then develop a taxonomy that links alternative systems of control with different
underlying corporate structures. Motivating the analysis is the belief that
shareholders will, within bounds, seek to adopt the most efficient means to
control managerial behaviour. The results of the analysis by Gagnon and
St-Pierre show that the distribution of voting rights in the corporation does
affect the precise means selected for control. In particular, they find that in
widely held corporations, the ratio of outside to inside board members increases
with the stock holdings of important outside shareholders (suggesting directorial
appointments as a means to monitor performance) and decreases with the
holdings of inside directors (suggesting entrenchment).
One of the key implications of the Gagnon and St-Pierre study is to
remind us that markets, albeit not always unerringly, are capable of devising
systems of control that support a multitude of corporate activities without having
to rely on external governmental intervention. In other words, so long as
23
DANIELS & MORCK
shareholders are able to access timely and accurate information about corporate
structure and performance, they should be able to pressure managers and
controlling shareholders to offer governance arrangements that are welfare
enhancing, thereby reducing the need for potentially destabilizing government
intervention. The propensity of markets to solve governance problems is
instructive, and worth bearing in mind when the nature and scope for legislation
in this area is contemplated.
The study "Executive Compensation and Firm Value" by University of
Toronto management professors Ramy Elitzur and Paul Halpern, involves a
systematic investigation of the compensation practices of public Canadian
companies, drawing on data generated pursuant to the recently enacted
amendments to the Regulations under the Ontario Securities Act. Elitzur and
Halpern cite data from the United States which shows the existence of a small
but positive relationship between the introduction of incentive-based
compensation arrangements (both short- and long-term) and share price
increases (Bhagat et aL, 1985; Brickley et al., 1985; Larcker, 1983; and
Tehranian & Waegelin, 1986). They also discuss studies that track a broader
range of firm-performance measures after the introduction of incentive-based
compensation arrangements, and find that a link exists between these schemes
and firm-performance improvements (Abowd, 1990). Nevertheless, their
research notes the existence of managerial earnings manipulation in cases
where accounting-based rather than market-based financial criteria are used to
undergird compensation schemes.
Elitzur and Halpern's empirical study focuses on the executive compen-
sation practices of a sample of 180 companies from the TSE 300 Index. Their
focus is on the difference in compensation practices between closely held and
widely held Canadian firms. In the case of a closely held firm, where manage-
ment already has a significant equity stake in the firm, conditioning compen-
sation on share-price changes would be superfluous and may even subject
managers to excessive levels of risk. However, the case for performance-based
compensation is stronger where the manager holds less stock, for instance,
where the company is widely held or where it is closely held but control is
secured by a dual-class share structure.
Elitzur and Halpern find that, regardless of ownership concentration,
both salary and total compensation are positively related to firm size.
However, in contrast to earlier studies, they find that performance, measured
either by accounting-, cash flow-, or market-based variables, has no effect on
the level of bonus, salary or total compensation for either closely held or
widely held firms. Further, their research did not find any relationship
between the percentage change in compensation and firm performance variables,
although they did find that compensation was positively influenced by the
existence of a poison pill. This latter effect seems to be stronger in the case
of closely held corporations, and suggests entrenchment. One final important
difference between closely held and widely held firms is the extent to which
24
CANADIAN CORPORATE GOVERNANCE: THE CHALLENGE
compensation levels persist despite a change in performance; the persistence
effect is greater in the closely held firms.
In the study titled "In High Gear: A Case Study of the Hees-Edper
Firms", Ron Daniels, Dean of Law at the University of Toronto, Randall
Morck of the Faculty of Business at the University of Alberta, and David
Stangeland of the Drake School of Management at the University of
Manitoba, point out that a conglomerate structure allows profitable subsidiaries
to bail out troubled ones (within the conglomerate) and so makes high(er)-
risk business strategies viable. They find that, while Hees-Edper companies
performed no better than independent firms of similar size in the same industries,
they were exposed to much higher risks. They find that this is due to both
higher leverage and higher-risk business strategies. Higher leverage is desirable
for a firm because interest payments are tax deductible while dividend payments
are not, but may serve little social purpose. Higher-risk business strategies may
serve the national interest if, as many critics argue, Canadian business is overly
conservative. They argue that Canadian public policy should not aim to
discourage the formation of conglomerates.
The study by Paul Halpern and Vijay Jog, "Bell Canada Enterprises:
Wealth Creation or Destruction?", tracks the performance of BCE, Canada's
largest conglomerate, since its establishment in 1983. The authors focus
specifically on BCE's growth and strategic direction and its value creation
performance. This inquiry is salient; BCE is insulated from the discipline of
the takeover market by virtue of its sheer size (it had $37 billion in assets in
1993) and certain regulatory impediments that limit foreign ownership. Also,
it is important to determine whether and how shareholders have controlled
managerial accountability problems in the conglomerate, especially given the
company's dispersed shareholdings. As mentioned earlier, there is an extensive
body of literature that argues that financial diversification and other motivations
for the conglomerate structure are suspect from a social welfare perspective,
and that they may simply reflect entrenched management's desire to stabilize
the firm by diversifying it, or to buy the growing industries, even though this is
not valued by shareholders.
Using a variety of measurement criteria, Halpern and Jog find that BCE
has demonstrated marginal performance over the period since its inception.
During the first six years, BCE operated like a large conglomerate, purchasing
and establishing companies, many of which were in areas unrelated to its core
activities. The source of financing for these transactions was the firm's steady
supply of cash, secured through mature product lines and a highly regulated
telecommunications' franchise. These results seem to confirm the existence of
entrenchment behaviour. Halpern and Jog do find, however, that BCE's
performance in the post-1989 period has improved (involving a return to
more focused corporate growth), and seems to have attracted some modest
recognition by shareholders. Nevertheless, shareholder support for the conglom-
erate is not as great as it could be, and the authors argue that this discount
25
DANIELS & MORCK
reflects the market's concern that management will repeat the mistakes of the
past, namely by diverting free cash flows (kicked off by the regulated telecom-
munications monopoly) to investments in wholly owned or portfolio companies
in related and semi-related businesses across the world. Further exacerbating the
company's problems is its strategic commitment to the management of assets, not
businesses.
Clifford Holderness of Boston College, a leading expert on closely held
firms in the United States, comments that closely held firms are more important
in the United States than is commonly realized, although they are not nearly
as pervasive as in Canada. He argues that dominant shareholders can and do
extract private benefits from the firms they control, but there must be constraints
on this. Otherwise, closely held firms would eventually disappear, and this is
not happening, even in the United States.
Vikas Mehrotra of the Faculty of Business at the University of Alberta
comments that sole federal jurisdiction in areas related to corporate gover-
nance is needed to prevent managers reincorporating their firms in provinces
that provide comfortable protection for insiders. He advances the argument
that disputes between dominant shareholders and small shareholders are the
central issue in Canadian corporate governance.
INSTITUTIONAL INVESTORS
ONE OF THE MOST IMPORTANT CHANGES TO CANADIAN capital markets, and
one which was frequently discussed during the conference, is the growing
significance of institutional owners. Spurred on by the retirement needs of an
aging population, large institutional investors (public and private pension
funds, mutual funds, insurance companies and banks) are blessed with ample
pools of capital that they have directed to investment in Canadian corporate
equity. For many Canadian and American commentators, the growth of
institutional ownership has profound implications for the control of managerial
behaviour in both widely held and closely held corporations. The claim is simple.
By virtue of their size, sophistication, and staying power, institutional investors
are capable of monitoring and disciplining both managers and controlling
shareholders. Nevertheless, the claim for institutional ownership is not with-
out its detractors. Many, such as Lakonishok et al. (1992), argue that the scope
for vigorous institutional activism is hobbled by a range of legal and organiza-
tional problems within institutional investors themselves.
The studies commissioned for this volume address the scope for, and
prospects of, institutional activism from a number of different perspectives.
"Do Institutional and Controlling Shareholders Increase Corporate Value?", the
study by Jeffrey Macintosh of the Faculty of Law, University of Toronto and
Lawrence Schwartz, a consulting economist in the private sector, addresses the
effects of both institutional and controlling shareholders on corporate value.
In the case of the former, the authors suggest that institutional shareholders will
26
CANADIAN CORPORATE GOVERNANCE: THE CHALLENGE
increase corporate value, although they concede the possibility that institutional
clout may be co-opted into being an unwitting ally of opportunistic manage-
ment. In the case of the latter, Macintosh and Schwartz are more agnostic; they
predict that controlling shareholders will engage both in more effective
monitoring of managers than non-control shareholders, and in redistributive
transactions that shift wealth from non-controllers to controllers. Given the
prospects of controlling shareholder opportunism in the form of wealth
redistributive transactions, Macintosh and Schwartz predict that institutional
investors will make controlling shareholders, not just management, a target of
their monitoring.
To test their hypotheses, Macintosh and Schwartz examined several
different performance measures for TSE 300 firms, which they correlate with
data on Canadian patterns of institutional ownership. They find a positive and
statistically significant relationship between both return on assets and return on
equity and institutional holdings. They also find some support for the hypothesis
that institutional monitoring tends to mitigate the danger of redistributive
transactions engineered by controlling interests. While the presence of institu-
tional investors is correlated with increased corporate value, the relationship
between controlling shareholders and corporate value is more ambiguous.
Macintosh and Schwartz find that, although firms with controlling shareholders
generate higher profits, these profits are siphoned off by controlling shareholders.
The study "Institutional Activism by Public Pension Funds: The
CalPERS Model in Canada?" by Steven Foerster of the Business School at the
University of Western Ontario considers the desirability of importing into
Canada the highly interventionist pattern of shareholder activism used by the
California Public Employees' Retirement System (CalPERS) in the United
States. CalPERS is not only the largest public pension fund in the United
States (assets of US$ 80 billion), but also the most activist. Over the last
several years, CalPERS has trained its sights on the most poorly performing
corporations in the United States. Each year, the fund identifies performance
laggards, and subjects the management of those firms to increasing pressure
(ranging from quiet, behind-the-scenes diplomacy to more public and stinging
forms of intervention, such as "Just Vote 'No'" proxy campaigns). Foerster
cites a recent study by Nesbitt (1994) that found that whereas the targets of
CalPERS activism underperformed the S&P Index by 60 percent prior to
CalPERS' involvement, post-intervention returns jumped dramatically - out-
performing the index by 40 percent.
Given the returns alleged to derive from CalPERS' activism, Foerster
considers why Canadian investors have yet to embrace this model of intervention.
On the basis of extensive interviews conducted with Canadian public fund
managers, Foerster traces the lack of enthusiasm for CalPERS' style intervention
to differences in "style", rather than to any fundamental difference in the
underlying regulatory structure of the two countries. That is, Canadian institu-
tional investors systematically favour a less confrontational style of dealing
27
DANIELS & MORCK
with performance problems than is manifest in the United States. Another
interesting point Foerster raises is that CalPERS systematically avoids closely
held firms. Since most Canadian firms fall into this category, it may be that
CalPERS' style of intervention is not well suited to dealing with managers who
are also dominant shareholders. It is important not to overstate the commit-
ment to tacit pressure and activism; even Foerster acknowledges the activism
of Canadian institutions (spearheaded by Fairvest) in responding to unfair
management initiated transactions.
In the study titled "Monitoring Incentives Facing Institutional Investors",
Michel Patry of the Ecole des Hautes Etudes Commerciales and Michel Poitevin
of the Universite de Montreal analyze the governance of institutional investors
themselves, and note that employees usually have little influence on how their
pension money is invested. They point out that in defined benefit pension
funds, the sponsoring organization has an incentive to maximize the fund's
return. However, that sponsor is often a corporation with its own governance
problems. Where the sponsor is not a corporation, it is usually a government
and therefore subject to the political favour trading that plagues the public
sector in Canada.
Several studies in the United States agree that portfolios of corporate
pension funds perform surprisingly poorly compared to both broad market
indices and mutual funds. The portfolio performance of public pension funds is
even worse.
In a summary of this evidence, Lakonishok et al. (1992) agree that
corporate treasurers, usually responsible for managing corporate pension funds,
are more likely to favour hiring outside portfolio managers than indexing.
Hiring outside managers gives them someone to blame for poor performance,
but still provides work for the corporate treasurer's department. Outside managers
must be evaluated, hired and fired; indexing is too easy a way to earn higher
returns, and would not justify a large bureaucracy in the corporate treasurer's
office. Since neither plan beneficiaries nor public shareholders have any input,
the interests of corporate treasurers take precedence.
Romano (1993) argues that similar circumstances prevail in public-sector
pension funds, with neither beneficiaries nor taxpayers having any serious
input into how their plans are managed. Hiring, firing, and evaluating outside
managers gives inside managers the same advantages as those in corporate
pension funds have. However, the additional complication of pressure to put
money into politically favoured investments may reduce performance even
further in public-sector funds.
Lakonishok et al. (1992) continue that outside portfolio managers for
pension funds are usually compensated with a management fee similar to that
charged by mutual funds. Thus, outside managers' incentives are to acquire and
keep a large portfolio to manage; increasing its value is less important to them.
Lakonishok et al. (1992) discuss evidence that corporate treasurers use
quarterly performance to evaluate the performance of outside portfolio managers.
28
CANADIAN CORPORATE GOVERNANCE: THE CHALLENGE
However, stock prices are known to exhibit mean reversion, that is, unusually
low stocks tend to go up and unusually high stocks tend to go down. Thus,
corporate treasurers systematically buy high and sell low. Also, there is evidence
that portfolio managers systematically sell "dog" stocks and replace them with
"high flying" stocks at quarterly reporting times. Having a few good stocks in
one's portfolio apparently impresses sponsors. Again, buying high and selling
low is not a recommended formula for financial success. Overall, Lakonishok
et at. (1992) conclude, poor governance within pension funds results in sponsors
reallocating their funds' assets among portfolio managers too often, in portfolio
managers trading too often, and in poor overall returns.
Patry and Poitevin suggest that few institutional investors have the
expertise necessary to intervene in the management of firms whose shares they
own. They also argue that the governance problems within pension funds
must be resolved before pension funds can be expected to improve the gover-
nance of corporations.
Patry and Poitevin contend that the current provision in the Income Tax
Act that restricts pension funds to investing no more than 20 percent of their
portfolios abroad probably does not contribute to better corporate governance.
A primary effect of this rule is to prevent pension funds from selling out of
poorly performing Canadian firms for lack of better alternative investments.
Although this might force the funds to voice their concerns to the managers
of such firms, it also reduces the power of pension funds to affect such firms by
dumping their stock. Patry and Poitevin consider mandatory indexing, a flip
tax on capital gains, greater legal liability for pension-fund managers, profes-
sional monitors, relational investing, and better disclosure of pension-fund
managers' compensation, and of pension funds' risk and comparative perfor-
mance. They discuss the pros and cons of these proposals in detail.
Brian Smith and Ben Amoako-Adu of the Department of Finance at
Wilfrid Laurier University, contribute the study on "Outside Financial
Directors and Corporate Governance". They find no consistent pattern relating
the performance of Canadian firms to the presence on their boards of directors
affiliated with financial institutions. This study is actually more wide-ranging
than its title suggests. The authors also examine insider ownership. Perhaps
because of the paucity of widely held Canadian firms and the limited disclosure
of their owners' stakes, no statistically discernable pattern is found in the data
for firms with insider ownership below 20 percent. Among firms with more
than 20 percent managerial ownership, they find a positive relationship with
share value similar to that found by Morck et al. (1988) in U.S. data. They
also find no statistically discernable relation between the number of outsiders
on the board and firm performance. However, they advance the important
point that Canadian disclosure rules fail to establish who really is an outsider.
A firm's lawyers, executives of its advertising firm, and executives of companies
that do business with it are not really outsiders, yet are so classified when they
sit on its board. Their fear of losing business with the firm may deter such
29
DANIELS & MORCK
directors from challenging the CEO. Perhaps the rules defining outside
directors should be more stringent.
Mark Huson of the Faculty of Business at the University of Alberta makes
some perceptive comments on the studies in Part IV, pointing out that the
defining theme of corporate governance in Canada is controlling shareholders,
not unsupervised managers as in the United States. Huson suggests that
Amoaku-Adu and Smith do not distinguish which directors are controlled by
management and which are really independent. Because of such problems in
most studies of outside directors, he argues that government ought to hold off
from requiring certain numbers of outsiders on boards.
Huson adds, however, that Amoaku-Adu and Smith's results are consistent
with dominant shareholders extracting disproportionate income from firms
they control. Can institutional investors limit this? Huson points out some
econometric problems in Foerster's analysis relating to the use of returns
excluding dividends. He argues that the analysis by Macintosh and Schwartz
does not distinguish between the possibility that institutional investors
improve share values and the possibility that improved share values attract
institutions. But despite these problems, he contends that institutional
investors probably do improve corporate governance in Canada. Huson does
not support Patry and Poitevin s idea of flip taxes to reduce churning in pension
funds. In his opinion funds should be free to divest themselves of investments
in poorly run firms, and the thinness of Canadian markets already constitutes
a barrier to this.
Huson advocates instead measures to reduce pension funds' costs in
confronting corporate governance problems. He proposes that pension funds
be allowed to communicate among themselves on corporate governance
issues, that institutional investors not be classified as controlling shareholders,
that valuation techniques in shareholder appraisal rights take into account
value lost due to things like poison pill adoptions, that outside board members'
pay be linked to share prices, and that pension funds be allowed to invest
abroad with no restrictions.
Michael Weisbach of the University of Arizona, one of the foremost
experts on boards of directors in the United States, suggests that Canadians
should not be too quick to imitate practices south of the boarder. Weisbach
suggests that activist U.S. pension funds like CalPERS are overrated, and he
argues that their apparent success may be due to mean reversion in stock
prices rather than to any real effect on corporate governance. He points out
that studies finding a statistical link between good company performance and
institutional investors as shareholders may not indicate a beneficial effect of
these investors on corporate governance. Rather, they may just be detecting
fund managers rushing to buy winners so their quarterly portfolio reports look
good. He supports Patry and Poitevin's call for increased indexing of pension
funds' portfolios, although he stops short of calling for mandatory indexing.
Weisbach argues that the current rule forcing Canadian pension funds to
30
CANADIAN CORPORATE GOVERNANCE: THE CHALLENGE
invest in Canada prevents them from diversifying as much as they should. He
points out that Canadian pension funds should invest very little in Canada in
order to insulate themselves from the Canadian business cycle.
INTERNATIONAL ASPECTS OF CORPORATE GOVERNANCE
The study titled "The Corporate Governance of Multinationals", by Randall
Morck of the Faculty of Business at the University of Alberta and Bernard
Yeung of the School of Business Administration at the University of Michigan,
argues for a minimalist approach to corporate governance legislation.
Globalization is fast making heavy-handed legislation of business impractical.
Canada is in competition with other countries for capital, knowledgeable
workers, and high value-added operations of multinationals. If the Canadian
government passes onerous laws, Canada will simply lose out to more friendly
jurisdictions. For this reason, the subsidiaries of foreign multinationals should
be treated like other Canadian companies with dominant shareholders.
The same globalization process means that Canadian companies will be
exposed to more bracing competition from abroad in coming years, and poor
corporate governance will therefore be more costly. Rather than attempt to
micromanage boards of directors, the emphasis should be on empowering
shareholders. Increasing shareholder power is a more flexible and more effective
way to improve corporate governance and thereby make Canadian firms globally
competitive.
The corporate governance issue being addressed here is the fair treat-
ment of minority shareholders by the dominant shareholder — in this case, the
foreign parent company. One way of achieving fair treatment is to require that
the boards of all closely held firms, including foreign controlled subsidiaries,
have conduct committees charged with monitoring non-arm's-length trans-
actions. These committees must have a majority of outside directors. If there is
a political necessity, they could also be required to be Canadian citizens,
although there is little economic rationale for this as long as they are sueable
in Canada. It would also make sense that all closely held firms, including
partially owned subsidiaries, disclose the details of all their non-arm's-length
transactions, and that their small shareholders have the right to launch class-
action suits against the dominant shareholder in cases of oppression.
The study by Lewis Johnson and Ted Neave of the School of Business at
Queen's University, "Corporate Governance and Supervision of the Financial
System", systematically links the governance structures of financial and market
intermediaries with the asset and liability mix of various institutions. The
authors do so by drawing on a transaction cost framework developed by Oliver
Williamson. They demonstrate a linkage between the nature of assets and
liabilities and the complexity and transparency of the monitoring arrange-
ments that obtain across different financial and market intermediaries. To
strengthen the performance of intermediaries, Johnson and Neave recommend
31
v
greater reliance on mandatory production of information where asset valuations,
liability valuations, or contingent risks are now opaque. Furthermore, they
endorse the value of continuing information release, rather than sudden
announcements of dramatic change.
"Banks and Corporate Governance in Canada", by Randall Morck of
the Faculty of Business at the University of Alberta and Masao Nakamura of
the Faculty of Commerce and Business Administration at the University of
British Columbia, describes the somewhat checkered history of the German
and Japanese banking systems. They argue that the alleged benefits of banks
as major shareholders are unlikely to materialize in Canada; indeed, they
suggest that the benefits are far from clear even in Germany and Japan. The
authors point out that Canadian banks are a poor choice for a watchdog as
they, almost alone among Canadian firms, are widely held and subject to all
the inefficiencies that implies. They recommend no move toward increasing the
role of Canadian banks in the corporate governance of non-financial firms.
Roberta Romano of Yale Law School, a leading expert in corporate
governance issues, comments that individuals have been remarkably creative
throughout history in structuring institutions to evade regulation. Although
she is concerned about the distribution effects of subsidies to higher education
and university research, which mainly benefit the upper middle class, she
agrees with Morck and Yeung that excessive regulation is to be avoided,
especially in an increasingly global economy. She points to the beginnings of
the Eurobond market as an example. The United States imposed taxes on
foreign bonds in the early 1960s with the result that the market simply
moved abroad. Overly heavy-handed corporate governance regulations
would simply fuel the search for administrative and legal structures that
evade them. She agrees with Morck and Nakamura that the Japanese and
German banking systems are probably not to be imitated. She points out
that the banking systems were in place long before the rapid post-war growth
of these two countries, and argues that a latecomer advantage (learning from
others' mistakes), and perhaps things like good education and a high savings
rate, were more important than corporate governance to that growth. She
argues that German banking was more an effect than a cause of that country's
rapid growth. Germany industrialized late, when the logistics of large-scale
manufacturing had been worked out by others. The optimal scale for
German industry was therefore larger than it had been during the period of
industrialization in England and France. The need for large blocks of capital
may have built the banks, and not the reverse. Romano points out that co-
determination in Germany assigns half the seats on a firm's supervisory board
or Aufsichtsrat to labour representatives. With half the board definitely not
representing shareholders and the other half doing so weakly at best, widely
held ownership is unlikely to be optimal in Germany, so another system had to
be developed. This, she argues, shows that corporate governance must be
thought of in the context of a country's overall economic system. German- and
32
DANIELS & MORCK
CANADIAN CORPORATE GOVERNANCE: THE CHALLENGE
Japanese-style banking makes little sense here, given the rest of Canada's
economic system.
In his comments Adrian Tschoegl, a noted expert on the Japanese
economy who teaches at the University of Pennsylvania's Wharton School,
emphasizes that small countries can be home bases for global companies, and
suggests that this is a feasible future for Canada. He then considers the roles of
debt and equity, and how they differ across countries, and argues that these
considerations might supplement the analysis in Johnson and Neave's study.
Tschoegl points to the work of Allen (1993), who argues that debt financing is
acceptable to investors in mature industries, where monitoring is easy. In
newer, less well-understood industries, where monitoring is difficult, equity is
predominant. He then provides insightful summaries of recent work on the
roles of markets and intermediaries, and again connects them to Johnson and
Neave's study. Finally, he comments at length on the study by Morck and
Nakamura, pointing out that Japanese keiretsu are unique among conglomerates
for their mutual cross holding. Firms own stock in each other collectively, and
no single firm may dominate. Japanese banks are owned by other keiretsu firms.
He adds that vertical monitoring is important in Japan: that firms monitor the
governance of their suppliers. In a discussion of the history of Japanese banking,
he notes that in the late 19th century Japan copied U.S. banking regulations,
but was dissatisfied with the results. A series of modifications modelled on
Belgian, British, and French laws and institutions followed, and led to the
present system. Thus, the current similarities of the Japanese and German
systems result from both retaining aspects of earlier practice that was once
common to much of Europe.
CORPORATE GOVERNANCE AND SOCIAL RESPONSIBILITY
ALTHOUGH A CONSIDERABLE AMOUNT OF INK HAS BEEN SPILLED on the issue of
the social responsibilities of directors, specifically the extent to which directors
owe duties to non-shareholder constituencies, the fact remains that public
concern with corporate misconduct typically involves a failure on the part of
the corporation to adhere to explicitly legislated duties and responsibilities. In
other words, the only question is how to ensure board compliance with
legislatively prescribed goals, not whether these goals should be pursued in the
first place.
The study by Ronald Daniels and Robert Howse of the Faculty of Law at
the University of Toronto, "Rewarding Whistleblowers: The Costs and
Benefits of an Incentive-Based Compliance Strategy", focuses on one relatively
under-utilized (in the Canadian context) instrument for achieving corporate
compliance with social responsibilities: whistleblower bounties. Daniels and
Howse argue that whistleblower bounties have considerable potential to
become a cost-effective means to enforce legislated responsibilities. The authors
provide two principal arguments in favour of such bounties. First, whistleblower
33
DANIELS & MORCK
bounties increase the effectiveness of sanctions by raising the probability that
misconduct will be detected, which means that the state will not have to use
excessive sanctions to secure social optimal penalties (the product of both
probability of detection and quantum of penalty). Second, whistleblower
bounties take advantage of existing information and control systems within
the corporation, thereby reducing the need for the state to establish more costly
and, ultimately, less effective external monitoring systems.
Nevertheless, despite the arguments in their favour, state reliance on
whistleblower bounties has proved to be extremely controversial. In the
United States, for instance, there has been intense criticism of the bounties
provided by the federal government under the False Claims Act. Critics allege
that the existence of such bounties distorts internal information flows, causes
managers to make lower-level employees over-invest in firm-specific capital
so as to magnify the downside costs of whistleblowing, and subverts the
ability of managers to create durable commitments to firm culture and team-
work. Daniels and Howse find these concerns overstated, and argue that careful
design and enforcement of whistleblower incentives can correct for many of
these problems.
The comment on the Daniels and Howse study by Jennifer Arlen of the
University of Southern California Law Center is sympathetic to enhanced
reliance on whistleblower bounties in the control of corporate crime.
However, she stresses that bounty provisions should not be enacted unless
accompanied by a thorough reform of criminal law. Arlen is concerned with
the interrelationship between bounties and the background system of criminal
and quasi-criminal sanctions. Specifically, if bounty awards are employed in
circumstances where corporations are in essence absolutely liable for agents'
crimes, the awards may result in increased corporate crime because they may
reduce the corporation's own efforts to reduce crime. Arlen thus argues for the
adoption of alternative corporate liability rules, such as mitigation rules,
negligence-based corporate liability or an evidentiary privilege .
The study, "Patient Capital? R&D Investment in Canada", by Ron
Giammarino of the Faculty of Commerce and Business Administration at the
University of British Columbia, is a thorough summary of recent research on
links between R&D spending and firms' share values. Corporate managers
often complain that focusing on maximizing share value necessitates adopting
a short-term planning horizon and forsaking long-term investments like R&D.
The inescapable conclusion of this study is that this complaint is bunk! High
R&D spending is statistically significantly related to above-average market-to-
book ratios (i.e., high share prices). We know increased R&D spending causes
share prices to rise and not the reverse; stock prices have been found to rise
significantly immediately upon announcements by high-tech firms of
increased R&D. Interestingly, this is true even when the firms in question
have quarterly operating losses. (However, in older industries, R&D - which
might rationally be seen as less valuable - does not increase, and may even
34
CANADIAN CORPORATE GOVERNANCE: THE CHALLENG E
reduce, share value.) Still, in general, shareholders like long-term investments
and would like to see more.
Why then do firms not make their shareholders happier and richer by
spending more on R&D than they do? Perhaps managers skimp on R&D
because long-term investments invite takeovers? If this were true, we should
expect to find higher R&.D in firms protected by poison pills and other anti-
takeover defences. In fact, after firms adopt such defences, R&.D tends to fall
significantly. Moreover, leveraged buyouts (LBOs), the type of takeover most
likely to divert earnings away from long-term investments, are extraordinarily
rare in industries where R&D spending is important. This suggests that raiders
usually stay away from high-R&D firms. When two R&D-intensive firms
merge, the R&D spending of the resulting firm is often lower than the combined
R&D of the two merged firms. But this is not evidence of inefficiency. In fact,
the motive for such mergers is often the savings attainable by rationalizing
activities like R&D.
The answer, Giammarino argues, may lie in how firms make investment
decisions. A recent survey by Jog & Srivastava shows that many Canadian
firms use out-of-date data and conceptually flawed capital budgeting decision
criteria in evaluating long-term projects. Accounting rates of return and pay-
back periods fall into this category. Even firms that use conceptually valid
methods such as net present values or internal rates of return often do not
employ them correctly. For example, an R&D investment typically involves
high risk in the early stages, but much greater certainty as the project develops.
This means a high discount rate should be used for the first cash flows, but a
lower rate is appropriate for cash flows in the more distant future. Failing to do
this is likely to bias firms against R&D spending. Also, it is important to
recognize that R&D investments have many of the characteristics of options.
There is a small chance of a big payoff, as in a call option. Valuation tech-
niques for options are complex, but evaluating R&D projects as options would
tend to give them higher values than standard net-present-value analyses
would assign. Managers' understandable aversion to high-risk investments like
R&D perhaps encourages them not to question the negative verdicts simpler
decision criteria produce.
Overall, the best way to increase R&D spending may be to educate
managers how to value R&D properly. Improve corporate governance in this
dimension, and firms' R&D spending will then take care of itself.
The study "Corporate Governance and Worker Education: An
Alternative View" by Alice Nakamura, of the Faculty of Business at the
University of Alberta, John Cragg, of the Faculty of Arts at the University of
British Columbia and Kathleen Sayers of International Wordsmiths, points
out that businesses may be loath to invest in worker training because of cost
considerations. For example, employees whose training firms pay for may leave
for highly paid jobs elsewhere, thus denying the firms a return on their invest-
ment in training. Yet in the new global economy, as Morck and Yeung point
35
DANIELS & MORCK
out, continuous innovation is critical to success. This requires that highly
educated employees create innovations and continuous education for other
employees to apply innovations. The solution to this under-investment in
training by employers is either to require workers to pay for their own training
or to provide public education. For egalitarian and other reasons, Canada has
focused on the latter. Because of the fiscal problems now facing all levels of
government in Canada, this decision is being re-evaluated.
Can firms be encouraged to invest more in training through changes in
corporate governance — for example by requiring worker representation on
boards? The answer given by Nakamura, Cragg and Sayers is "perhaps, but
other approaches would be much preferred". First, requiring workers on boards
might drive investment out of Canada. Second, employee representatives on
boards would protect the interests of existing employees, especially senior
employees, but would see little point in encouraging firms to spend money
training those who are now unemployed. Yet this is where the greatest social
need is. We add a third reason; worker representation on boards might also
simply marginalize boards. In Germany, which has mandatory labour represen-
tation on the Aufsichtsrat, or supervisory board, major decision-making has
been transplanted to the Vorstand, or management board, which consists of
directors who are also top executives. The Aufsichtsrat has become an ornament.
Nakamura, Cragg and Sayers argue that focusing on an alleged training
deficit detracts from other problems underlying many of Canada's social ills.
One such problem is the structure of Unemployment Insurance. They argue
for UI reforms that would allow employers who commit to job security to pay
lower UI taxes. Another underlying problem is the promotion of students who
are illiterate from grade to grade through primary and secondary schools. The
costs of correcting 12 wasted years are immense. Better monitoring of both
student achievements and teacher performance are imperative. A third under-
lying problem is the prohibitive expense of lengthy post-secondary education
for students from poor families and for students supporting families. Some type
of government subsidy would seem reasonable here. Finally, students often
have little information about what particular training would make them most
employable. To correct this, the authors argue that information on the
employment and average earnings of graduates of various post secondary
programs should be made available to the public.
Although Michael Trebilcock of the Faculty of Law at the University of
Toronto shares the skepticism of Nakamura et al. surrounding the scope for
worker-based governance structures to address job training objectives, he is
less sympathetic to the various alternative policy initiatives they prescribe.
Trebilcock believes the instruments suggested by Nakamura et al. are not in
themselves sufficient to address the need for job training or retraining. As a
starting point for policy reform, Trebilcock stresses the need to disaggregate
the potential demanders of job training or retraining services on the grounds
that the appropriate policy mix for each constituency can vary dramatically. In
36
CANADIAN CORPORATE GOVERNANCE: TH E CHALLENGE
determining the desired policy response for each group, Trebilcock argues for
more supply-side competition in training programmes. Trebilcock worries
about the excessive amount of centralized control exerted by the federal govern-
ment (through, for instance, the purchase of seats in community colleges).
Ultimately, Trebilcock believes that the most effective way to encourage
competition is through a demand-side scheme, such as that associated with
school voucher programmes.
Consulting economist Ron Hirshhorn's study, "The Governance of
Nonprofits", is useful in identifying the scope for governance problems in the
third or not-for-profit sector. The analysis is salient, given the growing reliance
of the state on nonprofit organizations to deliver a range of public goods and
services. Hirshhorn's analysis draws on a well-developed literature of organiza -
tional theory that explains the nonprofit organization as a response to market
failures that are too costly to solve by either for-profit or state providers. In
particular, Hirshhorn focuses on the perverse incentives that for-profit delivery
introduces in areas where outputs are difficult to measure. Hirshhorn stresses
the need to evaluate the rationale for nonprofit delivery on a case-by-case
basis, having regard to the elaborate criteria that he develops. Hirshhorn
cautions that mere evidence of some gap in the operation of private and
political markets does not support reliance on non-profit modes of delivery.
The calculus boils down to one of balancing the transaction cost savings in
some areas against the increases in enforcement costs in other areas.
To illustrate this analysis, Hirshhorn examines three case studies involving
community health care, local airport authorities and universities. Whereas
Hirshhorn regards the case for nonprofit delivery to be relatively robust for
community health care, he is skeptical of its value in the airport context
(given monopoly properties) and universities (given the complexity of defining
and measuring outputs).
Hirshhorn proposes several different avenues of policy reform designed to
strengthen the operation of no-profits. He argues generally for independent
management reviews of nonprofit performance and for the imposition of stringent
reporting requirements. Hirshhorn acknowledges that securing performance
improvements through the adoption of these measures will not be easy. The
difficulty is that many nonprofit services cannot be precisely defined and are even
more difficult to monitor. Indeed, it is often the case that these very properties are
the reason why for-profit providers eschew provision of these services.
Nevertheless, Hirshhorn is optimistic that precise performance measurements
can be created for a number of nonprofit services, improving significantly the
degree of product market pressure that can be directed at these providers. In any
event, Hirshhorn's focus on the role of the state in providing information on
nonprofits is well-placed. It is clear that in the globalized economy of the next
millennium, the role of the state will increasingly shift from that of a direct
producer of goods and services to that of an external monitor of, and a supplier of
information on, goods and services produced by others, particularly nonprofits.
37
DANIELS & MORCK
One particularly interesting set of recommendations favoured by
Hirshhorn relates to the paucity of effective governance provisions in the
Canada Corporations Act, Part II (CCA), which applies to Canadian nonprofits.
In contrast to the relatively crisp lines of accountability set out in standard
corporate legislation, Hirshhorn stresses the failure of nonprofit legislation to
specify comparable duties. He argues that "reasonable rules could be estab-
lished to determine those who qualify as significant stakeholders based on
their contributions to the organization".
Bruce Chapman's comment on Hirshhorn's study concentrates on the
charitable component of the non-profit sector. He develops a rationale for
non-profit delivery of charitable services that draws on a supply-side analysis,
rather than the demand-side analysis emphasized by Hirshhorn. Chapman
argues that charitable non-profits can be used to supply public goods in a way
that avoids the problematic and destabilizing political conflict that might
occur were the goods provided in the public sector. Chapman also claims that
the non-profit form not only prevents contract failure, but also permits donors
and investors on the supply side to control the specific nature of the in-kind
transfers that frequently characterize charitable non-profits. This rationale
supports the use of a disbursement obligation on charities that forces managers
of nonprofits to go back to their benefactors regularly for further funding,
thereby limiting the scope for agency drift. Chapman also argues in favour of
"line of activity" restrictions that limit the capacity of nonprofit managers to
stray from benefactor objectives.
ENDNOTES
1 M. Porter, Canada At the Crossroads: The Reality of a New Competitive
Environment, Ottawa: Business Council on National Issues/Supply and
Services Canada, 1991, pp. 10-12.
2 T. Hatzichronoglou, "Indications of Industrial Competitiveness: Results
and Limitations," in Technology and National Competitiveness, edited by J.
Niosi, Montreal: McGill-Queen's University Press, 1991, p. 191 (citing
OECD data).
3 Data from International Monetary Fund, reported in The Economist, World
Economy Survey, September 19-25, 1992, p. 17.
4 Statistical Abstract of the United States, various years.
5 Statistical Abstract of the United States, various years.
6 Michael Porter, The Competitive Advantage of Nations, New York: The Free
Press, 1990, p. 681.
7 R. J. Daniels, "Can Contractarianism be Compassionate?: Stakeholders
and Takeovers," University of Toronto Law Journal, 1994-
38
CANADIAN CORPORATE GOVERNANCE: THE CHALLENGE
8 The issue of the objective function of the firm is thoroughly canvassed in a
symposium issue of the University of Toronto Law Journal devoted to
Stakeholders and Corporate Governance.
9 347883 Alberta Ltd. v. Producers Pipelines Ltd. (1991), 80 D.L.R. (4th) 359
(Sask. C.A.); Remington Energy Ltd. v. joss Energy Ltd., unreported,
Alberta Court of Queen's Bench, per Fraser J., Dec. 17, 1993; Re MDC
Corporation and Regal Greetings & Gifts Inc. (1994) 17 OSCB 4971; and Re
Lac Minerals Ltd. and Royal Oak Mines Inc. (1994) 17 OSCB 4963.
10 Several examples can be cited. The tax incentives contained in the 1963
federal budget which lowered withholding taxes on dividends from 15 per-
cent to 10 percent for companies beneficially owned by Canadians to the
extent of at least 25 percent of their voting stock, and also where the parent
company and its associates held no more than 75 percent of the voting
shares and the stock of the subsidiary was listed on a Canadian exchange;
the establishment of the Foreign Investment Review Agency in 1974 and
its attention to Canadian share ownership as one of the criteria necessary
for entry into Canada; and the incentives set out in the Trudeau govern-
ment's National Energy Program for Canadian ownership.
BIBLIOGRAPHY
Allen, F. "Strategic Management and Financial Markets." Strategic Management journal, 14
(1993):! 1-22.
Barclay, M. and C. Holderness. "Private Benefits from Control of Public Corporations."
journal of Financial Economics, 25 (1989):371-97.
. "The Law and Large Block Trades. "Journal of Law and Economics, 35 (1992):265-94.
Barclay, M., C. Holderness and J. Pontiff. "Private Benefits from Block Ownership and
Discounts on Closed End Funds." Journa! of Finance, 33 (1993):263-91.
Beason, M. and H. Weinstein. "Growth, Targeting and Economies of Scale in Japan: 1960-
1990." Review of Economics and Statistics. 1994-
Berle, A. and G. Means. The Modern Corporation and Private Property. New York:
Macmillan Inc., 1932.
Bhagat, S., J. Brickley and R. Lease. "Incentive Effects of Stock Purchase Plans." Journal of
Accounting and Economics, 1 (1985):195-215.
Brickley, J., S. Bhagat and R. Lease. "The Impact of Long Range Managerial Compensation
Plans on Shareholder Wealth." Journal of Accounting and Economics, 7 (1985):115-29.
Comment, R. and G. Schwert. "Poison or Placebo? Evidence on the Deterence and Wealth
Effects of Modern Antitakover Measures." Journal of Financial Economics, 39, 1 (1995):3-4.
Daniels, R. "Can Contractarianism be Compassionate?: Stakeholders and Takeovers."
University of Toronto Law Journal, 1994.
Harris, D., R. Morck, J. Slemrod and B. Yeung. "Income Shifting in U.S. Multinational
Corporations." In Studies in International Taxation. Edited by A. Giovannini, G. Hubbard
and J. Slemrod. Chicago: University of Chicago Press, 1993.
39
DANIELS & MORCK
Hermalin, B. and M. Weisbach. "The Effects of Board Composition and Direct Incentives
on Firm Performance." Financial Management, 4 (1991):101-12.
Jarrell, G. and A. Poulson. "Dual Class Recapitalizations as Antitakeover Mechanisms: The
Recent Evidence." Journal o/Financial Economics, 20, 1/2 (1988):129-52.
Jensen, Michael. "The Agency Costs of Free Cash Flow." American Economic Review, 76
(1986):323-29.
Johnson, W., R. Magee, J. Nandu, J. Nagarakam, and H. Newman. "An Analysis of the
Stock Price Reactions to Sudden Executive Deaths." Journal of Accounting and
Economics, 1 (1985):151-74.
Lakonishok, J., A. Shleifer and R. Vishny. The Structure and Performance of the Money
Management Industry." Brookings Papers on Economic Activity and Microeconomics.
(1992):339-91.
Lang, L. and R. Stulz. "Does Corporate Diversification Create Value?" Unpublished
manuscript, 1992.
Larcker, D. "The Association Between Performance Plan Adoption and Corporate Capital
Investment." Journal of Accounting and Economics, 7 (1983):3-30.
Mace, M. Directors: Myth and Reality. Cambridge, MA: Harvard Business School Press, 1971.
McConnell, J. and H. Servaes. "Additional Evidence on Equity Ownership and Corporate
Value." Journal o/Financial Economics, 27, 2 (1990):595-610.
Megginson, W., R. Nash and M. Van Randenborgh. "The Financial and Operating
Performance of Newly Privatized Firms." Journal of Finance, 49 (1994):403-52.
Morck, R., A. Shleifer and R. Vishny. "Management Ownership and Corporate Performance:
An Empirical Analysis." Journal of Financial Economics, 20 (1988):293-316.
. "Alternative Mechanisms for Corporate Control." American Economic Review, 79,
4(1989):842-52.
Morck, R. and D. Stangeland. "Large Shareholders and Corporate Performance in Canada."
Unpublished manuscript, 1994.
Morck, R. and B. Yeung. "Why Investors Value Multinationality," Journal of Business, 64, 2
(1991):165-88.
. "Internalization: An Event Study." Journal of International Economics, 33
(1992):41-56.
Nesbitt, S. L. Rewards from Corporate Governance. California Public Employees Retirement
System. (Feb. 12, 1992): 1-5.
Romano, R. "Public Pension Fund Activism in Corporate Governance Reconsidered."
Columbia Law Review, 4 (1993).
Stangeland, D. "Issues in Corporate Control and the Performance of the Corporation."
Ph.D. Thesis, University of Alberta, 1994.
Tehranian, H. and J. Waegelin. "Market Reaction to Short Term Executive Compensation
Plan Adoption." Journal of Accounting and Economics, 7 (1985):131-44.
Weisbach, M. "Outside Directors and CEO Turnover." Journal of Financial Economics, 20
(1988):431-61.
40
Part II The Governance of
Canadian Corporations
This page intentionally left blank 
P. Someshwar Rao & Clifton R. Lee-Sing O
Industry and Science Policy Sector, Strategic Investment Analysis L*
Industry Canada
Governance Structure, Corporate
Decision-Making and Firm Performance
in North America
INTRODUCTION
F
OR THE PAST DECADE, a number of inter-related global trends have been
changing the world economy in a remarkable and fundamental way. The
enormity of these changes has been paralleled only by those experienced
during the nineteenth century Industrial Revolution. These developments
include rapid product, process and organizational innovations; shorter product
cycles; increased pace of business globalization; marked shifts in the comparative
advantage and competitive position of firms and nations; fierce competition
among firms and nations for markets, technology, capital and skilled employees;
the revolution in information technologies; dramatic reduction in transpor-
tation and communication costs; and the emergence of the Asia-Pacific
Region as a major player in the world economy.
These global developments have made it necessary for all countries to
become more flexible and to accelerate the pace of structural adjustment. A
nation's economic performance - absolute as well as relative — therefore,
depends on the willingness and ability of its firms to adapt to the constantly
changing environment. Slow economic growth, poor productivity performance,
stagnant real incomes and high unemployment in Canada and other indus-
trialized countries are attributed to the inability of firms to make necessary
modifications to their strategies and activities, and governments to adjust quickly
and decisively to these global changes.
The poor economic performance of the global economy, especially the
OECD countries, and the serious economic difficulties of many large and well-
known global companies such as GM, IBM and Eastman Kodak during the 1980s
and the early 1990s, strongly suggest that corporate internal control (corporate
governance) systems have failed to deal effectively with the challenges of
structural changes and adjustment. Since all the above-mentioned structural
43
RAO & LEE-SING
trends are expected to continue, if not intensify, in the future. The challenge
of adaptation and structural adjustment for Western firms and political systems,
therefore, is likely to continue for several decades.
Reactions and feedback from capital, product and factor markets, and legal,
political and regulatory systems, either individually or collectively, could effectively
address the problems of inadequate, ineffective and inefficient management.
However, as Jensen (1993) has eloquently argued, these control mechanisms are
either too blunt as instruments or too slow to act, resulting in very slow structural
adjustment, waste of productive resources, serious economic difficulties for firms
and employees, and poor overall economic performance. On the other hand, a
system of effective corporate governance - including the active participation of
shareholders in the direct and indirect management of the corporation through
the board of directors, and an arrangement of productive checks and balances
between the shareholders, board of directors and management of the corporation —
should increase corporate dynamism and flexibility, minimize the overall costs of
economic adjustment, and change and improve the global economic performance.
The recent upsurge of interest and research activity related to corporate
governance in Canada and other industrial countries is a reflection of the
growing recognition of the importance of corporate governance to the strong
economic performance of firms and nations. For example, in Canada, the TSE
report "'Where Were the Directors?' Guidelines For Improving Corporate
Governance in Canada"
1
 examines the role of the board of directors in corporate
governance and decision-making in Canada and recommends 18 measures to
improve current governance structures and practices.
To date, the corporate governance debate in Canada and elsewhere has
concentrated mainly on the role of the board of directors in ensuring that share-
holders' interests are met and agency costs are minimized. As a result, both the
research and policy debates have been too narrowly focused. In addition, much
of the past research on corporate governance in Canada is primarily qualitative
and is not based on rigorous empirical analysis. Current policy and research
efforts that focus simply on inter-country comparisons of corporate governance
environments cannot resolve these problems. For example, most Canadian firms
are closely held while the majority of American firms are widely held. Although
the problem of managers ignoring shareholders is prominent in widely held U.S.
firms, it is less of an issue in closely held Canadian firms. Therefore, research
and, ultimately, public policy should focus on problems that are specific to the
Canadian corporate governance environment.
The primary objective of this study is to conduct an in-depth, firm-specific
empirical analysis of the interaction between corporate governance, corporate
decision-making, and corporate performance in Canada. In particular, by using
extensive company data on Canadian and American firms, we will examine
empirically the role of corporate governance structure in corporate decision-
making and performance in Canada and compare the Canadian results with our
findings for the American companies. Our secondary objective is to provide a
44
general empirical background to the other studies in this volume, especially
the qualitative studies dealing with specific issues of corporate governance.
The governance variables include, among others, concentration of owner-
ship (e.g., widely held versus closely held); size and composition (e.g., inside versus
outside directors) of the board of directors; institutional ownership; inside
ownership; and the role of the CEO on the board of directors. Decision variables
include debt-to-asset ratio; capital-to-labour ratio; R&D intensity; and degree of
outward orientation (measured by the importance of foreign assets and foreign
sales in total assets and sales). Firm performance variables include accounting
measures (such as capital and labour productivity); sales and asset growth;
growth in earnings-per-share; and rates of return on assets and equity.
2
Following this Introduction, we discuss corporate governance structure,
disaggregated by firm size class (measured by sales), and by major industry
groupings in the section on the Governance System in Canada and the
United States. The section on the Analytical Framework describes in some
detail the framework we use to examine empirically the role of corporate
governance structure, corporate decision-making and corporate performance
in Canada and the United States. The Empirical Results are analyzed in the
next section, which discusses the importance of good corporate governance
and decision-making for healthy and robust corporate economic performance
in the two countries. Finally, we summarize the main findings of the study and
discuss their implications for possible action by corporations, institutions and
governments in the Conclusions.
Our findings suggest that corporate governance structures in Canada differ
significantly from those in the United States, especially with respect to the
nature and concentration of corporate ownership, institutional ownership,
inside ownership, and the composition of the board of directors. For example,
the concentration of corporate ownership is substantially higher in Canadian
than in American companies, but the concentration of institutional owner-
ship is considerably higher in American than in Canadian companies. More
important, however, is that corporate governance variables, especially in the
United States, appear to have a significant influence on the corporate perfor-
mance variables - directly as well as indirectly - through their influence on
corporate decision variables.
THE GOVERNANCE SYSTEMS IN CANADA
AND THE UNITED STATES
T
HIS SECTION PROVIDES definitions of corporate governance variables and
an examination of the governance structure in Canada, disaggregated by
six major size classes and by 11 major industry groups. The Canadian results
are then compared with the findings for the United States. The description of
both the database and the characteristics of Canadian and American samples,
especially industry and size distributions, appear in Appendix 1.
45
GOVERNANCE STRUCTURE, CORPORATE DECISION-MAKING & FIRM PERFORMANCE
RAO & LEE-SING
The corporate governance literature focuses on two major groups of
corporate governance variables: characteristics and interactions between a
firm's board of directors and management; and the composition of a firm's
ownership. These two groups of variables are further characterized as
either structures or practices.
Appendix 2 provides comparative summaries of the Canadian and
American governance environment.
3
 These summaries describe the role of
both structure and practice variables in shaping the governance environment.
However, the objective of this study is to set out an empirical analysis of the
governance environments in Canada and the United States. Therefore, the
summaries in Appendix 2 may best serve as a foundation for comparing the
Canadian and American environments.
The structural characteristics of the board and the management of a firm
include board size (Jensen, 1993 and Friedlaender, 1992), senior officer size
(Friedlaender, 1992), the inside director ratio (Jensen, 1993 and Friedlaender,
1992), the foreign director ratio, whether or not the CEO is the Chairperson
of the board of directors (Jensen, 1993), whether the CEO is on the board of
directors (Jensen, 1993), and the level of inside ownership (Jensen, 1993).
Firm-specific practices related to the board and management include
proxies of board culture (Jensen, 1993), financial expertise of the board
(Jensen, 1993), the level of legal liability taken by board members (Jensen,
1993), personnel characteristics (such as age, education and experience) of
the senior management (Friedlaender, 1992), corporate life cycle and age
(Morck &. Stangeland, 1994), and CEO interaction — such as information flow
and decision-making - with inside directors (Baysinger, 1990).
The structure of the ownership composition includes variables such as
institutional ownership/activism (Jensen, 1993), percentage held by the
largest shareholders (Morck & Stangeland, 1994), and concentration of
corporate ownership (Morck & Stangeland, 1994).
Firm-specific practices related to the composition of ownership include
the effects of differential voting rights (Morck & Stangeland, 1994), and the
founder/heir ownership relation (Morck & Stangeland, 1994).
Since it was very difficult to obtain reliable quantitative data on corporate
governance practices variables, this study focuses exclusively on the following
corporate governance structure variables.
• Concentration of Voting Shares and Ownership Control
• Level of Inside Ownership
• Institutional Ownership
• Number of Directors and Senior Officers
• Inside Director Ratio
• Foreign Director Ratio
• CEO on the Board or Chairperson of the Board of Directors
46
GOVERNANCE STRUCTURE, CORPORATE DECISION-MAKIN G & FIRM PERFORMANCE
CONCENTRATION OF VOTING SHARES AND OWNERSHIP CONTROL
CONCENTRATION OF CORPORATE OWNERSHIP, the percentage of voting shares
held by significant shareholders, and the number of significant shareholders all
measure the concentration of voting shares.
Concentration of Corporate Ownership or Corporate Control
The level of concentration of corporate ownership or corporate control is
measured by the number of voting shares held by one or a small group of
shareholders. This variable focuses on the defacto control of the voting shares.
There are three categories of ownership concentration: widely held control — in
which companies have no shareholder or group of related shareholders that
own, directly or indirectly, more than 20 percent of the voting shares; effective
control - in which companies have one shareholder or a small group of share-
holders owning, directly or indirectly, 20 percent to 49.9 percent of the voting
shares; and legal control — in which one or a small group of shareholders owns,
directly or indirectly, more than 50 percent of the voting shares of a company.
These definitions follow those used by Daniels & Macintosh (1991).
In the Canadian sample, 55.5 percent of the firms are legally controlled
while 21.4 percent and 23.1 percent are effectively controlled and widely
controlled, respectively (Figure 1). In contrast, less than 25 percent of the
U.S. firms are legally controlled, while 35.1 percent are effectively controlled
and 40.2 percent are widely held.
The majority of Canadian firms are legally controlled in all six size classes.
The differences in ownership structure between the two countries is more
pronounced for firms with over US$ 1 billion annual sales (see Appendix 3, Table
A3-5). These results are similar to the findings of Morck &. Stangeland (1994).
Unlike the size classes, the concentration of ownership differs significantly
across the major Canadian industry groups. For instance, the share of legally
controlled firms varies between a low of 32 percent in Mining to a high of 72
percent in Transportation and Public Utilities (Appendix 3, Table A3-6). The
Mining and Technology-Intensive Manufacturing industries have the majority
of firms in the widely held and effectively controlled categories, while the high
level of legally controlled firms in the Transportation and Public Utilities
industry might be simply a reflection of the provincial governments' owner-
ship of public utilities.
The levels of ownership concentration in the U.S. industries also differ
noticeably from those in Canada. The percent of legally controlled firms is
substantially lower in all the major U.S. industries. In the Finance, Insurance
and Real Estate industry, and in the Transportation and Public Utilities industry,
the majority of U.S. firms are widely held. In sharp contrast, about 70 percent
of Canadian firms in these two industries are legally controlled.
47
RAO & LEE-SING
FIGURE 1
CONCENTRATION OF OWNERSHIP IN CANADA AND THE UNITED STATES
Source: Based on Table A3-5.
Percentage of Voting Shares Held by Significant Shareholders
A shareholder that owns at least 10 percent of the voting shares in a company
is considered to be a "significant" shareholder. The percentage of voting shares
held by all significant or "10 percent" shareholders focuses on the defacto
control that may exist among these large block shareholders. This variable is
directly related to the corporate concentration variable, discussed above,
because the categorization of control - widely held, effective or legal - is based
on the percentage of voting shares held by all significant shareholders.
On average, the majority (53.6 percent) of voting stock in the Canadian
sample is held by one or more significant shareholders. The high concentration
of ownership is consistent with the majority of legally controlled firms in
Canada. The percentage of voting shares held by significant shareholders
increases with the size of the firm except for firms with sales over US$ 2 billion
(Appendix 3, Table A3-7). This result is consistent with the positive relation-
ship between the proportion of legally controlled firms and firm size.
Unlike the similarities between the two concentration variables by size
class, the concentration of ownership within industry groups differs from the
percentage held by the significant shareholders. The percentage held by
significant shareholders varies from a high of 80 percent in the Technology-
Intensive Manufacturing industry to a low of 12 percent in the Agriculture,
Forestry and Fishing industry (Appendix 3, Table A3-8). Although the
Technology-Intensive Manufacturing industry has the highest level of signifi-
cant shareholder ownership, it has a low number of legally controlled firms.
48
GOVERNANCE STRUCTURE, CORPORATE DECISION-MAKING & FIRM PERFORMANCE
Number of Significant Shareholders
The number of significant shareholders is another measure of the concentration
of voting shares within a firm. Although the number of 10 percent owners is
not directly related to the total percentage of the voting shares held by the
firms' significant shareholders, discussed above, it also indicates the concen-
tration of ownership within a firm.
In the Canadian sample, unlike the percentage of voting shares held by
significant shareholders, the number of significant shareholders tends to
decline with the size of the firm, except for the smallest size class (Appendix 3,
Table A3-9). This implies that the average dollar value of shares held by
significant shareholders is considerably higher in larger firms. Canadian firms
also have a higher number of significant shareholders across all size classes
than American firms.
The average number of significant shareholders varies between different
major industry groupings. Construction, Wholesale Trade, Services and
Labour-intensive Manufacturing tend to have more significant shareholders
than the norm, while Agriculture, Finance and Mining have fewer significant
shareholders. There is no systematic variation of the concentration of large
shareholders across the industry groupings in the Canadian and the American
samples (Appendix 3, Table A3-10).
LEVEL OF INSIDE OWNERSHIP
THE PERCENTAGE OF VOTING SHARES HELD BY "Insiders" follows the SEC's
definition of inside shareholders (traders). Inside shareholders include directors,
officers and affiliates of the firm.
In the Canadian sample, on average, over 21 percent of company shares
are held by insiders, compared to less than 10 percent in the American sample.
The percentage of voting shares held by insiders in Canadian firms declines
from an average of 35 percent in the smallest size class to 13 percent for firms
that are moderately sized and then increases to over 22 percent for the largest
size class. Insider ownership in the United States, on the other hand, tends to
decline systematically with the size of the firm (Figure 2).
Firms in the U.S. Retail Trade, Agriculture, Forestry and Fishing,
Wholesale Trade, and Services industries tend to have higher than average
levels of inside ownership (Appendix 3, Table A3-12). On the other hand,
Mining, Transportation and Public Utilities, and Technology-Intensive
Manufacturing tend to have lower than average levels of inside ownership.
The lack of inside ownership information for several Canadian industries
prevents Canada-U.S. comparisons by industry.
49
RAO &. LEE-SING
FIGURE 2
PERCENTAGE OF VOTING SHARES HELD BY INSIDERS, BY SIZE CLASS
Source: Based on Table A3-11.
INSTITUTIONAL OWNERSHIP
INSTITUTIONAL OWNERSHIP is THE PERCENTAGE of a firm's voting shares held
by institutional investors. The term "institutional owners", as used in this
study, is the same as that used by Disclosure Inc. Institutional owners
include banks and other financial institutions, pension funds, mutual funds,
and other corporations that own shares.
In the Canadian sample, institutional owners control about 38 percent
of the dollar value of shares, compared to 53 percent in the U.S. sample. The
apparent contradiction between the smaller proportion of legally controlled
firms and the higher levels of institutional ownership in the American sample
implies that there are a larger number of institutional holders in the United
States, each controlling only a small block of corporate shares.
Institutional ownership in the Canadian sample increases systematically
with firm size except for the largest size class. However, this increase is not
nearly as prominent as in the American sample where the range of institutional
ownership varies from a low of 16 percent in the smallest size class to over
55 percent in the largest size class. Canadian firms have higher levels of
50
GOVERNANCE STRUCTURE, CORPORATE DECISION-MAKIN G & FIRM PERFORMANCE
FIGURE 3
PERCENTAGE OF VOTING SHARES HELD BY INSTITUTIONS, BY SIZE CLASS
Source: Based on Table A3-13.
institutional ownership than their American counterparts in the two smallest
size classes. On the other hand, in other size classes, Canadian firms tend to
have significantly lower institutional ownership than similar size American
firms (Figure 3).
The Agriculture, Forestry and Fishing industry has the lowest level of
institutional ownership compared to the other Canadian industries (9.57 per-
cent versus the Canadian weighted average of 38.24 percent). Finance,
Resource-Intensive Manufacturing, Retail Trade and Wholesale Trade have
lower than average levels of institutional ownership. On the other hand,
Construction, Technology-Intensive Manufacturing, and Transportation and
Public Utilities have high levels of institutional ownership.
Unlike Canada, the level of institutional ownership does not vary substan-
tially across the major industry groups in the American sample. In the United
States the level varies from a low of 49.4 percent in the Retail Trade industry to a
high of 59.3 percent in the Finance, Insurance and Real Estate industry
(Appendix 3, Table A3-14).
51
RAO & LEE-SING
NUMBER OF DIRECTORS AND SENIOR OFFICERS
THE NUMBER OF DIRECTORS WHO SIT on a firm's board of directors is specified
and available from the company's annual reports. Similarly, data on the number
of senior officers (president, executive vice-presidents, senior vice-presidents,
vice-presidents, chiefs, treasurer, secretary, controller, comptroller, and other
comparable positions) is also available from the company's annual reports.
The average number of directors in Canadian firms is 9.25 while the
average number for American firms is 9.87. The number of directors increases
with firm size in both the Canadian and American samples. The number of
directors in the four smallest size classes are similar in Canada and the United
States. In the two largest size classes, however, the number of directors tends
to be significantly higher in the American sample than in the Canadian
sample. This discrepancy could be mainly a reflection of the larger size
American firms in the largest two size classes (Figure 4).
The number of directors across the major industry groups in the Canadian
sample tends to mirror that of the American sample. Construction and Services,
on average, has the fewest number of directors, while Transportation and Public
Utilities, and Resource-Intensive Manufacturing tend to have the largest number
of directors (Appendix 3, Table A3-16). Like the number of directors, the num-
FIGURE 4
NUMBER OF DIRECTORS BY SIZE CLASS
Source: TableA3-15.
52
GOVERNANCE STRUCTURE, CORPORATE DECISION-MAKING & FIRM PERFORMANCE
ber of senior officers increases with the size of the firm on the two country
samples (Appendix 3, Tables A3-17 and A3-18).
INSIDE DIRECTOR RATIO
THE INSIDE DIRECTOR RATIO REPRESENTS THE RATIO of inside directors (officers
of the firm who also sit on the Board) to the total number of directors.
4
The inside director ratio averages 20 percent in the Canadian sample
compared to 22 percent in the American sample. In addition, the American
insider ratio is higher than the Canadian ratio in all six size classes. The insider
ratio declines with firm size in the two samples - i.e., the larger firms tend to
have the smaller insider director ratio (Figure 5).
In Canada, the inside director ratio tends to be higher in the
Technology-Intensive Manufacturing, Wholesale Trade, Services, Mining, and
Construction industries. On the other hand, the ratio is lower in the Finance,
Insurance and Real Estate, and Transportation and Public Utilities industries.
With few exceptions, the industrial structure of the Canadian insider ratio
tends to be similar to the American (Appendix 3, Table A3-20).
FIGURE 5
INSIDE DIRECTOR RATIO BY SIZE CLASS
Source: Table A3-19.
53
RAO & LEE-SING
FOREIGN DIRECTOR RATIO
THE FOREIGN DIRECTOR RATIO REPRESENTS the ratio of the number of directors
that reside outside the nation where the firm is incorporated to the total number
of directors.
5
 This information is only available for the Canadian sample.
On average, only 15 percent of the directors of Canadian firms are residents
of foreign countries. The foreign director ratio tends to remain constant for
the first four size classes, between 18 percent and 19 percent. However, it
increases to 24 percent for firms with sales between US$ 1 and US$ 2 billion.
But, for the largest size class (sales over US$ 2 billion), the ratio averages only
11 percent (Figure 6).
Mining, Resource-Intensive Manufacturing, Services, and Wholesale
Trade industries have an above-average foreign director ratio. On the other
hand, the Labour-intensive Manufacturing, Mining and Construction industries
have a below-average ratio (Appendix 3, Table A3-22).
FIGURE 6
FOREIGN DIRECTOR RATIO IN CANADIAN FIRMS, BY SIZE CLASS
Source: Table A3-21.
54
GOVERNANCE STRUCTURE, CORPORATE DECISION-MAKING & FIRM PERFORMANCE
CEOvON THE BOARD OR CHAIRPERSON OF BOARD OF DIRECTORS
THIS VARIABLE INDICATES WHETHER the chief executive officer of the firm is
also the chairman of the board of directors, as specified and available from the
company's annual report.
On average, only 34-5 percent of Canadian firms have the CEO as the
chairperson, compared to almost 60 percent in the American sample (Figure 7).
In over 83 percent of Canadian firms, however, the CEO is also a member of
the firm's board of directors (Appendix 3, Table A3-23). There appears to be
no systematic relationship between whether or not the CEO is the chairperson
and firm size in the Canadian sample. But there is a strong positive relation-
ship between the CEO as chairperson and the size class. In contrast, the
proportion of firms with the CEO as chairperson does not vary systematically
across major industry groups (Appendix 3, Tables A3-24 and A3-26).
SUMMARY
THE CONCENTRATION OF CORPORATE OWNERSHIP is substantially higher in
Canada than in the United States. For instance, more than 55 percent of firms
are legally controlled, compared to less than 25 percent in the United States.
However, the institutional ownership is considerably higher in the United
States. These two results imply that there are a large number of institutional
investors in America each controlling only a small block of corporate shares.
FIGURE 7
CEO AS CHAIRPERSON
Source: Table A3-25.
55
RAO & LEE-SING
The number of directors and officers increases with firm size in the two
countries. The ratio of inside directors to total directors is higher in the
United States in all six size classes. Similarly, the proportion of firms with the
CEO as Chairperson of the Board of Directors is substantially higher in the
American corporations.
There appears to be, on average, a significant systematic relationship
between the corporate governance variables, firm size classes and major industry
groups. Therefore, in empirically examining the effect of governance variables
on corporate decision-making and corporate performance, the influence of size
and industry characteristics of firms must be taken into account.
ANALYTICAL FRAMEWORK
I
N THE PREVIOUS SECTION we examined seven main characteristics of the
corporate governance structure in Canada and the United States. We now
provide an empirical analysis of the influence of corporate governance on
corporate decision-making and corporate performance in the two countries.
This section will outline the analytical framework of the empirical (regression)
analysis reported in the next two sections.
The corporate governance structure is expected to affect directly the
corporate performance by having an effect on the managerial, technical and
adjustment efficiencies of the firm. In addition, the corporate governance
variables can indirectly influence the corporate performance through their
influence on the firm's strategies and decisions with regard to inputs, outputs,
innovations, markets, etc.
Therefore, the ability of corporate internal controls to deal effectively
with the challenges of structural adjustment can be analyzed by studying the
linkages between the three groups of variables: corporate governance structure
variables, corporate decision-making variables, and corporate economic
performance variables. The corporate governance structure variables used in
the regression analysis are those defined in the previous section.
Corporate decision-making variables represent activities of firms resulting
from both day-to-day and longer-term corporate strategies and decisions. They
include leverage in the firm — measured by the debt-to-assets ratio; the capital-
to-labour ratio - the ratio of assets to employees; the R&.D intensity — measured by
the ratio of R&.D to total sales; and the firm's degree of outward orientation -
measured in three ways: the ratio of foreign sales to total sales, the ratio of
foreign assets to total assets, and a dummy variable based on the presence of
foreign sales or assets.
Corporate economic performance variables measure the corporate
performance in terms of productivity, profitability, and growth. These include
the capital productivity ratio (sales to assets); labour productivity ratio (sales to
employees); sales growth; asset growth; the growth in capital and labour produc-
tivity; return on equity; return on assets; and the growth in earnings-per-share.
Table 1 includes and categorizes the variables used in the regression analysis.
56
GOVERNANCE STRUCTURE, CORPORATE DECISION-MAKIN G & FIRM PERFORMANCE
TABLE 1
CORPORATE GOVERNANCE, DECISION MAKING AND PERFORMANCE VARIABLES
VARIABLE
GROUP MEASURE VARIABLE
Governance Concentration of Ownership
Concentration of Ownership
Concentration of Ownership
Concentration of Ownership
Ownership Composition
Ownership Composition
Composition of Board
Composition of Board
Composition of Board
Composition of Board
Composition of Board
Composition of Board
Composition of Board
Decision Leverage
R&.D Intensity
Capital Labour Ratio
Outward Orientation
Outward Orientation
Outward Orientation
Performance Return
Return
Growth
Growth
Growth
Growth
Growth
Productivity
Productivity
Widely Held
Effectively Controlled
Percentage of Voting Shares Held by
all of the Significant Shareholders
1
Number of Significant Shareholders
1
Institutional Ownership
Insider Ownership
Inside Director Ratio
Foreign Director Ratio
Board Size
Officer Size
1
CEO is on the Board
CEO is the Chairperson
CEO is the Chairperson data not available
2
Debt-to-Asset Ratio
R&D-to-Sales Ratio
Assets-to-Employee Ratio
Foreign Sales to Total Sales
Foreign Assets to Total Assets
1
Presence of either Foreign Assets or Sales
1
Return on Equity
Return on Assets
Asset Growth
Sales Growth
Capital Productivity Growth
Labour Productivity Growth
Growth in Earnings-per-Share
Capital Productivity Ratio (Sales over Assets)
Labour Productivity Ratio (Sales over Employees)
Notes: This table contains a list of the variables described in the analytical framework section of the study. The
variables, shown on the far right are categorized into three main groups: corporate governance variables,
decision-making variables, and performance variables. These variables are further categorized into the
various measures shown in the second column.
1
 Variables that were included in the empirical framework of our study but were not included in the
regression outputs shown in Tables 5, 6, 7, and 8. These variables are highly correlated with other
variables that represent similar measures and their inclusion would lead to problems of multicollinearity.
2
 This variable is used in the Canadian regressions because several firms did not report this information.
Two sets of dummy variables are included in the regression equations to
control for the effects of firm size (sales) and type of industry on the corporate
decision-making and corporate performance variables. Six size classes are
represented by five dummy variables. Eleven major industry groupings are
represented by ten dummy variables. Table 2 gives the neumonics of the
dummy variables used in the regression analysis.
57
RAO & LEE-SING
TABLE 2
REGRESSION DUMMY VARIABLES
VARIABLE GROUP CLASS OR GROUP SIZE
SIZE (US$) 10 million to 50 million
50 million to 100 million
100 million to 500 million
500 million to 1,000 million
1,000 million to 2,000 million
2,000 million and greater
MAJOR INDUSTRY Agriculture, Forestry & Fishing
Construction
Finance, Insurance & Real Estate
Labour-intensive Manufacturing
Mining
Resource-Intensive Manufacturing
Retail Trade
Services
Technology-Intensive Manufacturing
Transportation & Public Utilities
Wholesale Trade
Note: This table contains a list of the size and industry dummy variables used in the regression analysis. Five sizes
of dummy variables and ten industry dummy variables were created using the indicator coding method.
The two control groups were "US$ 2,000 million and greater" and "Wholesale Trade".
The linkages between the three sets of variables can be analyzed with
the two models. The first formulation (Figure 8) analyzes the relationship
between the corporate governance structure and corporate decision-making
sets of variables, controlling for both the size and industry effects.
The second model (Figure 9) depicts the influences of the corporate
governance and decision-making variables on the performance variables,
controlling for industry and size effects.
Two sets of regression equations are used to generate the above models.
The first set of equations (Regression Equations, Set 1) are used to test for the
significance of the influence of corporate governance variables on decision-
making variables. The two sets of regressions are estimated for both the
Canadian and American samples.
58
GOVERNANCE STRUCTURE, CORPORATE DECISION-MAKING & FIRM PERFORMANCE
FIGURE 8
LINKAGES BETWEEN CORPORATE GOVERNANCE STRUCTURE
AND DECISION-MAKING
Corporate
Governance
Structure (X^)
=>
Size Industry
i 1
Corporate
Decision-
Making (Y
D
)
FIGURE 9
LINKAGES BETWEEN CORPORATE GOVERNANCE STRUCTURE,
DECISION-MAKING AND PERFORMANCE VARIABLES
Size Industry
Corporate Governance Structure (X^)
I
Corporate Decision-Making (X^)
—>
I A
Corporate
Economic
Performance (Yp)
59
RAO & LEE-SING
REGRESSION EQUATIONS, SET 1
Y
D1
 = f(X
G1
,..., X
GN
, SIZ, MI)
Y
DN
 = f(X
G1
,...,X
GN
,SIZ,MI)
where
XQJ, ... , XGN are the corporate governance structure variables,
Y
D]
, ... , Y
DN
 are the corporate decision variables,
SIZ are the size class dummy variables, and
MI are the major industry group dummy variables.
The second set of equations (Regression Equations, Set 2) is used to
analyze the relationship between the corporate governance structure variables,
corporate decision-making and corporate performance variables, again, con-
trolling for size and industry effects.
In this model, the total effect of a corporate governance variable on a
corporate performance variable is the sum of the two effects: the direct effect
and the indirect effect. The indirect effect measures the influence of the
governance variable on the performance variable operating through its impact
on the decision-making variable.
REGRESSION EQUATIONS, SET 2
Y
P1
 = fvXd,... , XQN, X
D1
,..., X
DN
, SIZ, MI)
YPN = I(XQI, ..., XG
N
, X
D1
,..., X
DN
, SIZ, MI)
where
XQJ, ... , XG
N
 are the corporate governance structure variables,
X
D1
, ... , X
DN
 are the corporate decision variables,
Y
P1
, ... , Y
PN
 are the corporate performance variables,
SIZ are the size class dummy variables, and
MI are the major industry group dummy variables.
60
GOVERNANCE STRUCTURE, CORPORATE DECISION-MAKING & FIRM PERFORMANCE
POSSIBLE EFFECTS OF CORPORATE GOVERNANCE VARIABLE S
THE REGRESSION RESULTS ARE DISCUSSED IN THE NEXT SECTION. First, however,
we will review from the corporate governance literature a priori relationships
among the governance structure, decision-making and performance variables.
However, as described below, these relationships are confounded by many
interactions.
Concentration of Ownership
Morck & Stangeland (1994) examined the relationship between concentration
of ownership and firm performance in Canada and the United States. The
managers of widely held firms can be neither effectively monitored nor
controlled by the widely dispersed and often unsophisticated shareholders who
hold only a small number of shares ("small" shareholders). Therefore, the
decisions and performance of widely held firms could be adversely influenced by
the divergence of interests of the managers and shareholders. On the other hand,
sophisticated shareholders who own large numbers of shares ("large" shareholders)
will have the ability and incentive to monitor effectively the decisions and per-
formance of managers. But the concentration of corporate ownership could result
in an unhealthy and undemocratic concentration of economic power.
6
In addition, the effect of corporate ownership on firm performance could
be significantly influenced by the nature or composition of the ownership
concentration - the proportion of voting shares held by institutional investors
and insiders, the number of significant shareholders, etc. Therefore, it is difficult
to predict a priori the relationship between the level of concentration of
ownership and its effect on decision-making and performance.
In an attempt to reach a better understanding of the relationship between
the level of concentration of ownership and firm decision-making and perfor-
mance, we disaggregated the two country samples into the three groups: widely
held, effectively controlled and legally controlled. The averages of governance,
decision and performance variables for the three groups are displayed in
Appendix 4, Table A4-1.
There appears to be a systematic negative relationship between the rate
of return on assets and equity and the level of ownership concentration in the
United States. The relationship between rate of return and concentration of
ownership is not as clear for the Canadian firms because the average rates of
return on equity and the average return on assets are the same for widely held
firms as for legally controlled firms.
On the other hand, growth of sales and assets in American firms appears
to be positively related to ownership concentration. In the Canadian sample,
legally controlled firms, on average, have much lower sales and asset growth
than the other two groups (Appendix 4, Figure A4-2).
61
RAO &. LEE-SING
It is important to note that these are partial results because they do not
take into account the size and industry effects or the interactions between the
governance and decision-making variables. Therefore, these provide only a cursory
glimpse of the relationship between performance and ownership concentration.
7
Institutional Investors
According to several commentators, American institutional investor activism
(for example, the California Public Employees Retirement System [CalPERS],
Corporate Partners, Allied Investment Partners, etc) has a significant positive
effect on firm performance because these well-informed investors can assess
firm management strategies and activities in an unbiased way and can there-
fore exert pressure on the board for change and dynamism through either voting
or selling mechanisms. In addition, institutional investors could take a more
active role in the management of the firm by obtaining a seat on the board.
However, the effectiveness of institutional investors could be severely
limited by the legal, tax and regulatory constraints (Jensen, 1993). Furthermore,
institutional investors, such as public pension funds, may be risk-averse and shy
away from activism for enhancing the firm performance. Nevertheless, the cost
of a proxy fight has fallen from US$ 1 million to less than US$ 5,000
8
 (The
Economist, 1994) thereby minimizing the constraints on institutional activism.
Insider Ownership
A high degree of ownership by managers and directors could well align their
interests in the firm with those of the shareholders. Managers and directors,
whose remuneration and personal wealth are closely tied to the firm's
performance, would prefer to emphasize firm performance over other objectives.
But, as the level of inside ownership increases, management could become
more entrenched and fail to act in the interests of other shareholders. Thus,
the relationship between inside ownership and firm performance is confounded
by the interaction of these two effects. The interaction between the entrench-
ment and the incentive alignment effects has been analyzed by Morck (1994).
Inside Directors
Baysinger & Hoskisson (1990) reviewed the body of literature that focuses on
inside directors. They argue that insiders have access to information that is
relevant to assessing both the managerial competence and the strategic
desirability of initiatives. They also state that outside directors, although they
are more open and objective, lack the amount and quality information needed
to perform their roles. Hence, the inside directors ratio can improve the effec-
tiveness of a firm's decision-making. Moreover, agency theory states that
inside directors should perform just as well or better than outside directors
62
GOVERNANCE STRUCTURE, CORPORATE DECISION-MAKING & FIRM PERFORMANCE
because their reputations and economic well-being are tied directly to the
performance of the firm.
On the other hand, inside directors may not be as objective as outside
directors if the CEO is a member and/or the chairperson of the board. Also, out-
side directors may bring insight and objectivity to the decision-making process
from their involvement and association with other organizations and outside
sources. It is uncertain whether the positive effects of inside directors outweighs
the negative effects; thus, the relationship between inside directors and decision-
making and firm performance cannot be predicted unambiguously a priori.
CEO as Chairperson
Proponents of separating the offices of CEO and chairperson argue that the
ability of the board to function independently and effectively will be compro-
mised when the CEO is also the chairperson of the board. Separation of CEO
and chairperson positions could improve the decision-making process and
performance of the firm in three ways (Bacon, 1993). First, the relationship
between the board (the overseer of management) and management will
become clear. Second, with an independent leader the board will become more
effective and better organized. Third, the board's responsibility to look after
shareholders' interests will come to the forefront.
Conversely, a combined CEO/chairperson role would enhance the
information flow between the Board and management and improve the co-
operation and co-ordination between the two bodies. It is not clear whether
the advantages of separation of the two positions dominate the disadvantages.
Board Size
The TSE report on Corporate Governance (1994) and Jensen (1993) looked
into the debate of board size and its influence on the effectiveness of the
board. Large boards bring a diversity of views and experience, increase the
opportunity for a broad geographic representation, and provide extensive
director resources for constituting board committees to deal effectively with
complex issues. However, beyond a certain threshold of board size, the
information flow and decision-making could become more difficult and cumber-
some, and the directors might lose their sense of responsibility and accountability.
Again, it is unclear whether board size has a positive or negative effect on the
decision and performance variables.
SUMMARY
IN SHORT, IT IS DIFFICULT TO PREDICT, a priori, the relationship among the gover-
nance, decision and performance variables. Against this background, the
regression results are discussed in the next section.
63
RAO & LEE-SING
EMPIRICAL RESULTS
T
HIS SECTION BEGINS WITH a discussion of the degree of empirical association
between the corporate governance structure and decision-making variables,
after controlling for the effects of size and industry characteristics. We then
examine the degree of association between the governance structure and firm
performance variables, discussing the roles of governance structure and decision-
making on performance, along with the size and industry effects. Finally, we
discuss the direct and indirect effects of the governance structure on the firm
performance through its impact on the decision-making variables.
9
CORPORATE GOVERNANCE AND DECISION-MAKING
Tables 3 and 4 display the Canadian and American regression results of Model 1
as shown in Figure 8, described in the previous section. The size of the
F-statistics indicates significant associations among the decision, governance,
size and industry variables in Canada and the United States. However, in general,
the American results are much stronger than the Canadian regressions.
TABLE 3
CANADIAN CORPORATE GOVERNANCE AND DECISION-MAKING REGRESSION RESULTS
DEBT
ASSETS
ASSETS
EMPLOYEE
R&D
SALES
FOREIGN SALES
TOTAL SALES'
(Constant)
Board Size
CEO is Chair
CEO is Chair n/a
Foreign Director Ratio
Inside Director Ratio
Inside Ownership
Institutional Ownership
Widely Held
Effectively Controlled
F
r?
n
0.3293
3
-0.0100
3
0.0494
-0.0052
-0.0488
-0.0685
0.0000
-0.0002
-0.0272
0.0298
4.2963
0.1603
566
-3360297
3
101307
3
-75628
-149821
404900
-1428028°
15161
C
32952
a
967610°
405216
10.1254
0.2862
631
0.3156
C
-0.0195
b
-0.0349
-0.0456
-0.1678
-0.0451
0.0008
-0.0002
-0.0026
-0.0188
1.0894
0.2664
89
0.0843
-0.0021
0.2095
3
0.1568
3
0.4290
3
0.0082
-0.0007
-0.0009
C
-0.0419
-0.0234
3.6924
0.1485
533
a=l % b = 5% c=10% t- statistic significance
Notes: This table summarizes the Canadian regression results illustrated in Figure 8, Model 1. Each column
represents one of the equations depicted in Regression Equations Set 1. The contents of each cell indicate
the coefficient value and it's significance in the regression equation. Industry and Size dummy variables,
described in Table 2, have been included in the regressions as control variables but are omitted from this
summary table.
1
 The ratio of Foreign Sales to Total Sales, a measure of the firm's outward orientation, was chosen for
this Table. The other two measures of outward orientation, the ratio of Foreign Assets to Total Assets
and the Presence of Foreign Sales or Assets, have very similar regression results.
64
GOVERNANCE STRUCTURE, CORPORATE DECISION-MAKING & FIRM PERFORMANCE
TABLE 4
AMERICAN CORPORATE GOVERNANCE AND DECISION MAKING REGRESSION RESULTS
DEBT
ASSETS
ASSETS
EMPLOYEE
R&D
SALES
FOREIGN SALES
TOTAL SALES'
(Constant)
Board Size
CEO is Chair
Inside Director Ratio
Inside Ownership
Institutional Ownership
Widely Held
Effectively Controlled
F
r2
0.2985
3
-0.0003
0.0147
C
-0.0129
-0.0005
b
-0.0006
3
-0.0507
3
-0.0279
b
17.5080
0.2221
1372
a = l% b=5% c
427934
a
-4628
-8903
-80081
-1402
1475
-57818
-35763
72.6868
0.3541
2940
0.4813
3
-0.0011
0.0085
-0.0703
-0.0001
-0.0006
0.0056
0.0162
3.3642
0.0921
753
0.1418
3
0.0004
-0.0152
C
-0.0168
-0.0002
0.0003
0.0019
0.0063
23.3684
0.2544
1530
= 10% t - statistic significance
Notes: This table summarizes the American regression results illustrated in Figure 8, Model 1. Each column
represents one of the equations depicted in Regression Equations Set 1. The contents of each cell indicate
the coefficient value and its significance in the regression equation. Industry and Size dummy variables,
described in Table 2, have been included in the regressions as control variables but are omitted from this
summary table.
1
 The ratio of Foreign Sales over Total Sales, a measure of the firm's outward orientation, was chosen
for this Table. The other two measures of outward orientation, the ratio of Foreign Assets to Total Assets
and the presence of Foreign Sales or Assets, have very similar regression results.
The debt-toasset ratio, a measure of leverage or riskiness of the firm, is
significantly negatively related to institutional ownership and insider owner-
ship, and positively related with the concentration of ownership. In addition,
firms in the smallest size class have lower debt-to-asset ratios than firms in the
largest size class.
The Canadian debt-to-asset ratio regression results are considerably
weaker than the U.S. findings. The leverage of firms in Canada is significantly
related to only one governance variable. The debt-to-asset ratio and board size
are negatively related. Firms in the three smallest size classes have significantl y
lower leverage than firms in the largest size class.
Unlike the results on the debt-to-asset ratio, the governance variables
have a stronger influence on capital intensity in Canada than in the United
States. In Canada, board size, institutional ownership and insider ownership
are positively related to capital intensity. In contrast, corporate ownership
concentration and the inside director ratio are weakly negatively related to
the capital-labour ratio. Firms in the three smallest size classes have higher
capital intensity than firms in the largest size class.
65
RAO & LEE-SING
None of the governance variables is significantly related to capital intensity
in the American sample. In addition, unlike Canada, firms in the two smallest
size classes have slightly lower capital intensity than firms in the largest size
class. On the other hand, the capital-to-labour ratio and the industry character-
istics are strongly related in the American sample.
The research and development intensity, measured by the ratio of R&D
to sales, is not significantly related to any of the governance structure variables
in the American sample. The R&D intensity of firms with sales between
US$ 1 billion and US$ 2 billion is significantly higher than firms in the
largest size class with sales over US$ 2 billion.
In the Canadian sample, only board size is significantly negatively related
to the R&D-to-sales ratio. There is a weak positive relationship between the
R&D intensity and firm size.
In the United States, firms that have a combined CEO-chairperson have
significantly lower outward orientation, measured by the ratio of foreign sales
to total sales, than firms that do not combine the two positions. Not surpris-
ingly, firms in the smallest five size classes have lower outward orientation than
firms in the largest size class.
In Canada, the outward orientation is strongly positively related to firms
with the CEO as the chairperson of the board. Also, as expected, there is a
strong positive relationship between the foreign director ratio and the outward
orientation of the firm. The outward orientation of medium-size Canadian
firms is not significantly different from that of the largest firms. However, the
outward orientation of firms in the smallest size class (with sales less than
US$ 50 million) have significantly lower outward orientation than firms in the
largest size class.
Institutional ownership seems to have a much more positive influence
on corporate decision-making in the United States than in Canada. On the
other hand, the relationship between the inside director ratio and the decision
variables is similar in the two countries. Firms with a higher inside director
ratio, other things remaining constant, seem to have lower leverage, lower
capital intensity, lower R&D intensity and lower outward orientation.
American firms with high levels of inside ownership tend to have low
leverage. Inside ownership in Canadian firms is significantly positively related
to capital intensity. In American firms, leverage increases with the concentra-
tion of ownership. That is, legally controlled firms tend to have higher debt-
to-asset ratios compared to the effectively controlled and widely held firms. In
Canada, the capital intensity of the widely held firms is somewhat higher than
in the legally controlled group firms.
In both countries, a CEO-chairperson is associated with high levels of
leverage. In Canada, outward orientation of firms with CEO as the chairperson
tends to be higher than in the firms with the two positions separated.
The relationships between the size of the board of directors and the
decision variables are stronger in Canada than they are in the United States.
66
GOVERNANCE STRUCTURE, CORPORATE DECISION-MAKING & FIRM PERFORMANCE
In Canada, the size of the board is negatively related to both the degree of
leverage and the R&D intensity. Board size in Canada is positively related to
the capital intensity of the firm.
In the two countries, firms in the two smallest size classes have lower leverage
and lower outward orientation than firms in the largest size class. In Canada, how-
ever, smaller firms tend to have higher capital intensity than the larger firms.
CORPORATE PERFORMANCE AND GOVERNANCE STRUCTURE
TABLES 5 AND 6 DISPLAY THE CANADIA N AND AMERICAN regression results for the
corporate performance variables depicted in Figure 9 (Model 2). The size of the
F-statistics suggests that the equations explain the inter-firm variation in the
corporate performance variables fairly clearly. But, like the equations for corporate
decision variables, the Canadian results are not as robust as the American results.
Profitability
For the U.S. sample, both the ROE and the ROA are significantly positively
related to institutional ownership, inside ownership, and inside director ratio.
Widely held and effectively controlled firms exhibit significantly higher
profitability than the legally controlled firms. Similarly, the profitability of
firms in the largest size class (sales over US$ 2 billion) is significantly better than
firms in the two smallest size classes (firms with less than US$ 100 million).
In the case of Canadian firms, only one of the governance variables
(inside director ratio) is significantly positively related to profitability. Like
the U.S. results, the profitability of firms in the two smallest size classes is
significantly inferior to that of firms with sales over US$ 2 billion.
Growth
The relationship between the two growth measures and the governance vari-
ables for American firms is very similar to the findings for the two profitability
measures described above. The growth of sales and assets are significantly posi-
tively related to institutional ownership, inside ownership, and the inside direc-
tor ratio. Similarly, the growth performance of firms in the two smallest size
classes is significantly weaker than firms with sales over US$ 2 billion. However,
the growth performance of firms with sales between US$'500 million and US$ 1
billion is significantly better than firms in the largest size class.
In the Canadian sample, the asset growth is significantly positively
correlated with the CEO as the Chairperson of the Board, the foreign director
ratio, the inside director ratio, and size of the Board. However, sales growth is
somewhat positively related to only the size of the Board. Unlike the U.S.
results, the growth performance of firms with sales under US$ 100 million is
somewhat better than the performance of firms with sales over US$ 2 billion.
67
RETURN
ON
EQUITY
RETURN
ON
ASSETS
SALES
GROWTH
ASSET
GROWTH
SALES
EMPLOYEE
SALES
ASSETS
LABOUR
PROD.
GROWTH
CAPITAL
PROD.
GROWTH
EPS
GROWTH
TABLE 5
CANADIAN CORPORATE PERFORMANCE AND CORPORATE DECISION-MAKING REGRESSION RESULTS
(Constant) 0.1765
b
 0.0391
Assets/Employee 0.0000 0.0000
Debt/Assets -0.1997
3
 -0.0532
b
Board Size -0.0040 0.0003
CEO is Chair -0.0025 0.0108
CEO is Chair n/a -0.0444 -0.0033
Foreign Director Ratio 0.0558 -0.0044
Inside Director Ratio 0.2112
a
 0.0691
b
Inside Ownership -0.0004 0.0002
Institutional Ownership -0.0001 0.0000
Widely Held -0.0045 -0.0050
Effectively Controlled 0.0095 -0.0139
F 2.3287 1.7335
r? 0.1219 0.0937
n 463 463
0.0419
0.0000
b
-0.0234
0.0005
0.0955
C
-0.0342
0.0733
0.0422
-0.0003
-0.0005
-0.0018
0.0069
2.5521
0.1307
463
-0.0733
0.0000
-0.0485
0.0047
C
0.1203
b
-0.0331
0.2960
3
0.1112
C
-0.0004
-0.0005
0.0203
0.0210
2.8663
0.1460
463
1077695
3
O
a
139295
-20842
3
101535
-60893
-440007
8906
-672
-211
-13106
-22336
9.2013
20.3543
463
2.4883
a
O.OOOO
3
-0.7129
3
-0.0301
3
-0.2206
-0.0092
-0.2477
0.1922
0.0014
0.0003
-0.1585
-0.0453
13.4619
0.4453
463
0.0680
0.0000
b
-0.0349
0.0000
0.0669
-0.0384
0.0876
0.0232
-0.0004
-0.0006
-0.0209
-0.0007
2.3230
0.1217
463
0.1152
C
0.0000
0.0250
-0.0042
C
-0.0247
-0.0011
-0.2227
b
-0.0690
0.0001
-0.0001
-0.0221
-0.0141
1.2780
0.0708
463
0.9702
0.0000
-1.1762
b
-0.0380
0.0819
0.2190
0.9288
0.4137
0.0014
-0.0058
-0.3349
-0.2455
1.2529
0.0695
463
b = 5% c = 10% t - statistic significance
Notes: This table summarizes the Canadian regression results illustrated in Figure 9, Model 2. Each column represents one ot the equations depicted in Regression Equations Set 2.
The contents of each cell indicate the coefficient value and its significance in the regression equation. Industry and Size dummy variables, described in Table 2, have been
included in the regressions as control variables but have been omitted from this summary table.
a+ 1%
TABLE 6
AMERICAN CORPORATE PERFORMANCE AND CORPORATE DECISION-MAKING REGRESSION RESULTS
RETURN
ON
EQUITY
RETURN
ON
ASSETS
SALES
GROWTH
ASSET
GROWTH
SALES
EMPLOYEE
LABOUR CAPITAL
SALES PROD. PROD. EPS
ASSETS GROWTH GROWTH GROWTH
(Constant) 0.0187 0.0296
b
 0.0435 0.0743
b
 547177° 2.1237
3
 0.1054
3
 -0.0307 -0.0657
Assets/Employee 0.0000
b
 0.0000
b
 0.0000
 b
 0.0000
a
 O
a
 O.OOOO
3
 0.0000
b
 0.0000 0.0000
Debt/Assets -0.1904
3
 -0.0500
3
 -0.0521
b
 -0.1541
3
 -69544 -0.4930
3
 -0.0235 0.1020
3
 -0.1348
Board Size -0.0001 -0.0003 0.0001 0.0001 -947 -0.0042 -0.0005 0.0000 0.0068
CEO is Chair -0.0011 -0.0014 -0.0060 0.0023 3613 -0.0058 -0.0067 -0.0083 0.0128
Inside Director Ratio 0.0975
3
 0.0263
b
 0.0500
b
 0.0487
b
 63148 0.0784 -0.0173 0.0012 0.4879
3
Inside Ownership 0.0008
3
 0.0002
b
 0.0004
b
 0.0005
b
 649
C
 0.0021
b
 0.0002 0.0000 0.0004
Institutional Ownership 0.0012
3
 0.0005
a
 0.0007
3
 0.0007
3
 176 -0.0016
C
 0.0001 0.0000 0.0035
b
Widely Held 0.0427
3
 0.01 ll
b
 0.0110 0.0082 -26840 -0.0591 -0.0076 0.0028 -0.0832
Effectively Controlled 0.0201 0.0036 0.0149 0.0051 -22908 0.0740 -0.0072 0.0097 -0.2321
3
F 10.4983 7.0906 4.3160 5.9267 28.4526 46.9150 1.6636 2.8886 2.2912
r
2
 0.1591 0.1133 0.0722 0.0965 0.3389 0.4581 0.0291 0.0495 0.0397
n 1357 1357 1357 1357 1357 1357 1357 1357 1357
a = 1% b = 5% c = 10% t - statistic significance
Notes: This table summarizes the American regression results illustrated in Figure 9, Model 2. Each column represents one of the equations depicted in Regression Equations, Set 2.
The contents of each cell indicate the coefficient's value and it's significance in the regression equation. Industry and Size dummy variables, described in Table 2,
have been included in the regressions as control variables but are omitted from this summary table.
RAO & LEE-SIN G
Productivity
Both labour and capital productivity of American firms are significantly
positively related to the inside ownership ratio. Like profitability and growth,
productivity levels of firms in the two smallest size classes are significantly
lower than the levels of firms in the largest size class. Labour and capital
productivity of Canadian firms are significantly negatively correlated with the
size of the Board. Productivity levels of Canadian firms with less than US$ 100
million are also significantly lower than the firms with sales over US$ 2 billion.
CORPORATE PERFORMANCE AND CORPORATE DECisiON'MAKiNG
PROFITABILITY AND THE GROWTH PERFORMANCE of American firms are
significantly negatively related to the degree of leverage (Tables 5 and 6).
Although the profitability measures are also significantly negatively related to
leverage in Canada, sales and asset growth of Canadian firms are only weakly
related to leverage. In both the Canadian and American samples, leverage has
a strong negative effect on capital productivity. However, leverage does not
have a significant effect on labour productivity in the two countries.
The capital-to-labour ratio has a strong negative relationship with both
profitability and growth in the American equations. Canadian firms' profitability
and growth, however, are weakly (negatively) correlated with capital intensity.
But, as expected, labour productivity of both American and Canadian firms is
strongly positively related to the capital-to-labour ratio.
TOTAL EFFECT OF GOVERNANCE VARIABLES ON CORPORATE PERFORMANCE
HERE, WE WILL SUMMARIZE THE DIRECT AND INDIRECT impact of governance
variables. Recall that the model in Figure 8 investigates the relationship
between corporate decision-making and corporate governance structure variables,
controlling for size and industry effects. Also recall that the model in Figure 9
investigates the direct relationship between corporate performance and
corporate governance structure variables, controlling for size, industry and
corporate decision-making variables. These two models are depicted by the
first two columns that appear under each country in Table 7. The third column,
direct and indirect performance, summarizes the two effects (net impact of
direct and indirect) on performance. These total effects are very similar to the
results obtained by the reduced form equations.
10
In Table 7, the label in each cell corresponds to the signs and levels of
significance found for the corporate governance variable (row). For example,
the association between decision-making and institutional ownership in
Canada is labelled "mixed". "Mixed" means that the number of times the
governance variable is significantly positively related to the decision-making
and/or performance variables is similar to the number of times that it is
70
GOVERNANCE STRUCTURE, CORPORATE DECISION-MAKING & FIRM PERFORMANCE
TABLE 7
SUMMARY OF REGRESSION ANALYSIS
Institutional
Ownership
Inside
Ownership
Concentration
of Ownership
Board Size
Foreign
Director Ratio
Inside
Director Ratio
CEO/Chairman
DECISION-
MAKING
Indirect
Mixed
Weak
Positive
No Impact
Positive
Weak
Positive
Weak
Negative
Weak
Positive
CANADA UNITED STATES
PERFORMANCE
Direct
No Impact
No Impact
No Impact
Negative
Mixed
Positive
Weak
Positive
Direct &
Indirect
No Impact
No Impact
No Impact
Weak
Negative
Weak
Positive
Positive
Weak
Positive
DECISION-
MAKING
Indirect
Weak
Positive
Weak
Positive
Weak
Negative
No Impact
n/a
Weak
Negative
Weak
Negative
PERFORMANCE
Direct
Strong
Positive
Strong
Positive
Negative
No Impact
n/a
Strong
Positive
No Impact
Direct &
Indirect
Strong
Positive
Strong
Positive
Weak
Negative
No Impact
n/a
Positive
No Impact
significantly negatively related. The institutional ownership variable is found
to be both significantly positively and negatively related to decision-making
variables in the model's set of regression equations — i.e., institutional owner-
ship is significantly positively related to the capital-to-labour ratio and signif-
icantly negatively related to outward orientation. "No impact" means that the
corporate governance variable in question was not found to be significant, either
positively or negatively, in the two sets of regression equations. "Weak positive"
or "weak negative" means that the governance variable was significant in one of
the equations in the two sets of regression equations. "Positive" or "negative"
and "strong positive" or "strong negative" means that the governance variable
was significant in two and more than two of the regression equations, respectively.
SUMMARY
IN GENERAL, THE CORPORATE PERFORMANCE OF AMERICAN FIRMS is positively
correlated (both directly and indirectly) by institutional ownership, the inside
director ratio, and inside ownership, and negatively correlated with corporate
ownership concentration. The Canadian results are not as strong and robust as
the American findings.
71
RAO & LEE-SING
The inside director ratio and the foreign director ratio seem to have a
positive influence on the performance of Canadian firms. Similarly, the size of
the Board seems to exert a positive influence (mainly indirectly via leverage
and capital formation) on the economic performance of Canadian firms. The
growth performance of Canadian firms is also significantly positively related to
the foreign director ratio. Likewise, the growth performance of companies
where the CEO is also the Chairperson of the Board is significantly better than
companies where the two positions are not held by one person.
In general, the economic performance (productivity, growth and
profitability) of both Canadian and American firms with sales under
US$ 100 million is considerably inferior to the performance in the largest
size class (sales over US$ 2 billion).
CONCLUSIONS
T
HE MAJOR OBJECTIVE OF THIS STUDY has been to examine the corporate
governance structure in Canada and the United States, and to provide an
empirical analysis of the degree of association between the governance, corporate
decision, and performance variables in the two countries. Toward this goal,
drawing from three major sources of company data, a large database on gover'
nance, decision and performance variables for 766 Canadian and 3,000
American firms was created. Some of the major findings of our study follow.
• The majority of Canadian firms in all size classes and in most
industry groups is legally controlled (one or a small group of share-
holders owning, directly or indirectly, more than 50 percent of the
voting shares of the company). On average, 55 percent of
Canadian firms are legally controlled, compared to less than
25 percent of American firms.
• Differences in the ownership structure of American and
Canadian companies are more pronounced for very large firms.
• In the Canadian sample, on average, over 20 percent of company
shares are held by insiders (directors or officers of the firm),
compared to less than 10 percent in the United States.
• On the other hand, U.S. firms, on average, exhibit a much higher
level of institutional ownership (percentage of shares held by
institutional investors) than Canadian firms (53.3 percent versus
38.2 percent).
• In the two samples, the number of directors and senior officers
increases with firm size (measured by total sales).
72
GOVERNANCE STRUCTURE, CORPORATE DECISION-MAKING & FIRM PERFORMANCE
• On average, the inside director ratio (the number of directors
who are also officers of the firm over the total number of directors)
in the Canadian sample (20 percent) is lower than in the U.S.
sample (22 percent).
• The foreign director ratio averages 15 percent for Canadian firms.
• On average, only 34-5 percent of Canadian firms have the CEO
as the Chairperson of the Board, compared to over 60 percent for
U.S. firms.
• Profitability, productivity and growth performance of American
firms are significantly positively correlated (both directly and
indirectly) by institutional ownership, the inside ownership
ratio, and the inside director ratio.
• The economic performance of widely held and effectively
controlled American firms is significantly better than the perfor-
mance of legally controlled firms.
• In Canada, the inside director ratio has a positive effect on
corporate performance.
• The growth performance of Canadian firms, especially asset
growth, is positively related to the foreign director ratio.
• The size of the Board, and whether the CEO is the Chairperson
of the Board, do not seem to matter much for corporate perfor-
mance in the two countries.
In short, our findings indicate that the governance structure variables are
not strongly correlated with the corporate decision-making and performance
variables. However, the governance variables related to ownership — institu -
tional ownership, inside ownership, and concentration of ownership — are
strongly correlated with performance variables in the United States.
These results imply the need for different policy approaches in the two
countries. In the United States, governments and corporate actions aimed at
improving institutional activism, increasing the inside ownership and reducing
the ownership concentration would improve corporate performance. In
Canada, on the other hand, government and corporate efforts to improve
corporate governance practices such as reducing executive entrenchment and
differential voting rights, enacting minority shareholder provisions, enforcing
director liability, and enhancing disclosure requirements might be more
relevant for improving corporate performance.
73
RAO & LEE-SING
Efforts to develop data on corporate governance practices would be very
helpful in shedding further light on the importance of corporate governance
for firms' adaptability, flexibility and dynamism. Future Industry Canada
research in this area may focus on this effort.
ENDNOTES
1 The Toronto Stock Exchange Committee on Corporate Governance in
Canada, "'Where Were The Directors?' Guidelines For Improved Corporate
Governance in Canada," Draft Report (May 1994).
2 This study uses accounting-based, rather than market-based, measures of
firm performance. Market-based performance measures are both easy to
compute and reliable. However, they cannot be used to evaluate a firm's
decision-making and performance in absolute terms, but only in relation
to market expectations of the firm. The expected performance of a well-
managed firm will be built into the firm's current stock price and only
unexpected deviations will be picked up by the future stock performance.
3 The Canadian and American Corporate Governance Environment
Summaries are from Framework Conditions for Industry Draft Agenda,
OECD, 1994.
4 The definition of inside directors used in this study differs from that used by
Amoako-Adu and Smith (1995). Their definition also includes the
directors appointed from employees of parent companies, subsidiaries and
affiliates. Therefore, our estimate of the inside director ratio will be biased
downward, compared to the estimates of Amoako-Adu and Smith (1995).
However, it is difficult to determine, a priori, which of the two measures
captures better the independence of directors.
5 The Canadian Business Corporations Act (CBCA) uses country of citizen-
ship to distinguish between domestic and foreign directors. Because the
Compact Disclosure data sources used in this study do not include citizen-
ship information, we use the country of residency to estimate the foreign
director ratio.
6 The concentration of economic power in Canadian and American firms
should also take into account executive entrenchment. However, like
concentration, executive entrenchment is significantly higher in Canada.
Canadian executives are much more entrenched than their American
counterparts for several reasons: the absence of class actions by shareholders
in the Canadian legal system; the consequent difficulty of dismantling
poison pills and other takeover defenses in Canada; and the coattails
provisions in Canadian corporate acquisitions that, while benefiting small
shareholders, make the transfer of corporate control more expensive. The
74
GOVERNANCE STRUCTURE, CORPORATE DECISION-MAKING & FIRM PERFORMANCE
widespread use of multiple classes of shares in Canada adds further to the
entrenchment of Canadian executives.
7 The relationship between firm performance and ownership concentration,
taking into account size, industry, governance structure, and corporate
decision-making variables, is discussed later in this section.
8 Patricia Lipton, Head of the Wisconsin State Retirement System.
9 In the section on the Analytical Framework of this study, we assume that the
relationships among the governance, decision-making, and performance
variables are linear. However, we tested for the presence of non-linear
relationships among the variables, but did not include the results. In
general, the non-linear regression results are very similar to the results
from the linear model. Nevertheless, the rate of return and growth perfor-
mance variables are found to be somewhat non-linearly related to the
Inside Director ratio.
10 The reduced form equations regress the performance variables only on the
governance, industry and size variables. However, a Table summarizing the
results of the reduced equations, which would be similar to Tables 5 and 6,
is not included.
75
RAO & LEE-SING
APPENDIX 1
SOURCES AND CHARACTERISTICS OF THE DATABASE
T
HIS APPENDIX BRIEFLY DESCRIBES the data sources, data grouping, and the
characteristics of sample companies in the two countries.
SOURCES OF MICRO DATA
THE DATABASE FOR THIS STUDY IS DEVELOPED FROM three main sources of
company financial data: Compact Disclosure Canada, Compact Disclosure
SEC, and Compact Disclosure Worldscope. All three databases are provided in
CD-ROM format by Disclosure Incorporated.
1
 When necessary, information
was supplemented by Moody's International Company Database provided by
Moody's Investors Services.
The Compact Disclosure Canada database consists of over 8,500
Canadian public and private companies and Crown corporations. Companies
included in the database are those that are incorporated in Canada, either
federally or in one of the provinces or territories, and that trade on a
Canadian exchange. Company records include annual financial statement
data, interim financial statement detail, stock issue data, legal and company
status indicators, and textual data.
The Compact Disclosure SEC database consists of over 12,000 public
companies that file reports with the U.S. Securities and Exchange Commission.
Company data is extracted from the company's 10K or 20F statements.
The Compact Disclosure Worldscope database consists of over 11,000
public companies from 40 countries. Company data is obtained from annual and
periodic reports filed with each of the national stock exchange commissions.
Two criteria were used to select the firms that comprise the database for
this study. First, only those firms that disclosed a full set of corporate gover-
nance variables are included. Second, only those firms that had sales and
assets values greater than US$ 10 million are included. The final database
consists of 766 Canadian firms and 3,000 American firms.
The two country samples refer to December 1993, or the latest reported
date. However, data related to both level and growth values of the firms'
corporate decision and corporate performance variables are based on five-year
averages for the period 1988 to 1993.
2
Each firm in the sample is categorized by a size class and a major industry
group. The size of the firm is measured by the total dollar value of sales. The
firm's major industry group is based on the firm's primary business activity.
Individual firm data is aggregated into these size classes or major industry groupings
using the weighted average method, which permits the major industry or size
averages to capture the relative importance of the firms within the grouping.
3
76
GOVERNANCE STRUCTURE, CORPORATE DECISION-MAKING & FIRM PERFORMANCE
CHARACTERISTICS OF SAMPLE FIRMS
Size Distribution
The firms in our study vary in size from a minimum of US$ 10 million in sales
and assets to US$ 138 billion in sales and US$ 219 billion in assets for the
United States sample and US$ 16.5 billion in sales and US$ 102 billion in
assets for the Canadian sample. Total sales and assets of the Canadian firms
are US$ 397 billion and US$ 1.2 trillion while total sales and assets for the
American sample are US $3.6 trillion and US $7.0 trillion.
In an effort to examine the differences in governance structure between
firms of different sizes, both the Canadian and the American firms are grouped
into six size classes based on the dollar value of their total sales. Table Al-1
displays the size classes used in this study.
Although the size groupings are chosen somewhat arbitrarily, they seem
to capture adequately the differences in governance variables by the size of
sales. In general, the Canadian sample is dominated by more smaller firms
than the American sample. Over 80 percent of the Canadian firms have sales
less than US$ 500 million while only just over 70 percent of the American
firms fall into the same category. On the other hand, the American sample is
dominated by extremely large firms — almost 19 percent of the American firms
have sales over US$ 1 billion compared to only about 11 percent of the
Canadian firms (Figure Al-1). However, the average size of sales and assets for
Canadian firms with sales below US$ 2 billion compares very favourably with
those of their U.S. counterparts (Table A3-1).
The most striking feature of the sample firms is that although large firms
(sales over US$ 1 billion) account for a small proportion (less than 20 per-
cent) of total firms, they contribute to over 70 percent of sales and assets in
the two samples (Table A3-2).
TABLE Al- 1
SIZE CLASSES DETERMINED BY SALES VOLUME
SIZE CLASS SALES CRITERIA (US$ MILLIONS)
1 10 to 50
2 50 to 100
3 100 to 500
4 500 to 1,000
5 1,000 to 2,000
6 > 2,000
77
RAO & LEE-SING
FIGURE Al-1
SAMPLE DISTRIBUTION OF CANADIAN AND AMERICAN FIRMS BY SIZE CLASS
Source: Based on Table A3-1.
Industry Distribution
In order to capture the differences in governance structure and practices across
industrial activities, the 3766 companies are grouped into three digit industry
groups based on their U.S. Standard Industrial Classification 1986 (SIC) code
as determined by the companies' primary business activity. These industries
are further aggregated into 11 major industry groups.
4
 Table Al-2 describes the
major industry groupings used in this study.
Table A3-3 depicts the average sales and asset value of firms in the 11
major industries. Unlike the size classes, the average sales size of American
firms tend to be at least twice as large as their Canadian counterparts in all the
major industry groups except in the Agriculture, Forestry and Fishing indus-
tries, and the Finance, Insurance and Real Estate industries. The higher ratios
of average sales to average assets for the U.S. firms implies that, on average,
capital productivity of the American firms is significantly higher than the pro-
ductivity of Canadian firms.
5
78
GOVERNANCE STRUCTURE, CORPORATE DECISION-MAKING & FIRM PERFORMANCE
TABLE A1-2
COMPONENTS OF THE MAJOR INDUSTRY GROUPINGS
MAJOR INDUSTRY GROUPING INDUSTRY
Agriculture, Forestry and Fishing
Construction
Finance, Insurance and Real Estate
Labour-intensive Manufacturing
Mining
Resource-Intensive Manufacturing
Retail Trade
Services
Technology-Intensive Manufacturing
Transportation and Public Utilities
Wholesale Trade
Depository Institutions
Non-depository Institutions
Securities and Brokers
Insurance
Other Financial Services
Clothing
Furniture and Fixtures
Leather and Products
Miscellaneous Manufactured Goods
Printing and Publishing
Textiles
Fabricated Metals
Food and Products
Lumber and Wood
Non-Metallic Minerals
Paper and Allied
Petroleum Refining
Primary Metals
Tobacco
Commercial Services
Health Services
Other Services
Aircraft and Parts
Chemicals and Allied
Communications Equipment
Computer and Office
Electrical Products
Light Machinery
Machinery, excluding Electrical
Miscellaneous Electrical Products
Motor Vehicles and Equipment
Other Transportation Equipment
Rubber and Products
79
RAO & LEE-SING
FIGURE A1-2
SAMPLE DISTRIBUTION OF CANADIAN AND AMERICAN FIRMS BY INDUSTRY GROUPING
Agriculture, Forestry &. Fishing
Construction
Finance, Insurance & Real Estate
Labour-intensive Manufacturing
Mining
Resource-Intensive Manufacturing
Retail Trade
Services
Technology-Intensive Manufacturing
Transportation & Public Utilities
Wholesale Trade
Source: Based on Table A3-4-
Over three-quarters of the Canadian firms are from just five major industries:
Finance, Insurance and Real Estate; Mining; Resource-Intensive Manufacturing;
Technology-Intensive Manufacturing; and Transportation and Public Utilities.
In addition, these five major industries account for 85 percent of the total
sales and over 95 percent of the total assets in the Canadian sample
(Appendix 3, Table A3-4). The Finance, Insurance and Real Estate industry,
alone, represents close to one-quarter of all Canadian firms and accounts for
nearly three-quarters of the total assets of the Canadian sample.
The Finance, Insurance and Real Estate industry also plays an important
role in the American sample with just over 20 percent of the number of firms
and just over half of the total assets. The Technology-Intensive Manufacturing
80
GOVERNANCE STRUCTURE, CORPORATE DECISION-MAKING & FIRM PERFORMANCE
and Services industries are better represented in the American sample, while
the Mining industry has a higher proportion of firms, sales and assets in the
Canadian sample (Figure A1-2).
Although the sample selection method did not appear to have intro-
duced any systematic bias toward any major industry in the two countries, it is
interesting to note that the Canadian sample has a higher proportion of firms
in Finance, Insurance and Real Estate, the Mining, and the Resource-
Intensive Manufacturing industries than in the United States. On the other
hand, the American sample has a higher proportion of Technology-Intensive
Manufacturing and Services industries. The industrial distribution of firms in
the two samples reflects the comparative advantage position of the two coun-
tries (Eden, 1994, and the Industry Canada Working Paper Number 1).
ENDNOTES TO APPENDIX 1
1 Disclosure Incorporated, 5161 River Road, Bethesda, MD 20816.
2 Although accounting-based measures are ideally suited to measure the
effectiveness of managers' decision-making and firm performance, the
numerous conventions that are necessary in implementing an accounting
system often leave us perplexed about accounting measures. To overcome
both these ambiguities (which tend to be resolved over extended time
periods) and short-term fluctuations, our analysis includes a five-year aver-
age of accounting variables.
3 The weights used in the averaging process are the value of the variables in
the denominators of the ratios being averaged. For instance, employment
values are used as weights in calculating the average value of the
Asset/Employment ratio.
4 This industry grouping is consistent with the commonly used practices of
the (former) Economic Council of Canada and the OECD.
5 The average value of capital productivity, the ratio of average sales to aver-
age assets, is 0.3416 for Canada and 0.5139 for the United States. The
capital productivity, excluding the Mining industry, averages 0.3326 and
0.5023 for Canada and the United States, respectively, implying that the
lower average value in Canada is not due to the inclusion of the Mining
industry.
APPENDIX 2
GOVERNANCE ENVIRONMENT IN NORTH AMERICA
(See Table A2-1 and Table A2-2 on the following pages.)
8J
TABLE A2-1
CANADIAN GOVERNANCE ENVIRONMENT
CONCENTRATION
OF OWNERSHIP
AND CONTROL
Laws and Competition
Regulations and certain
financial and
investment
legislation
regulate the
concentration
of ownership
and control.
Standard High level of
Practices corporate
ownership
concentration.
Approval
required for
certain
mergers and
acquisitions.
STRATEGIC
ROLE OF
BOARDS
Directors are
to manage or
supervise the
management
of corporations.
Directors must
retain ultimate
control over
the corporation.
Directors
manage or
supervise the
management
of corporations.
Institutional
investors are
increasing their
influence.
STRATEGIC
INFORMATION
OF OWNERS
Corporate and
securities laws
ensure
widespread
information
available to
the public.
Strategic
information is
protected.
Detailed
annual,
quarterly
information on
public
companies
is provided.
Analytical
information
available from
securities
industry.
USE OF TAKE-
OVERS AS
RESTRUCTURING
TOOLS
Takeovers
usually
allowed as
governance
tools. Minority
shareholders
generally
protected.
Corporate and
corporation
articles
generally do not
discourage take-
overs. Few anti-
takeover
devices upheld
by courts.
CONCENTRATION
OF
CREDITORS
Guidelines
with respect to
federal
deposit-taking
institutions (DTIs)
discourage
concentration
of credit positions.
Standard
practice for
larger firms is
small-size bank
loans and
dispersed bond
ownership.
COMBINATION OF
EQUITY AND CREDIT
BY "UNIVERSAL"
INVESTORS
DTIs are not
restricted in
combining
debt and equity.
Limits are imposed
on the amount of
equity DTIs provide
for non-financial
firms.
Equity and
debt are
separate
instruments.
Financial non-
DTls are
increasing
their equity
holdings
substantially.
USE OF
BANKRUPTCY
AS EXIT
TOOLS
Comprehensive
exit procedures
provided for
corporations
and financial -
institutions.
Procedures
aimed at
survival.
Creditor or
debtor may
invoke both
exit and
reorganization
processes.
RATE OF
RETURN OF
INVESTMENT
Corporate law
stresses
maximization
of shareholder
value as key
objective.
Equity
securities
values are
traded
according to
returns.
TABLE A2-1
Implicit
Rules
Observed
Outcomes
(CONT'D)
CONCENTRATION
OF OWNERSHIP
AND CONTROL
Corporations
aiming at
concentrated
ownership
generally go
private.
Vast majority
of Canadian
corporations
are privately
held. Over
80% of public
corporations
have a
dominant
shareholder.
Source: Corporate Governance Branch,
STRATEGIC
ROLE OF
BOARDS
Boards work
closely with
senior
management.
They are
increasingly
independent.
Boards are
increasing
management
and
supervision of
management
of corporations
to avoid
potential
liabilities.
industry Canada.
STRATEGIC
INFORMATION
OF OWNERS
Widespread
use of financial
information.
Aggregate
statistical
information
available.
Public
information
widely used.
Strategic
information
generally not
available.
USE OF TAKE-
OVERS AS
RESTRUCTURING
TOOLS
General
acceptance of
takeovers as
part of
adjustment
process.
Increasing
number of
takeovers.
Poison pills
often
disallowed by
courts.
CONCENTRATION
OF
CREDITORS
DTls and
bondholders
keep arm's-
length
relations wit h
borrowing
companies.
Highly
dispersed
credit positions
with respect to
large firms.
COMBINATION OF
EQUITY AND CREDIT
BY "UNIVERSAL"
INVESTORS
Ownership and
creditorship
entail different
rights and
responsibilities.
Financial non-
DTls have
increased their
involvement in
corporate
governance.
USE OF
BANKRUPTCY
AS EXIT
TOOLS
Bankruptcy
recognized as
a risk of
business.
Process seen
as providing
an orderly,
efficient and
fair real location
of assets.
Liquidations
outnumber
reorganizations.
Recent
legislation
promotes more
reorganizations
and survivals
of existing
entities.
RATE OF
RETURN OF
INVESTMENT
Corporations
recognized as
profit
maximizing
organization.
Average
Return on
Capital
1933 = 16.2%
1992=16.0%
TABLE A2-2
UNITED STATES GOVERNANCE ENVIRONMENT
Laws and
Regulations
Standard
Practices
CONCENTRATION
OF OWNERSHIP
AND CONTROL
Securities and
investment
legislation
contains, and
discourages,
the concentration
ot ownership
and control.
(Private stock
exchange rules
discourage
concentration
of ownership.)
Dispersed
owners do not
usually
exercise their
proxy rights.
STRATEGIC
ROLE OF
BOARDS
Regulatory
rules
discourage
shareholders
entering the
boards. But
legally the
board rules and
fiduciary duties
of directors
were recently
strengthened.
Standard
corporate
charters and
practice do not
provide tor
strategic/
operational role
for boards.
STRATEGIC
INFORMATION
OF OWNERS
Company and
securities law
prescribe
widespread
public financia l
information.
Strategic and
prospective
information
from within
firms is
classified.
Detailed
quarterly
financial
information,
additional
analytical
information by
securities
firms.
"Price sensitive"
insider
information
not diffused.
USE OF TAKE-
OVERS AS
RESTRUCTURING
TOOLS
Legal
recognition of
take-overs as
governance
tool, recent
state anti-take-
over
legislation.
Corporate
charters
designed to
permit take-
overs; recent
toleration of
anti-takeover
devices.
CONCENTRATION
OF
CREDITORS
Banking
legislation
discourages
concentration
of credit
positions.
Standard
practice is
small-size
bank loans and
dispersed
bond
ownership.
COMBINATION OF
EQUITY AND CREDIT
BY "UNIVERSAL"
INVESTORS
Combination of
equity and
debt forbidden
by law.
Equity and
debts are
stringently
separated
instruments
but recent rise
in the use of
hybrids.
USE OF
BANKRUPTCY
AS EXIT
TOOLS
Elaborated
bankruptcy
legislation. In
bankruptcy,
creditor claims
may be
subordinated
to company
survival.
Standard
bankruptcy
procedures are
available.
Creditors are
incited to use
them rather
than negotiate
out-of-court
rescue
arrangements.
RATE OF
RETURN OF
INVESTMENT
Company law
stresses
profitability as
key corporate
objective.
Equity
securities
valued and
ownership
trade
according to
returns.
TABLE A2-2 (CONT'D)
CONCENTRATION
OF OWNERSHIP
AND CONTROL
Implicit Corporations
Rules aiming at
concentrated
ownership and
control are
expected to go
"private" (no
public issuance of
securities).
Observed Dispersed
Outcomes ownership in
industry, but
some recent
consolidation
through
institutional
ownership.
Source: OECD Framework Conditions for
STRATEGIC
ROLE OF
BOARDS
The board
needs to
tunction as an
amicable and
advisory body
to the CEO.
But this under-
standing is
changing for a
more inde-
pendent role.
Large majority
of boards
dominated by
management,
but recent
cases of
reversals.
Industry, Draft.
STRATEGIC
INFORMATION
OF OWNERS
Widespread
credibility and
use of financial
information.
Other more
strategic
information
should be made
available only if
evenly available
to all investors.
Excellent
diffusion of
public
information,
containment ot
insider
strategic
information
USE OF TAKE-
OVERS AS
RESTRUCTURING
TOOLS
Perceived
legitimacy and
acceptance of take-
overs as part of
adjustment
process.
Large number
of take-overs
but recent
curbs in
several states.
CONCENTRATION
OF
CREDITORS
Banks and
bondholders
keeps arm's-
length relations
with borrowing
companies.
Highly
dispersed
credit
positions.
COMBINATION OF USE OF
EQUITY AND CREDIT BANKRUPTCY
BY "UNIVERSAL" AS EXIT
INVESTORS TOOLS
Ownership and Bankruptcy risks
creditorship are recognized
entail essentially and accepted.
different rights
and
responsibilities.
Very few Large number
investors of exits settled
have an via
owner and bankruptcies.
creditor Possible
perspective. excess in the
number of
bankruptcies.
RATE OF
RETURN OF
INVESTMENT
Companies
recognized as
as profit-
maximizing
organizations.
US-BW 1000
companies,
return on
equity
1993 = 18.4%
1992 = 14.9%
RAO & LEE-SING
APPENDIX 3
CORPORATE GOVERNANCE STRUCTURE:
DETAILED TABULATIONS
TABLE A3-1
DATABASE SUMMARY
AVERAGE ASSETS BY
- NUMBER OF FIRMS, AVERAGE SALES AND
SIZE CLASS (US$ MILLIONS)
CANADA
SALES CLASS
10 to 50
50 to 100
100 to 500
500 to 1,000
1,000 to 2 ,000
> 2,000,000
#OF
FIRMS
284
120
221
54
37
50
AVERAGE
SALES
25,808
72,422
238,012
708,941
1,375,530
4,796,452
AVERAGE
ASSETS
92,733
228,667
439,353
3,548,411
2,567,314
14,532,434
UNITED STATES
#OF
FIRMS
725
491
948
279
204
353
AVERAGE
SALES
28,066
71,760
237,393
681,204
1,411,930
8,064,429
AVERAGE
ASSETS
117,643
186,368
464,385
1,385,208
2,454,547
15,614,259
TABLE A3 -2
DATABASE DISTRIBUTION - FIRMS, SALES AND ASSETS BY SIZE CLASS (US$ MILLIONS)
CANADA
SALES CLASS
10 to 50
50 to 100
100 to 500
500 to 1,000
1,000 to 2,000
> 2,000,000
%OF
FIRMS
37.1
15.7
28.9
7.0
4.8
6.5
% OF TOTAL
SALES IN
SAMPLE
2
2
13
10
13
60
% OF TOTAL
ASSETS IN
SAMPLE
2
2
8
16
8
62
UNITED STATES
%OF
FIRMS
24.2
16.4
31.6
9.3
6.8
11.8
% OF TOTAL
SALES IN
SAMPLE
1
1
6
5
8
79
% OF TOTAL
ASSETS IN
SAMPLE
1
1
6
6
7
79
86
GOVERNANCE STRUCTURE, CORPORATE DECISION-MAKIN G &. FIRM PERFORMANCE
TABLE A3-3
DATABASE SUMMARY - NUMBER OF FIRMS, AVERAGE SALES AND AVERAGE ASSETS
BY MAJOR INDUSTRY GROUPING (US$ MILLIONS)
CANADA
MAJOR
INDUSTRY
GROUPING
Agriculture, Forestry
& Fishing
Construction
Finance, Insurance
& Real Estate
Labour-intensive
Manufacturing
Mining
Resource-Intensive
Manufacturing
Retail Trade
Services
Technology-Intensive
Manufacturing
Transportation
& Public Utilities
Wholesale Trade
Simple Average
TOTAL
No. OF
FIRMS
4
9
186
34
136
99
34
53
92
82
37
766
AVERAGE
SALES
219,492
102,962
749,953
298,532
232,421
509,715
818,165
106,200
682,665
610,032
481,884
519,087
AVERAGE
ASSETS
115,817
90,279
4,650,208
254,979
469,695
580,386
333,030
109,232
316,210
1,399,244
187,504
1,519,712
UNITED STATES
TOTAL
No. OF
FIRMS
7
34
660
175
108
281
208
361
747
292
127
3,000
AVERAGE
SALES
108,216
659,264
767,971
526,732
1,492,697
2,090,011
1,850,427
347,574
1,474,520
1,731,911
922,843
1,201,821
AVERAGE
ASSETS
113,009
596,291
5,590,339
388,407
1,473,019
1,956,505
1,104,318
373,289
1,682,977
2,977,955
294,044
2,338,690
87
RAO & LEE-SING
TABLE A3-4
DATABASE DISTRIBUTION - FIRMS, SALES AND ASSETS
BY MAJOR INDUSTRY GROUPING
CANADA
MAJOR
INDUSTRY GROUPING
Agriculture,
Forestry & Fishing
Construction
Finance,
Insurance & Real Estate
Labour-intensive
Manufacturing
Mining
Resource-Intensive
Manufacturing
Retail Trade
Services
Technology-Intensive
Manufacturing
Transportation
& Public Utilities
Wholesale Trade
%OF
FIRMS
0.5
1.2
24.3
4-4
17.8
12.9
4.4
6.9
12.0
10.7
4.8
%OF
TOTAL
SALES
0
0
35
3
8
13
7
1
16
13
4
%OF
TOTAL
ASSETS
0
0
74
1
5
5
1
0
2
10
1
UNITED STATES
%OF
FIRMS
0.2
1.1
22.0
5.8
3.6
9.4
6.9
12.0
24.9
9.7
4.2
%OF
TOTAL
SALES
0
1
14
3
4
16
11
3
31
14
3
%OF
TOTAL
ASSETS
0
0
53
1
2
8
3
2
18
12
1
GOVERNANCE STRUCTURE, CORPORATE DECISION-MAKING & FIRM PERFORMANCE
TABLE A3 -5
CONCENTRATION OF OWNERSHIP BY SIZE CLASS
CANADA - % OF FIRMS
SALES (US$ MILLIONS)
10 to 50
50 to 100
100 to 500
500 to 1,000
1,000 to 2 ,000
> 2,000
Average of All Classes
TABLE A3-6
WIDELY
HELD
23.9
20.8
22.6
20.4
21.6
30.0
23.1
EFFECTIVE
CONTROL
23.6
24.2
20.4
25.9
8.1
12.0
21.4
CONCENTRATION OF OWNERSHIP BY MAJOR
LEGAL
CONTROL
52.5
55.0
57.0
53.7
70.3
58.0
55.5
UNITED STATES - % OF FIRMS
WIDELY
HELD
45.6
30.1
31.8
39.4
47.6
62.3
40.2
EFFECTIVE
CONTROL
31.0
39.3
37.2
39.4
34.8
28.6
35.1
LEGAL
CONTROL
23.4
30.6
31.0
20.8
17.6
9.1
24.7
INDUSTRY GROUPING
CANADA - % OF FIRMS
MAJOR
INDUSTRY GROUPING
Agriculture, Forestry
& Fishing
Construction
Finance, Insurance
& Real Estate
Labour-intensive
Manufacturing
Mining
Resource-Intensive
Manufacturing
Retail Trade
Services
Technology-Intensive
Manufacturing
Transportation &
Public Utilities
Wholesale Trade
Average of All Groups
WIDELY
HELD
50.0
22.2
15.1
17.6
42.6
21.2
11.8
15.1
24.0
17.0
32.4
23.1
EFFECTIVE
CONTROL
0.0
11.1
15.6
20.6
25.0
15.2
26.4
37.7
38.0
11.0
13.5
21.4
LEGAL
CONTROL
50.0
66.7
69.3
61.8
32.4
63.6
61.8
47.2
38.0
72.0
54.1
55.5
UNITED STATES - % OF FIRMS
WIDELY
HELD
28.6
26.5
58.2
33.1
31.5
36.7
31.7
21.6
35.3
56.5
33.9
40.2
EFFECTIVE
CONTROL
42.8
17.6
29.5
29.7
32.4
40.2
33.2
44.6
41-9
20.5
37.0
35.1
LEGAL
CONTROL
28.6
55.9
12.3
37.2
36.1
23.1
35.1
33.8
22.8
23.0
29.1
24.7
RAO & LEE-SING
TABLE A3-7
PERCENTAGE HELD BY ALL OF THE SIGNIFICANT SHAREHOLDERS, BY SIZE CLASS
SALES (US$ MILLIONS) CANADA - WEIGHTED AVERAGE
10 to 50 50.07
50 to 100 52.65
100 to 500 51.71
500 to 1,000 52.32
1,000 to 2,000 58.19
> 2,000,000 53.30
Weighted Average of All Classes 53.55
TABLE A3-8
PERCENTAGE HELD BY ALL OF THE SIGNIFICANT SHAREHOLDERS
BY MAJOR INDUSTRY GROUPING
MAJOR
INDUSTRY GROUPING CANADA - WEIGHTED AVERAGE
Agriculture, Forestry & Fishing 12.26
Construction 54-75
Finance, Insurance & Real Estate 42.25
Labour-intensive Manufacturing 51.49
Mining 53.98
Resource-Intensive Manufacturing 37.27
Retail Trade 58.69
Services 44-40
Technology-Intensive Manufacturing 80.63
Transportation & Public Utilities 63.77
Wholesale Trade 61.31
Weighted Average of All Groups 53.55
90
GOVERNANCE STRUCTURE, CORPORATE DECISION-MAKING & FIRM PERFORMANCE
TABLE A3 -9
NUMBER OF SIGNIFICANT SHAREHOLDERS WITH AT LEAST
10 PERCENT OWNERSHIP BY SIZE CLASS
SALES (US$ MILLIONS)
10 to 50
50 to 100
100 to 500
500 to 1,000
1,000 to 2,000
> 2,000
CANADA -
WEIGHTED AVERAGE
1.37
1.52
1.20
1.18
1.12
0.93
Weighted Average of All Classes 1 .03
UNITED STATES -
WEIGHTED AVERAGE
0.91
1.13
1.02
0.89
0.72
0.44
0.54
TABLE A3- 10
NUMBER OF SIGNIFICANT SHAREHOLDERS WITH AT LEAST
10 PERCENT OWNERSHIP BY MAJOR INDUSTRY GROUPING
CANADA -
MAJOR INDUSTRY GROUPING WEIGHTED AVERAGE
Agriculture, Forestry & Fishing 1 .60
Construction 3.30
Finance, Insurance & Real Estate 1.75
Labour-intensive Manufacturing 2.45
Mining 1.76
Resource-Intensive Manufacturing 2.23
Retail Trade 2.31
Services 2.67
Technology-Intensive Manufacturing 2.06
Transportation & Public Utilities 1 .88
Wholesale Trade 3.62
Weighted Average of All Groups 1 .03
UNITED STATES -
WEIGHTED AVERAGE
2.44
1.76
1.38
2.18
1.23
1.68
1.97
1.81
1.41
1.31
1.63
0.54
91
RAO & LEE-SING
TABLE A3- 11
PERCENTAGE OF VOTING SHARES HELD BY INSIDERS,
SALES (US$ MILLIONS)
10 to 50
50 to 100
100 to 500
500 to 1,000
1,000 to 2,000
> 2,000
Weighted Average of All Classes
BY SIZE CLASS
CANADA - UNITED STATES -
WEIGHTED AVERAGE WEIGHTED AVERAGE
34.58
24.52
13.18
15.13
20.98
22.18
21.49
25.71
25.77
22.30
15.31
14.13
7.46
9.60
TABLE A3- 12
PERCENTAGE OF VOTING SHARES HELD BY INSIDERS,
MAJOR INDUSTRY GROUPING
Agriculture, Forestry & Fishing
Construction
Finance, Insurance & Real Estate
Labour-intensive Manufacturing
Mining
Resource-Intensive Manufacturing
Retail Trade
Services
Technology-Intensive Manufacturing
Transportation & Public Utilities
Wholesale Trade
Weighted Average of AH Groups
CANADA -
BY MAJOR INDUSTRY GROUPING
UNITED STATES -
WEIGHTED AVERAGE WEIGHTED AVERAGE
n/a
n/a
2.69
9.97
54.14
26.82
n/a
24.13
1.99
22.15
n/a
21.49
17.88
10.58
10.76
13.90
3.05
9.50
19.52
14.99
6.99
5.18
16.27
9.60
Note: n/a = data not available
92
GOVERNANCE STRUCTURE, CORPORATE DECISION-MAKING & FIRM PERFORMANCE
TABLE A3-13
INSTITUTIONAL OWNERSHIP BY
SALES (US$ MILLIONS)
10 to 50
50 to 100
100 to 500
500 to 1,000
1,000 to 2 ,00 0
> 2,000
Weighted Average of All Classes
SIZE CLASS
CANADA -
WEIGHTED AVERAGE
31.36
31.59
33.05
36.86
44.26
38.78
38.24
UNITED STATES -
WEIGHTED AVERAGE
16.36
27.23
40.58
46.22
51.34
55.53
53.27
TABLE A3 -14
INSTITUTIONAL OWNERSHIP BY MAJOR INDUSTRY GROUPING
MAJOR INDUSTRY GROUPING
Agriculture, Forestry &. Fishing
Construction
Finance, Insurance & Real Estate
Labour-intensive Manufacturing
Mining
Resource- Intensive Manufacturing
Retail Trade
Services
Technology-Intensive Manufacturing
Transportation & Public Utilities
Wholesale Trade
Weighted Average of All Groups
CANADA -
WEIGHTED AVERAGE
9.57
44.20
28.75
39.58
40.27
26.70
24.16
22.98
68.73
52.82
20.31
38.24
UNITED STATES -
WEIGHTED AVERAGE
56.19
56.28
59.28
51.02
55.68
51.52
49.40
51.69
56.56
44.72
54.46
53.27
93
RAO & LEE-SING
TABLE A3- 15
NUMBER OF DIRECTORS BY SIZE CLASS
SALES (US$ MILLIONS)
10 to 50
50 to 100
100 to 500
500 to 1,000
1,000 to 2,000
> 2,000
Weighted Average of All Classes
CANADA -
WEIGHTED AVERAGE
7.34
8.27
9.81
11.92
11.89
16.66
1433
UNITED STATES -
WEIGHTED AVERAGE
7.38
8.29
9.62
11.62
12.70
17.74
16.36
TABLE A3- 16
NUMBER OF DIRECTORS BY MAJOR
MAJOR INDUSTRY GROUPING
Agriculture, Forestry & Fishing
Construction
Finance, Insurance & Real Estate
Labour-intensive Manufacturing
Mining
Resource-Intensive Manufacturing
Retail Trade
Services
Technology-Intensive Manufacturin g
Transportation & Public Utilities
Wholesale Trade
Weighted Average of All Groups
INDUSTRY GROUPING
CANADA -
WEIGHTED AVERAGE
12.28
8.93
10.79
11.28
10.54
12.51
11.83
9.60
11.24
13.70
11.40
14.33
UNITED STATES -
WEIGHTED AVERAGE
14.73
10.15
15.41
13.43
15.15
18.42
13.54
12.51
18.89
14.71
12.05
16.36
94
GOVERNANCE STRUCTURE, CORPORATE DECISION-MAKIN G &. FIRM PERFORMANCE
TABLE A3- 17
NUMBER OF OFFICERS BY SIZE CLASS
CANADA -
SALES (US$ MILLIONS) WEIGHTED AVERAGE
10 to 50
50 to 100
100 to 500
500 to 1,000
1,000 to 2,000
> 2,000
Weighted Average of All Classes
5.24
6.51
8.47
10.39
10.05
16.25
13.45
UNITED STATES -
WEIGHTED AVERAGE
9.02
8.96
9.76
11.48
12.03
15.86
14.84
TABLE A3- 18
NUMBER OF OFFICERS BY MAJOR INDUSTRY GROUPING
MAJOR INDUSTRY GROUPING
Agriculture, Forestry & Fishing
Construction
Finance, Insurance & Real Estate
Labour-intensive Manufacturing
Mining
Resource-Intensive Manufacturing
Retail Trade
Services
Technology-Intensive Manufacturing
Transportation & Public Utilities
Wholesale Trade
Weighted Average of All Groups
CANADA -
WEIGHTED AVERAG E
10.14
4.43
17.80
10.92
9.05
10.45
11.43
11.29
10.94
13.05
11.38
13.45
UNITED STATES -
WEIGHTED AVERAGE
13.79
12.31
18.27
12.31
14.24
16.57
12.23
12.61
14.21
14.87
11.35
14.84
95
RAO & LEE-SING
TABLE A3- 1 9
INSIDE DIRECTOR RATIO BY SIZE CLASS
SALES (US$ MILLIONS)
10 to 50
50 to 100
100 to 500
500 to 1,000
1,000 to 2,000
> 2,000
Weighted Average of All Classes
CANADA -
WEIGHTED AVERAGE
0.30
0.30
0.22
0.19
0.18
0.19
0.20
UNITED STATES -
WEIGHTED AVERAGE
0.39
0.35
0.31
0.25
0.24
0.21
0.22
TABLE A3 -20
INSIDE DIRECTOR RATIO BY MAJOR INDUSTRY GROUPING
MAJOR INDUSTRY GROUPING
Agriculture, Forestry & Fishing
Construction
Finance, Insurance & Real Estate
Labour-intensive Manufacturing
Mining
Resource-Intensive Manufacturin g
Retail Trade
Services
Technology-Intensive Manufacturing
Transportation & Public Utilities
Wholesale Trade
Weighted Average of All Groups
CANADA -
WEIGHTED AVERAGE
0.23
0.25
0.14
0.21
0.27
0.21
0.18
0.29
0.30
0.14
0.29
0.20
UNITED STATES -
WEIGHTED AVERAGE
0.13
0.28
0.23
0.27
0.31
0.19
0.24
0.27
0.20
0.19
0.32
0.22
96
GOVERNANCE STRUCTURE, CORPORATE DECISION-MAKING & FIRM PERFORMANCE
TABLE A3-21
FOREIGN DIRECTOR RATIO IN CANADIAN FIRMS BY SIZE CLASS
SALES (US$ MILLIONS) CANADA - WEIGHTED AVERAGE
10 to 50 0.19
50 to 100 0.19
100 to 500 0.18
500 to 1,000 0.18
1,000 to 2,000 0.24
> 2,000 0.11
Weighted Average of All Classes 0.15
TABLE A3-22
FOREIGN DIRECTOR RATIO IN CANADIAN FIRMS BY MAJOR INDUSTRY GROUPING
MAJOR INDUSTRY GROUPING CANADA - WEIGHTED AVERAGE
Agriculture, Forestry & Fishing n/a
Construction 0.14
Finance, Insurance & Real Estate 0.13
Labour-intensive Manufacturing 0.10
Mining 0.17
Resource-Intensive Manufacturing 0.19
Retail Trade n/a
Services 0.23
Technology-Intensive Manufacturing 0.15
Transportation & Public Utilities 0.17
Wholesale Trade 0.24
Weighted Average of All Groups 0.15
Note: n/a = data not available
97
RAO & LEE-SING
TABLE A3 -23
CEO is ON THE BOARD OF DIRECTORS BY SIZE CLASS
SALES (US$ MILLIONS)
10 to 50
50 to 100
100 to 500
500 to 1,000
1,000 to 2, 000
> 2,000
Average of All Classes
CANADA -
No
23.7
19.2
12.3
7.4
11.1
10.2
17.0
% OF FIRMS
YES
76.3
80.8
87.7
92.6
88.9
89.8
83.0
TABLE A3-24
CEO is ON THE BOARD OF DIRECTORS BY MAJOR INDUSTRY GROUPING
CANADA - % OF FIRMS
MAJOR INDUSTRY GROUPING No YES
Agriculture, Forestry & Fishing 25.0 75.0
Construction 22.2 77.8
Finance, Insurance & Real Estate 18.8 81.2
Labour-intensive Manufacturing 18.2 81.8
Mining 20.6 79.4
Resource-Intensive Manufacturing 16.3 83.7
Retail Trade 14.7 85.3
Services 14-0 86.0
Technology-Intensive Manufacturing 16.7 83.3
Transportation & Public Utilities 9.9 90.1
Wholesale Trade 16.2 83.8
Average of All Groups 17.0 83.0
98
GOVERNANCE STRUCTURE, CORPORATE DECISION-MAKING & FIRM PERFORMANCE
TABLE A3 -25
CEO is CHAIRPERSON OF THE BOARD BY SIZE CLASS
SALES (US$ MILLIONS)
10 to 50
50 to 100
100 to 500
500 to 1,000
1,000 to 2,000
> 2,000
Average of All Classes
CANADA
No
73.3
100.0
53.3
71.4
83.3
50.0
65.5
- % OF FIRMS
YES
26.6
0.0
46.6
28.6
16.7
50.0
34.5
UNITED STATES -
No
55.3
46.2
40.1
31.7
29.4
18.2
40.7
% OF FIRMS
YES
44.7
53.8
59.9
68.3
70.6
81.8
59.3
TABLE A3 -26
CEO is CHAIRPERSON OF THE BOARD BY MAJOR INDUSTRY GROUPING
CANADA - % OF FIRMS
MAJOR INDUSTRY GROUPING
Agriculture, Forestry & Fishing
Construction
Finance, Insurance & Real Estate
Labour-intensive Manufacturin g
Mining
Resource-Intensive Manufacturing
Retail Trade
Services
Technology-Intensive Manufacturing
Transportation & Public Utilitie s
Wholesale Trade
Average of All Groups
No
n/a
n/a
30.0
100.0
81.2
75.0
n/a
100.0
83.3
25.0
n/a
65.5
YES
n/a
n/a
70.0
0.0
18.8
25.0
n/a
0.0
16.7
75.0
n/a
34.5
UNITED STATES - % OF FIRMS
No
50.0
38.2
45.7
36.0
37.0
39.9
33.2
38.2
43.4
34.7
44-1
40.7
YES
50.0
61.8
54-3
64.0
63.0
60.1
66.8
61.8
56.6
65.3
55.9
59.3
Note: n/a = data not available.
APPENDIX 4
(See TableA4-l, Figure A4-1 and Figure A4-2 on following pages.)
99
TABLE A4-1
CONCENTRATION OF OWNERSHIP TABULATIONS
CANADA
VARIABLE
Number of Firms
Return on Equity
Return on Assets
Sales Growth
Asset Growth
WIDELY
HELD
177
0.06
0.03
0.15
0.14
Sales/Employee ($) 449,484
Sales/Asset
Labour Prod. Growth
Capital Prod. Growth
EPS Growth
Debt/Assets
0.78
0.14
0.01
-0.13
0.21
Assets/Employee ($) 888,503
Presence of Foreign Sales or Assets
Institutional Ownership
Inside Ownership
Board Size
Inside Director Ratio
Foreign Director Ratio
CEO is Chairperson of the Board
0.23
6.18
4.18
9.61
0.24
0.03
0.26
EFFECTIVELY
CONTROLLED
164
0.08
0.02
0.14
0.13
LEGALLY ALL
CONTROLLED FIRMS
425
0.06
0.03
0.08
0.07
398,493 412,880
1.08
0.14
0.01
-0.08
0.27
646,944 1,761
0.17
18.75
17.83
8.70
0.28
0.03
0.29
1.00
0.08
0.01
-0.16
0.26
,103
0.12
48.33
26.24
9.31
0.26
0.04
0.50
766
0.07
0.03
0.11
0.10
418,908
0.97
0.11
0.01
-0.14
0.25
1,324,969
0.16
32.26
19.34
9.25
0.26
0.04
0.35
WIDELY
HELD
1206
0.06
0.03
0.08
0.11
233,417
0.90
0.07
-0.02
-0.02
0.20
796,027
0.36
31.69
11.59
10.45
0.30
n/a
0.61
UNITED STATES
EFFECTIVELY
CONTROLLED
1054
0.04
0.03
0.12
0.13
219,865
1.15
0.11
-0.01
-0.09
0.20
544,701
0.46
41.18
18.90
9.85
0.30
n/a
0.58
LEGALLY
CONTROLLED
740
0.02
0.02
0.13
0.14
248,713
1.25
0.13
-0.01
-0.02
0.23
415,791
0.36
30.37
36.04
8.96
0.33
n/a
0.59
ALL
FIRMS
3000
0.05
0.03
0.11
0.13
232,420
1.07
0.10
-0.02
-0.05
0.21
613,140
0.40
34.70
20.19
9.87
0.31
n/a
0.59
Note: n/a = not available.
GOVERNANCE STRUCTURE, CORPORATE DECISION-MAKING & FIRM PERFORMANCE
FIGURE A4-1
RETURN PERFORMANCE GROUPED BY CONCENTRATION OF OWNERSHIP
FIGURE A4-2
GROWTH PERFORMANCE GROUPED BY CONCENTRATION OF OWNERSHIP
Source: Table A4-1 for both Figures.
101
RAO & LEE-SING
ACKNOWLEDGEMENTS
W
E ARE GRATEFUL TO Ross Preston and Denis Gauthier for their encourage-
ment and support on this project, and to Ron Daniels and Randall
Morck for their comments and suggestions at various stages of the study. We
would also like to thank David Stangeland, Giovanni Barone-Adesi and Lee
Gill for their many useful comments on the draft paper. Thanks are also due to
Ashfaq Ahmad and Marc Legault for their assistance throughout.
BIBLIOGRAPHY
Amoako-Adu, B. and B. F. Smith. "Outside Financial Directors and Corporate Governance
-Draft," 1995.
Bacon, J. "Corporate Boards and Corporate Governance." The Conference Board, Numbe r
1036(1993).
Baysinger, B. D. and H. N. Butler. "Corporate Governance and the Board of Directors:
Performance Effects of Changes in Board Composition," Journal of Law, Economics and
Organization, 1, 1 (1985):101-24.
Baysinger, B. D. and R. E. Hoskisson. "The Composition of Boards of Directors and
Strategic Control: Effects on Corporate Strategy," Academy of Management Review, 15, 1
(1990):72-87.
Beck, S. "The Corporation and Canadian Society," presented at the Canadian Corporate
Governance: A Multi-Disciplinary Perspective Conference held at the C.D. Howe
Institute. (February 1994).
Clarkson, M. B. E. and M. C. Deck. "'Straddling Fences Makes it Difficult to Walk in a
Straight Line': A Commentary Prepared in Response to '"Where Were The Directors?'
The Draft Report of the TSE Committee on Corporate Governance in Canada May
1994." The Centre for Corporate Social Performance and Ethics, (July 1994).
Dalton, D. R., I. F. Kesner, and P. L. Rechner. "Corporate Governance and Boards of
Directors: An International, Comparative Perspective." Advances in International
Comparative Management, 3 (1988):95-105.
Daniels, R. and P. Halpern. "The Canadian Quandary: Accounting for the Survival of the
Closely Held Corporation." Draft presented at the Canadian Corporate Governance: A
Multi-Disciplinary Perspective Conference held at the C.D. Howe Institute. (February
1994).
Daniels, R. and J. Macintosh. "Towards a Distinctive Canadian Corporate Law Regime."
Osgoode Hall Law Journal, 29, 4 (Winter 1991):864-933.
Densetz, H. and K. Lehn. "The Structure of Corporate Ownership: Causes and
Consequences." Journal of Political Economy, 93, 6 (1985).
Dey, P. et al. "Where Were The Directors?", The TSE Committee on Corporate
Governance in Canada. (May 1994).
Donaldson, G. "Voluntary Restructuring: The Case of General Mills." Journal of Financial
Economics, 27 (1990):! 17-141.
102
GOVERNANCE STRUCTURE, CORPORATE DECISION-MAKIN G & FIRM PERFORMANCE
The Economist. "A Survey of Corporate Governance." (January 29th, 1994).
Eden, L., Multinationals in North America, The Industry Canada Research Series. Calgary:
University of Calgary Press, 1994.
Friedlander, A. F., E. R. Berndt and G. McCullough. "Governance Structure, Managerial
Characteristics, and Firm Performance in the Deregulated Rail Industry." Brookings
Papers: Microeconomics. 1992.
Herzel, L. "Corporate Governance Through Statistical Eyes." Journal of Financial Economics,
27(1990):581-93.
Industry Canada, Institutional Activism in Canada (Draft), Corporate Governance Branch,
September 1994.
Industry Canada. Economic Integration in North America: Trends in Foreign Direct Investment
and the Top 1,000 Firms, Working Paper Number 1. (January 1994).
Jensen, M. C. "The Modern Industrial Revolution, Exit, and the Failure of Internal Control
Systems." Journal o/Finance, 48, 3 (July 1993):831-80.
Jensen, M. C. and W. Meckling. "Theory of the Firm: Managerial Behaviour, Agency Costs
and Ownership Structure." Journal of Financial Economics, 3 (October 1976):305-60.
Macintosh, J. G. "The Role of Institutional and Retail Investors in Canadian Capital
Markets." Osgoode Hall Law Journal, 31,2 (1993).
Morck, R. K. and D.A. Stangeland. "Corporate Performance and Large Shareholders"
(Working Paper No. 4-94). Institute for Financial Research, Faculty of Business,
University of Alberta, 1994.
Morck, R. K. "On the Economics of Concentrated Ownership," Draft presented at the
Canadian Corporate Governance: A Multi-Disciplinary Perspective Conference held at
the C.D. Howe Institute, February 1994.
Neave, E. H., "Organizational Economics and Directors' Control (Draft presented at the
Canadian Corporate Governance: A Multi-Disciplinary Perspective Conference held at
the C. D. Howe Institute)," February 1994.
Porter, M. E. "Capital Disadvantage: America's Failing Capital Investment System."
Harvard Business Review. (September-October 1992):65-82.
. "Capital Choices: Changing the Way America Invests in Industry." The Council
on Competitiveness and The Harvard Business School. 1992.
Roe, M. J. "Political and Legal Restraints on Ownership and Control of Public Companies."
Journal of Financial Economics, 27 (1990):! 17-41.
Romano, R. "A Cautionary Note on Drawing Lessons from Comparative Corporate Law."
Yale Law Journal, 102 (1993): 2021-37.
. "Public Pension Fund Activism in Corporate Governance Reconsidered."
Columbia Law Review, 93, 4 (May 1993).
Roth, C. W. "Concentration of Ownership and the Composition of the Board: An
Examination of Canadian Publicly-Listed Corporations " (Draft). Industry Canada,
1994.
103
This page intentionally left blank 
Vijay Jog
School of Business
Carleton University
A/it Tulpule O
Corporate Renaissance Group ^J
Ottawa
Control and Performance:
Evidence from the TSE 300
INTRODUCTION
S
INCE THE PUBLICATION OF the Jensen & Meckling (1976) study on agency
theory, a great deal of attention has been given to the potential for conflict
between owners and managers and the effect of that conflict on firm performance.
The existing empirical evidence has focused on the investigation of the
determinants and consequences of ownership structure, as well as on the
relationship between ownership structure and firm characteristics such as size,
leverage, and expenditure on research and development. The increasing number
of mergers and takeovers, and leverage and management buyouts in the 1980s
added impetus to this empirical research.
The relationship between ownership concentration and corporate
performance continues to be a debatable issue. The existing empirical evidence
consists mainly of U.S. studies, which test the cross-sectional relationship
between ownership and corporate performance, the latter measured either by
accounting criteria or by Tobin's Q, with little emphasis on the temporal
stability of results.
Summarizing their empirical research, Jensen & Warner (1988) conclude
that "the precise effects of stock holdings by managers, outside shareholders,
and institutions are not well understood, and the inter-relationships between
ownership, firm characteristics, and corporate performance require further
investigation". Similarly, Jog & Schaller (1989) assert that "... a lot needs to
be known before any direct connection can be made between the ownership
structure and the market value of equity".
The purpose of this study is to examine the possible relationship
between corporate control (as measured by the controlled ownership of the
firm) and the accounting and stock market performance of Canadian firms
listed on the Toronto Stock Exchange (TSE). The study focuses on the
Canadian companies that belonged to the TSE 300 Composite Index in the
14 years between 1978 and 1991 inclusive.
105
JOG & TULPULE
The study is organized as follows: a review of the research to date is
followed by sections that present the main thesis of the study, describe the
data employed, and discuss the methodology used. The last section presents
results and conclusions.
PREVIOUS RESEARCH
T
HE EXISTING RESEARCH can be reviewed broadly under two headings:
theoretical and empirical.
THEORETICAL RESEARCH
MOST OF THE BASIC TENETS OF THE AGENCY THEORY are well described in Berle
& Means (1932) and Jensen & Meckling (1976). In effect, there are at least
four sources of conflict that can arise between management (agents) and
external shareholders (principals).
• Management's tendency to consume some of the firm's resources
by way of perquisites.
• Managers have a greater incentive to shirk their responsibilities
as their equity interest in a firm falls.
• Managers may forgo profitable but risky projects, as they bear the
cost of failure more than the widely diversified shareholders.
• Managers may misrepresent the quality of future investment
projects, thereby causing the outside investors to demand a higher
risk premium.
In order to ensure that agents (management) act in the best interest of the
principals, the principals (owners) incur certain costs. These may include pay-
ment of incentives to the manager, as well as monitoring costs to "limit the
aberrant activities of the [managers]". Jensen & Meckling (1976) identify
three components of these agency costs: the monitoring expenditures by the
principals, the bonding expenditure by the agent
1
 and the residual loss.
2
It is these agency costs, resulting from the conflict between ownership
and managerial performance, that provide an explanation of the relationship
between corporate ownership and firm performance. It is argued that one way
to reduce this conflict, and thereby the agency costs, would be to have managers
own a substantial portion of the company's shares, thus automatically aligning
the managers' interests with those of the owners, and ensuring that the managers
act in the best interests of the owners. The higher the proportion of managerial
ownership, the lower the agency costs would be, and the better the performance
of the firm.
3
 Similar arguments can be extended to cases where a significant
block of shares is owned by a small number of investors who can, in turn, exert
106
CONTROL AND PERFORMANCE: EVIDENCE FROM THE TSE 300
exert a higher degree of monitoring on managers, thereby improving perfor-
mance. If this is true in either case, then one would expect to find a positive
relationship between corporate control and managerial performance: the higher
the stake in the ownership of the firm by either management or a significant
owner, the better the performance is likely to be.
On the other hand, Stulz (1988) argues that firm value increases as
managerial ownership increases, but firm value would decrease if insiders own
a large portion of the shareholders' equity. In the context of a possible tender
offer, Stulz shows that managerial ownership affects both the probability of a
takeover bid and the premium. When managers do not own any equity in the
firm, the offer premium is small. As managerial ownership increases from zero, the
premium offered also increases. However, a very high percentage of managerial
ownership will, in fact, be an impediment to a takeover attempt, thereby
reducing the firm's value. Stulz thus predicts a convex relationship between
firm value and managerial ownership, implying that a certain level of insider
ownership may exist, at which firm value is maximized .
Other theories, however, contend that the ownership of a firm and its
performance are unrelated. Although conflict between owners and managers
may indeed exist, there are other mechanisms that effectively resolve the
conflict and align the interests of the owners and managers. Such alternative
mechanisms include external monitors (auditors, bankers, institutional
investors), performance-related compensation schemes, choice of managers
with aligned and similar interests (Demsetz, 1983), the existence of an
efficient labour market (Fama, 1980), a competitive product market (Hart,
1983), large shareholder blocks (Stiglitz, 1985; Shleifer & Vishny, 1986), and
an increase in corporate disclosure demanded by government regulators
(e.g., the Canada Business Act) or by securities legislators (e.g., the Ontario
Securities Commission). According to these theories, alternative mechanisms
will effectively counter the agency problem, and so the performance of a firm
will be unrelated to its ownership structure.
In contrast to the above-mentioned theories which suggest a positive or
no relationship, Muellar (1986) postulates that corporate ownership and
performance are negatively related. He argues that even if its ownership is fairly
large, management may still divert funds for its personal use, provided the
benefits of such diversion are greater than the cost of receiving less return
from its ownership of the firm. Figure 1 shows a graphical representation of the
existing theoretical work.
EMPIRICAL RESEARCH
TABLE 1 SHOWS THE STATE OF EMPIRICAL RESEARCH in the area of corporate
performance and managerial ownership. Five conclusions emerge from a
review of this work. First, the findings are mixed, although there is some
evidence that performance is positively related to insider/manager ownership.
107
JOG & TULPULE
FIGURE 1
RELATIONSHIP BETWEEN FIRM PERFORMANCE AND OWNERSHIP STRUCTURE
AS SUGGESTED BY THEORETICAL RESEARCH
JM Jensen & Meckling (1976)
S Stulz(1988)
M Muellar(1986)
FIGURE 2
RELATIONSHIP BETWEEN FIRM PERFORMANCE AND OWNERSHIP STRUCTURE
AS EVIDENCED BY EMPIRICAL RESEARCH
MSV Morck, Shleifer & Vishny (1988)
CHH Chen, Hexter & Hue (1990)
W Wruck(1989)
DL Demsetz&.Lehn(1985)
NS Neun&Santerre(1986
HS Holderness & Sheehan (1991)
108
CONTROL AND PERFORMANCE: EVIDENCE FROM THE TSE 300
TABLE 1
SUMMARY OF EMPIRICAL RESEARCH
AUTHORS
Vance
Pfeffer
Demsetz & Lehn
Neun & Samterre
Morck, Shleifer
& Vishny
Holderness
&. Sheehan
Kim & Lyn
Wruck
Chen, Hexter,
&Hue
Oswald &. Jahera
YEAR
1964
1972
1985
1986
1988
1991
1988
1989
1990
1991
PERFORMANCE
Net Income
Profit Margin,
ROE
ARR / BV
Net Profit Margin
Tobin's Q
ARR, Tobin's Q
Tobin's Q
Firm Value
Tobin's Q
Excess Stock
Returns
OWNERSHIP
Directors' Share-
holding
Managerial Share-
holding
Top 5 Share-
holders' holdings
Dominant Stock
Ownership
Directors' Share-
holdings
Majority Share-
holding
Insider Ownership
Insider Ownership
Managerial
Ownership
Directors' &
Officers' Share-
holdings
FINDING
Positive
Relationship
Positive
Relationship
No
Relationship
S-shaped
Relationship
Increase,
Fall, Increase
No
Conclusions
Positive Relationship
Increase,
Fall, Increase
Large Firms: Positive
Relationship
Small Firms: No
Relationship
Positive
Relationship
Second, in a majority of the studies, performance is measured based on Tobin's
Q and is considered only for a specific year or two; the studies have not
attempted to show temporal stability of these results. This makes generalizations
about the relationship suspect. Third, there is some evidence of curvilinearity,
which is consistent with both the alignment and entrenchment hypotheses.
Fourth, there is some evidence (Oswald & Jahera, 1991) that both size and
ownership are statistically significant factors in explaining the performance of
a firm, an observation that seems also to corroborate those of Vance (1964),
Pfeffer (1972), and Kim & Lyn (1988).
4
 Finally, the definition of control varies
significantly across studies, thereby making the task of comparison quite
difficult. Figure 2 summarizes the existing empirical evidence from the U.S.-
based studies.
109
JOG & TULPULE
SUMMARY
IN SUMMARIZING THE THEORETICAL ARGUMENTS and the empirical evidence, it
can be said that the performance of a firm may depend upon its ownership
structure (as a manifestation of the agency problem), but no unambiguous
statements can be made about either the importance or the nature of the
dependency. It depends upon the efficacy of alternate mechanisms, including
disclosure and business rules imposed by regulators and government policies.
Another observation about the existing empirical research is that the use of
varied performance measures makes it difficult to draw unambiguous conclusions.
In addition, most studies are cross-sectional in nature and provide only a snap-
shot analysis (at a specific point in time). Such analyses do not show how
strongly the results would hold over a longer period of time or how robust
they are.
The empirical research presented in this section also reveals that all these
studies have been conducted using U.S. data. It is not clear whether these results
are applicable to the Canadian situation. The only other empirical study in the
Canadian context is by Morck & Stangeland (1994), who comment on some of
the conspicuous differences between American and Canadian companies.
5
 Thus,
it would be of considerable interest to know more, not only about the relation-
ship between performance and ownership, but also about the Canadian context,
especially where there is a significant difference of ownership percentage
between Canadian and American publicly traded firms.
TERMS OF REFERENCE
T
HIS STUDY ASKS the following questions: do companies that are closely
held perform better than those that are widely held? And, if so, does the
degree of ownership provide any indication of differential performance? This
relationship between corporate ownership and performance is examined, using
the ownership and performance data of Canadian companies listed on the
TSE 300 Composite Index, spanning the years from 1978 to 1991 inclusive.
MEASURES OF PERFORMANCE AND CONTROL
ALTHOUGH TOBIN'S Q, AS A MEASURE OF PERFORMANCE, appears to dominate
the U.S. studies, a lack of readily available replacement values for the sample
companies makes it virtually impossible to use this ratio in this study. Instead,
we follow two methods for evaluating firm performance. First, we use the stock
market returns as a measure of performance, as they reflect long-term
performance from an investor's viewpoint. Second, for a subset of the sample
firms (manufacturing sector), we employ widely used accounting-based
performance measures. This subset includes market-to-book ratio of equity as
a proxy for the Tobin's Q.
no
CONTROL AND PERFORMANCE: EVIDENCE FROM THE TSE 300
Control - i.e., how closely held (concentrated) or widely held (diffused)
within a company - is measured by the percentage of the outstanding equity of
the firm which is tradeable on the stock exchange. A firm whose entire out-
standing equity is tradeable on the stock exchange is considered to have very
diffused control (widely held), while a firm that has only a very small portion
of its outstanding equity tradeable on the stock exchange is considered to be
highly controlled (closely held). In addition, in order to account for obvious
differences in the degree of "assets in place" between manufacturing and non-
manufacturing firms, we evaluate firm performance in two separate groups,
based on their industry classification.
6
THE DATA
THE SAMPLE COMPANIES IN THIS STUDY include only those companies that were
included in the Toronto Stock Exchange 300 Composite Index (TSE 300)
between 1978 and 1991 inclusive. Control is measured using "float" percentages.
These indicate the percentage of shares available for outside ownership and are
used to calculate relative weights in the index, and can also be used to estimate
the degree of control. The data are obtained from the index listing published in
the TSE Review. The values are those at December 31 each year. The industry
groupings used here are based on the TSE 300 groupings.
In contrast with the United States, where there are a large number of
small banks, the Canadian banking industry is characterized by a small number
of large banks. Because of government-imposed ownership restrictions, all of
these banks are widely held. Also, utility companies in Canada are largely
regulated by the federal and provincial governments. For these reasons, the
sample used in this study excludes both banks and utility companies from the
data set.
To calculate stock market performance, monthly returns for individual
companies were drawn from the Toronto Stock Exchange/University of
Western Ontario (TSE/Western) database, which is modeled after the
University of Chicago Center for Research in Security Prices (CRSP) database,
spanning a period of 14 years from 1978 to 1991. The accounting data on
individual firms is gathered from the Stock Guide database maintained by
Stock Guide Publications Ltd. Due to the unavailability of such data on a
consistent basis prior to 1988, analysis based on the accounting measures is
restricted to the years 1988 to 1991.
The data set is comprised of information on a total of 613 companies, over
the 14-year period between 1978 and 1991. Table 2 shows the breakdown by
industry of the 613 companies, as classified according to the Industry Codes used
by the index. Due to the listing-delisting feature of the TSE 300 Index, the
constituency of the index is fluid. Table 3 shows the number of companies
forming part of the data set each year. The rows represent the "from" year and
the columns indicate the "to" year. The numbers along the principal diagonal
111
JOG&TULPULE
TABLE 2
FIRMS IN DATA SET: BREAKDOWN BY INDUSTRY
INDUSTRY
NUMBER INDUSTRY
1
3
4
5
6
7
8
9
11
12
14
MANUFACTURING
Metals and Minerals
Oil and Gas
Paper and Forest Products
Consumer Products
Industrial Products
Construction
Transportation
Manufacturing Sub-Total
NON MANUFACTURING
Pipelines
Communications
Merchandising
Management Companies
Non-Manufacturing Sub-Total
Manufacturing & Non-Manufacturing Total
2
10
13
EXCLUDED FROM ANALYSIS
Gold
Utilities
Financial Services (Banks)
Total Sample
NUMBER
63
113
24
44
71
23
10
348
6
30
57
23
116
464
59
25
65
613
% OF GROUP
18.10
32.47
6.90
12.64
20.40
6.61
2.87
100.00
5.17
25.86
49.14
19.83
100.00
% OF TOTAL
13.58
24.35
5.17
9.48
15.30
4.96
2.16
75.00
1.29
6.47
12.28
4.96
25.00
100.00
thus indicate the number of companies forming part of the data set in the given
year. The figure in cell C (i, j) indicates the number of companies forming part
of the data set, which were listed in year i, and continued to be listed through
year j. Table 3 also reveals that the data set includes at least 236 companies and
at the most 264 companies (see the diagonal numbers) for any given year.
112
TABLE 3
THE NUMBER OF COMPANIES IN THE DATA SET (1978-1991)
YEAR
3
 1978 1979 1980 1981 1982 1983 1984
1978 240 216 204 185 180 176 165
1979 236 218 196 190 183 172
1980 241 215 207 197 182
1981 249 234 218 200
1982 247 228 208
1983 247 225
1984 250
1985
1986
1987
1988
1989
1990
1991
«
1985 1986 1987 1988
155 141 126 114
162 148 132 120
169 153 138 124
182 161 144 129
188 167 148 129
198 176 156 137
219 194 173 152
243 210 189 166
249 223 194
254 220
257
1989
112
117
121
125
124
129
142
156
183
206
237
262
1990
107
112
117
121
120
124
137
150
173
195
221
245
260
1991
101
106
111
116
115
120
131
141
159
176
196
218
232
264
Note:
 a
 Horizontal rows indicate the "from" year; vertical columns indicate the "to" year.
JOG & TULPULE
TABLE 4
PORTFOLIO FORMATION AND RETURN CALCULATION PROCEDURE
PORTFOLIO
FORMATION
VARIABLE
Control
WEIGHT
VARIABLE 1
FOR
PORTFOLIO RETURNS
Equal
WEIGHT
VARIABLE 2
FOR
PORTFOLIO RETURNS
Size
SECTOR
GROUPS
Manufacturing,
Non-manufacturing
METHODOLOGY
S
INCE THIS STUDY USES both accounting and stock-market data, we adopt
two distinctly different approaches in our enquiry into firm performance.
STOCK-MARKET-BASED PERFORMANCE
To ANALYZE THE PERFORMANCE OF FIRMS that differ in their degree of control, we
employ a portfolio-grouping approach. Specifically, at the beginning of each
period all firms are first ranked in ascending order of control based on their year-
end value of the previous year. They are then placed in four groups repre-
senting the four quartiles. Within each group, the portfolio return is calcu-
lated in two ways — firs t by equally weighting the returns and then by
weighting the returns based on the market capitalization of the firms.
7
 In each
case, the portfolios are revised annually using the ownership and market
capitalization data as of December 31st each year, and calculated separately for
the manufacturing and non-manufacturing sectors. Table 4 shows this portfolio-
grouping procedure.
Since the portfolio approach is used to analyze the relationship between
control and performance, we use the well-known Sharpe measure for comparing
the relative performance. The Sharpe measure uses the ratio of average returns
and the corresponding standard deviation as a measure of performance. The
significance of the Sharpe portfolio performance measure, is calculated using
the Jobson & Korkie (1986) test.
8
 In addition, non-parametric and parametric
tests are also performed to test for the number of months and the relative
amount by which a particular portfolio outperformed another.
ACCOUNTING-BASED PERFORMANCE MEASURES
As NOTED EARLIER IN TABLE 1, many of the previous studies concentrate on
accounting-based performance measures. To facilitate comparisons, we
114
CONTROL AND PERFORMANCE: EVIDENCE FROM THE TSE 300
attempt to evaluate performance of our sample firms based on the eight most
commonly used measures: asset turnover, gross margin, return on assets, return
on equity, debt equity, capital expenditure-to-net fixed assets, interest coverage
and market-to-book-value ratio. The last can serve as a proxy for Tobin's Q.
Due to data availability constraints, and in order to ensure within-group
homogeneity, we conduct this analysis only for the years 1988 through 1991
and only for our manufacturing firm sub-sample.
9
RESULTS
AGGREGATE PORTFOLIO RESULTS - STOCK MARKET DATA
IN ORDER TO PROVIDE AN OVERVIEW of the potential differences between the
two sectors, the overall results of the sectoral portfolios are shown in Table 5,
before displaying the detailed breakdowns. Panel A of the table shows cumulative
wealth based on annual revisions for each of the three portfolio-weighting
schemes: equally weighted, control weighted and size weighted. The last column
shows the comparable annual values for the TSE 300.
10
These results indicate that the two sectors had significantly different
performance: the non-manufacturing sector portfolio, irrespective of the
weighting scheme, generated higher wealth than the manufacturing sector
portfolio. Moreover, the systematic risk of the non-manufacturing portfolio is
lower than that of the manufacturing portfolio. Panel C of the table (based on
average return percentage) indicates that the higher overall performance of
the non-manufacturing firms is due to the significantly higher performance
during the middle period, 1982-1986. Investigating the within-sector returns
shows the effect of the weighting variable. In the manufacturing sector, the
size weighted portfolio dominates the equally weighted portfolio indicating
that, based on the entire 1978-1991 period, large firms performed better than
small firms in this sector. However, the reverse is true for the non-manufacturing
sector. A comparison of the control weighted portfolio for the manufacturing
sector with the equally weighted portfolio shows that the former under-
performed the latter, indicating that a portfolio strategy of investing in
manufacturing companies based on the percentage of control would have
generated lower returns than either the equally weighted or the size weighted
strategy. The results for the non-manufacturing sector are exactly the opposite.
These results, shown without any statistical tests, indicate that if there is a
positive relationship between control and performance, it is likely to exist in
non-manufacturing-sector firms and not in manufacturing-sector firms.
However, it should also be noted that, based on the sub-period analysis in Panel C
of Table 5, the relationship is not constant in all periods. Therefore, if anything
can be concluded from these overall results, it is that there appears to be a
(temporally) positive, unstable, relationship between control and performance
in the non-manufacturing sector only.
115
JOG & TULPULE
TABLE 5
CUMULATIVE WEALTH AGGREGATES
PANEL A
Growth of $1 cumulative investment in aggregate portfolios of manufacturing and non-
manufacturing firms, revised annually. Growth of $1 cumulative investment in the TSE 300 Index.
COMPARABLE
EQUALLY WEIGHTED CONTROL WEIGHTED SIZE WEIGHTED VALUES FOR
YEAR MFR NON MFR NON MFR NON TSE 300
1977 1.00 1.00 1.00 1.00 1.00 1.00 1.00
1978 1.39 1.39 1.42 1.43 1.35 1.31 1.29
1979 2.28 1.61 2.28 1.59 2.20 1.53 1.88
1980 3.20 2.32 3.17 2.19 2.90 2.13 2.44
1981 2.71 2.12 2.90 1.89 2.80 1.95 2.20
1982 2.66 2.57 2.79 2.33 2.57 2.23 2.32
1983 3.46 4.21 3.72 3.83 3.50 3.68 3.14
1984 3.08 4.29 3.26 3.83 3.22 3.65 3.06
1985 3.47 6.57 3.73 5.96 3.75 5.29 3.83
1986 3.50 7.71 3.70 7.24 4.11 6.44 4.17
1987 3.83 7.57 4.00 7.44 4-41 6.21 4-41
1988 4.28 9.97 4.72 9.78 4.90 7.90 4.90
1989 4.71 10.64 5.18 10.30 5.64 8.73 5.95
1990 3.55 6.87 3.89 6.21 4.75 6.50 5.07
1991 3.70 7.76 4.14 7.11 5.15 7-11 5.68
PANEL B
Average monthly returns and their standard deviations, average monthly excess returns and their
standard deviation, Sharpe measure and JK-Z test (H
0
: Sh
PT
p = Sh-j-
SE
 300), portfolio beta.
Average Return (%) 0.97 1.36 1.01 1.34 1.23 1.33 1.17
Standard
Deviation (%) 6.07 5.38 6.12 5.94 6.12 5.66 5.07
Excess Return (%) 0.05 0.44 0.09 0.42 0.34 0.41 0.25
Standard
Deviation (%) 6.11 5.44 6.17 5.99 6.16 5.71 5.13
Sharpe Measure 0.01 0.08 0.01 0.07 0.06 0.07 0.05
JK-Z Test 0.09 1.37 -0.82 0.38 1.36 0.43 0.00
Beta 1.07 0.80 1.02 0.83 1.09 0.87 1.00
Note: ** Statistically significant at the 1% level.
116
CONTROL AND PERFORMANCE: EVIDENCE FROM THE TSE 300
TABLE 5 (CONT'D)
PANEL C
Average monthly returns and their standard deviations, average monthly excess returns and their
standard deviations, Sharpe measure and JK-Z test (H
0
: Shpjp = Sh
TSE300
), portfolio beta.
EQUALLY WEIGHTED
MFR NON
(1978-1981)
Average Return (%) 2.33
Standard Deviation (%) 6.88
Excess Return (%) 1 .27
Standard Deviation (%) 6.99
Sharpe Measure 0.18
JK-Z Test 2.25**
Beta 1.11
(1982-1986)
Average Return (%) 0.49
Standard Deviation (%) 5.74
Excess Return (%) -0.38
Standard Deviation (%) 5.78
Sharpe Measure -0.07
JK-Z Test -3.13**
Beta 1.05
(1987-1991)
Average Return (%) 0.35
Standard Deviation (%) 5.52
Excess Return (%) -0.50
Standard Deviation (%) 5.55
Sharpe Measure -0.09
JK-Z Test -1.18
Beta 1.02
1.69
5.08
0.63
5.21
0.12
-0.11
0.68
2.26
5.09
1.39
5.13
0.27
3.37**
0.76
0.21
6.00
-0.64
6.03
-0.11
-1.16
0.98
CONTROL WEIGHTED
MFR NON
2.45
6.68
1.39
6.79
0.20
2.41**
1.03
0.47
5.88
-0.39
5.91
-0.07
-2.73**
1.01
0.38
5.72
-0.46
5.75
-0.08
-0.78
1.00
1.45
5.41
0.39
5.53
0.07
-0.26
0.69
2.36
5.56
1.48
5.68
0.26
3.06**
0.80
0.24
6.44
-0.61
6.48
-0.09
-0.91
1.03
SIZE WEIGHTED
MFR NON
2.38
7.38
1.32
7.46
0.18
1.52
1.13
0.83
5.93
-0.04
5.97
-0.01
-1.71
1.08
0.49
5.58
-0.37
5.60
-0.07
-0.58
1.06
1.53
5.61
0.47
5.72
0.08
-0.93
0.76
2.14
5.55
1.27
5.58
0.23
2.74
0.85
0.35
5.70
-0.50
5.72
0.09
-0.9
1.00
COMPARABLE
VALUES FOR
TSE 300
1.83
5.92
0.77
6.03
0.13
0.00
1.00
1.18
4.69
0.31
4-70
0.07
0.00
1.00
0.64
4.74
-0.22
4.76
-0.05
30.00
1.00
Note: ** Statistically significant at the 1% level.
117
JOG & TULPULE
TABLE 6
RELATIONSHIP OF VARIOUS PORTFOLIO STRATEGIES AS SHOWN IN TABLES 7-10
PORTFOLIO WEIGHT VARIABLE 1 WEIGHT VARIABLE 2
FORMATIO N FORPORTFOLIO FORPORTFOLIO
VARIABLE RETURNS RETURNS
Control Equal Size
Manufacturing (Table 7) Manufacturing (Table 8)
Non-manufacturing (Table 9) Non-manufacturing (Table 10)
Panel B of Table 5 shows the statistical significance of the differences
between portfolio returns and the TSE 300. Panel C shows mean return, standard
deviation and the values of the Sharpe measure and the corresponding JK-Z
tests based on the monthly returns for each portfolio compared to the TSE 300
Index for each of the sub-periods. Although there is no significantly superior
performance by any of the portfolios over the TSE 300 based on the entire
period (Panel B), the performance in selected sub-periods (Panel C) is in some
cases superior (or inferior) to the TSE 300.
CONTROL-BASED PORTFOLIO RESULTS - STOCK MARKET DATA
TABLE 6 SHOWS THE TABLE REFERENCES corresponding to the procedure set out
in Table 4- Tables 7 through 10 present the information for various portfolio
formation strategies in the same format as Table 5.
Table 7 displays the results of portfolios created through annual revision
(1YR), for the companies classified as belonging to the manufacturing sector
(MFR), where portfolios are formed using control as the portfolio formation
variable (CTRL) with portfolio returns created using equally weighted returns.
All the tables thus represent results organized as per the periodicity of the
revision for the sector under consideration, followed by the portfolio forma-
tion variable and the variable used to weight the individual firm returns. Also,
each table has three panels identical to the pattern set out in Table 5. These
outline the annual wealth index values for each of the four portfolios (panel
A); the results of the JK-Z tests for the entire period (panel B); followed by the
sub-period results (panel C). Collectively, Tables 7 through 10 enable us to
make inferences about the relationship between corporate control and share-
holder performance.
Manufacturing Firms
Tables 7 and 8 show the performance results for the control-based portfolios
formed from firms belonging to the manufacturing sector. From panel A of
these tables, It can be seen that as control increases, performance initially
118
CONTROL AND PERFORMANCE: EVIDENCE FROM THE TSE 300
decreases, attains a peak, and then stays almost the same (for equally weighted)
or declines slightly (size-weighted). Thus, a strategy of investing in manufacturing
firms with larger control but with equal weights would have generated higher
wealth for the investor; average monthly returns are highest for Portfolio 3
which, in fact, outperforms the TSE 300. Panel B of the tables indicates that
Portfolio 3 outperformed all the other portfolios based on the JK-Z test over
the entire period. Unfortunately, as seen from panel C of the tables, the under-
or over-performance of the portfolios is sub-period dependent. Similar to the
aggregate results, most of the over-performance in this sector occurred during
the period from 1978 to 1981. However, Portfolio 3 was still the best
performer of those under consideration.
Non-manufacturing Firms
Tables 9 and 10 show the relationship between control and performance of
non-manufacturing firms. Here again, similar results are obtained: Portfolio 3
show superior performance but Portfolio 1 shows equally superior performance.
Thus, the relationship here is less linear than that found in the manufacturing
sector. Casual observation of these tables indicates that, as control is reduced
(moves from widely held to some), performance decreases in this sector; it then
increases as control increases, but declines or stays the same at higher levels of
control. Similar to the non-manufacturing sector, relative performance also
varies across sub-periods.
Figure 3 plots the cumulative wealth of each of the portfolios in Tables 7
through 10 for the entire period as well as for the three sub-periods. The
difference between the two sectors, based on the entire period evaluation can
also be seen in Figure 3. It is interesting to compare the shape of the lines in
Figure 3 for the four graphs. Although the individual shapes vary, casual obser-
vation indicates that the third-quartile portfolio generally outperforms the
other quartiles, and that the second and fourth quartile portfolios under-
perform the first and the third quartile in relative terms. This pattern is more
pronounced in the non-manufacturing sector than in the manufacturing
sector. These results contradict the typical relationships shown by "snap-shot"
analysis of accounting performance and ownership structure. Figure 4 shows
the shape of our overall relationship for the entire period in comparison to
those found by others (as shown in Figure 2). Although our definitions of both
ownership and performance are quite different from previous studies, our
results cast some doubt on the validity of snap-shot analysis of the relationship
between performance and control-related variables. If anything, our results
indicate that, from an investor perspective, it is better to invest in a diversified
portfolio of either widely held firms or high control firms and avoid investing
in low control or very high control firms. There is only a slight re-affirmation
of the Chen, Hexter & Hue results, where a slight decline in performance in
the "highest ownership" portfolio of firms was found.
119
JOG & TULPULE
TABLE 7
MANUFACTURING FIRM SAMPLE - EQUAL WEIGHTING
(1 YEAR, MFR, CTRL, EQL)
PANEL A
Growth of $1 cumulative investment in aggregate portfolios of manufacturing and non-manufacturing
firms, revised annually. Growth of $1 cumulative investment in the TSE 300 Index.
YEAR PORTFOLIO 1 PORTFOLIO 2 PORTFOLIO 3 PORTFOLIO 4 MANUFACTURING
1977 1.00 1.00 1.00 1.00 1.00
1978 1.35 1.42 1.45 1.38 1.39
1979 2.19 2.35 2.37 2.20 2.28
1980 3.01 3.33 3.32 2.96 3.20
1981 2.28 2.54 3.09 2.98 2.71
1982 2.17 2.25 3.05 2.92 2.66
1983 2.61 2.90 4.21 4.04 3.46
1984 2.07 2.53 3.61 3.69 3.08
1985 2.37 3.08 4.36 4.03 3.47
1986 2.30 2.60 4.96 3.96 3.50
1987 2.35 2.73 5.72 4.30 3.83
1988 2.71 3.05 6.76 5.32 4.28
1989 2.97 3.40 7.80 5.56 4.71
1990 2.61 2.51 5.94 4.01 3.55
1991 2.79 2.32 6.36 4.56 3.70
PANEL B
Average monthly returns and their standard deviations, average monthly excess returns and their
standard deviations, Sharpe measure and JK-Z test (H^: Shppp = ShjsE^o), portfolio beta.
PORTFOLIO 1 PORTFOLIO 2 PORTFOLIO 3 PORTFOLIO 4 MANUFACTURING
Average Return (%) 0.82 0.71 1.28 1.07 0.97
Standard Deviation (%) 6.46 6.34 5.86 5.61 6.07
Average Excess Return (%) -0.10 -0.21 0.36 0.15 0.05
Standard Deviation (%) 6.51 6.39 5.91 5.66 6.11
Sharpe Measure -0.01 -0.03 0.06 0.03 0.01
JK-Z Test -0.60 -1.09 1.78** 0.55 0.08
Beta 1.15 1.11 1.05 0.96 1.07
Note: ** Statistically significant at the 1 % level.
120
CONTROL AND PERFORMANCE: EVIDENCE FROM THE TSE 300
TABLE 7 (CONT'D)
PANEL C
Average monthly returns and their standard deviations, average monthly excess returns and their
standard deviations, Sharpe measure and JK-Z test (H
0
: Shpjp = Sh
TS
£3
00
), portfolio beta.
1978-1981 PORTFOLIO 1 PORTFOLIO 2
Average Return (%)
Standard Deviation (%)
Average Excess Return (%)
Standard Deviation (%)
Sharpe Measure
JK-Z Test
Beta
2.03
7.70
0.96
7.81
.0.12
-0.15
1.24
2.24
7.41
1.17
7.54
0.16
1.34
1.20
PORTFOLIO 3
2.58
6.43
1.52
6.55
0.23
4.41**
1.03
PORTFOLIO 4 MANUFACTURING
2.49
6.06
1.42
6.16
0.23
4.06**
0.96
2.33
6.97
1.27
7.07
0.18
1.95**
1.11
1982-1986
Average Return (%)
Standard Deviation (%)
Average Excess Return (%)
Standard Deviation (%)
Sharpe Measure
JK-Z Test
Beta
0.20
6.11
-0.68
6.14
-0.11
-4.00**
1.09
0.21
5.83
-0.66
5.87
-0.11
-4.37**
1.07
0.96
5.88
0.08
5.93
0.01
-1.46
1.12
0.61
5.20
-0.27
5.24
-0.05
-2.52**
0.91
0.49
5.74
-0.38
5.78
-0.07
-3.13**
1.05
1987-1991
Average Return (%)
Standard Deviation (%)
Average Excess Return (%)
Standard Deviation (%)
Sharpe Measure
JK-Z Test
Beta
0.49
5.66
-0.37
5.68
-0.06
-0.60
1.08
-0.02
5.76
-0.88
5.79
-0.15
-2.42**
1.01
0.56
5.24
-0.30
5.27
-0.06
-0.30
0.99
0.39
5.53
-0.47
5.57
-0.08
-0.95
1.00
0.34
5.66
-0.51
5.69
-0.09
-1.08
1.02
Note: ** Statistically significant at the 1 % level.
121
JOG & TULPULE
TABLE 8
MANUFACTURING FIRM SAMPLE - SIZE WEIGHTING
(1 YEAR, MFR, CTRL, SIZE)
PANEL A
Growth of $1 cumulative investment in size-weighted portfolios of manufacturing firms, revised
annually by ranking firms by control.
YEAR PORTFOLIO 1 PORTFOLIO 2 PORTFOLIO 3 PORTFOLIO 4 MANUFACTURING
1977 1.00 1.00 1.00 1.00 1.00
1978 1.33 1.42 1.42 1.24 1.35
1979 1.98 2.34 2.38 2.10 2.20
1980 2.47 3.51 3.37 2.26 2.90
1981 1.93 3.44 3.42 2.41 2.80
1982 1.94 2.10 3.67 2.55 2.57
1983 2.67 3.01 4.75 3.57 3.50
1984 2.39 2.90 4.17 3.40 3.22
1985 2.89 3.66 4.64 3.79 3.75
1986 3.11 3.99 5.22 4.10 4.11
1987 3.84 3.67 5.90 4.24 4-41
1988 4.57 4.05 6.05 4.94 4.90
1989 5.16 5.15 7.20 5.03 5.64
1990 4.51 4.52 6.40 3.57 4.75
1991 4.26 4.85 7.96 3.54 5.15
PANEL B
Average monthly returns and their standard deviations, average monthly excess returns and their
standard deviations, Sharpe measure and JK-Z test (H
0
: Shp
TF
 = Sh
TSE
 300). portfolio beta.
PORTFOLIO 1 PORTFOLIO 2 PORTFOLIO 3 PORTFOLIO 4 MANUFACTURING
Average Return (%) 1.06 1.16 1.45 0.93 1.23
Standard Deviation (%) 6.27 6.50 6.52 5.95 6.12
Average Excess Return (%) 0.14 0.24 0.53 0.01 0.34
Standard Deviation (%) 6.32 6.54 6.55 5.97 6.16
Sharpe Measure 0.02 0.04 0.08 0.00 0.06
JKZTest 0.63 0.82 2.27** 0.00 1.36
Beta 1.15 1.11 1.06 0.97 1.09
Note: ** Statistically significant at the 1 % level.
122
CONTROL AND PERFORMANCE: EVIDENCE FROM THE TSE 300
TABLE 8 ( CONT'D)
PANEL C
Average monthly portfolio returns and their standard deviations, average monthly portfolio excess
returns and their standard deviations, Sharpe measure and JK-Z Test (H
0
: Shp
TF
 = SH
TSE
 3
00
),
portfolio beta.
1978-1981 PORTFOLIO 1 PORTFOLIO 2
Average Return (%)
Standard Deviation (%)
Average Excess Return (%)
Standard Deviation (%)
Sharpe Measure
JK-Z Test
Beta
1.63
7.14
0.57
7.26
0.08
-2.12**
1.15
2.93
8.07
1.87
8.15
0.23
3.88**
1.27
PORTFOLIO 3
2.87
7.55
1.81
7.64
0.24
3.06**
1.13
PORTFOLIO 4 MANUFACTURING
2.08
6.83
1.02
6.89
0.15
0.51
0.98
2.38
7.38
1.32
7.46
0.18
1.52
1.13
1982-1986
Average Return (%)
Standard Deviation (%)
Average Excess Return (%)
Standard Deviation (%)
Sharpe Measure
JK-Z Test
Beta
0.96
5.71
0.08
5.74
0.01
-1.67**
1.11
0.43
6.12
-0.44
6.19
-0.07
-2.59**
1.00
0.89
6.23
0.02
6.26
0.00
-1.72**
1.17
1.05
5.77
0.18
5.80
0.03
-0.78
1.02
0.83
5.93
-0.04
5.97
-0.01
-1.71**
1.08
1987-1991
Average Return (%)
Standard Deviation (%)
Average Excess Return (%)
Standard Deviation (%)
Sharpe Measure
JK-Z Test
Beta
0.72
6.15
-0.14
6.17
-0.02
0.77
1.18
0.47
5.16
-0.39
5.16
-0.08
-0.97
0.99
0.88
5.81
0.03
5.84
0.00
-1.95**
1.15
-0.10
5.23
-0.96
5.26
-0.18
—3 24**
0.94
0.49
5.58
-0.37
5.60
-0.07
-0.58
1.06
Note: ** Statistically significant at the 1% level.
123
JOG & TULPULE
TABLE 9
NON-MANUFACTURING FIRM SAMPLE - EQUAL WEIGHTING
(1 YEAR, NON-MFR, CTRL, SIZE)
PANEL A
Growth of $1 cumulative investment in equally weighted portfolios of non-manufacturing firms,
revised annually by ranking firms by control.
NON-
YEAR PORTFOLIO 1 PORTFOLIO 2 PORTFOLIO 3 PORTFOLIO 4 MANUFACTURING
1977 1.00 1.00 1.00 1.00 1.00
1978 1.42 1.20 1.55 1.38 1.39
1979 1.73 1.39 1.86 1.45 1.61
1980 2.59 2.32 2.39 1.93 2.32
1981 2.76 1.95 2.17 1.60 2.12
1982 3.34 2.20 2.58 2.15 2.57
1983 5.42 3.34 4.70 3.33 4.21
1984 5.73 3.66 4.38 3.24 4.29
1985 8.39 5.92 6.89 4.80 6.57
1986 9.34 6.43 9.50 5.29 7.71
1987 8.17 6.00 10.64 5.24 7.57
1988 11.32 7.42 14-19 6.72 9.97
1989 13.29 7.51 14-07 7.45 10.64
1990 9.04 4-61 8.98 5.30 6.87
1991 9.66 5.92 9.24 6.34 7.76
PANEL B
Average monthly returns and their standard deviations, average monthly excess returns and their
standard deviations, Sharpe measure and JK-Z test (H
0
: Shp
TF
 = Sh
TSE
 3
00
), portfolio beta.
NON-
PORTFOLIO 1 PORTFOLIO 2 PORTFOLIO 3 PORTFOLIO 4 MANUFACTURING
Average Return (%) 1.49 1.21 1.50 1.25 1.36
Standard Deviation (%) 5.09 5.33 5.78 5.35 5.38
Average Excess Return (%) 0.57 0.29 0.58 0.33 0.44
Standard Deviation (%) 5.13 5.39 5.85 5.40 5.44
Sharpe Measure 0.11 0.05 0.10 0.06 0.08
JK-Z Test 1.83** 0.90 1.81** 1.08 1.37
Beta 0.65 0.71 0.88 0.88 0.80
Note: ** Statistically significant at the 1 % level.
124
CONTROL AND PERFORMANCE: EVIDENCE FROM THE TSE 300
TABLE 9 ( CONT'D)
PANEL C
Average monthly portfolio returns and their standard deviations, average monthly portfolio excess
returns and their standard deviations, Sharpe measure and JK-Z Test (H^: Shpjp = SHj3£ 300),
portfolio beta.
NON-
1978-1981 PORTFOLIO 1 PORTFOLIO 2
Average Return (%)
Standard Deviation (%)
Average Excess Return (%)
Standard Deviation (%)
Sharpe Measure
JK-Z Test
Beta
2.26
4.87
1.19
4-98
0.24
1.96**
0.60
1.52
5.03
0.46
5.15
0.09
-0.76
0.67
PORTFOLIO 3
1.76
5.13
0.70
5.27
0.13
0.11
0.69
PORTFOLIO 4 MANUFACTURING
1.13
5.41
0.07
5.53
0.01
-2.49**
0.76
1.69
5.08
0.63
5.21
0.12
-0.11
0.68
1982-1986
Average Return (%)
Standard Deviation (%)
Average Excess Return (%)
Standard Deviation (%)
Sharpe Measure
JK-Z Test
Beta
2.15
4.62
1.28
4.66
0.27
3.40**
0.68
2.13
4.93
1.25
4.96
0.25
3.55**
0.81
2.64
5.62
1.77
5.69
0.31
4.30**
0.89
2.11
4.56
1.24
4.58
0.27
3.09**
0.62
2.16
5.09
1.39
5.13
0.27
3 37**
0.76
1987-1991
Average Return (%)
Standard Deviation (%)
Average Excess Return (%)
Standard Deviation (%)
Sharpe Measure
JK-ZTest
Beta
0.21
5.53
-0.64
5.55
-0.12
-1.30
0.88
0.03
5.80
-0.82
5.85
-0.14
-1.75**
0.92
0.15
6.22
-0.70
6.26
-0.11
-1.46
1.08
0.49
5.96
-0.37
5.99
-0.06
-0.36
1.06
0.21
6.00
-0.64
6.03
-0.11
-1.16
0.98
Note: ** Statistically significant at the 1% level.
125
JOG & TULPULE
TABLE 10
NON-MANUFACTURING FIRM SAMPLE - SIZE WEIGHTING
(1 YEAR, NON-MFR, CTRL, SIZE)
PANEL A
Growth of $1 cumulative investment in size-weighted portfolios of non-manufacturing firms,
revised annually by ranking firms by control.
YEAR PORTFOLIO 1 PORTFOLIO 2 PORTFOLIO 3 PORTFOLIO 4
NON-
MANUFACTURING
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1.00
1.24
1.54
2.22
2.27
2.71
4.19
3.89
5.33
6.25
5.53
7.27
9.29
6.92
7.53
1.00
1.11
1.29
2.25
2.03
2.41
3.58
4-02
6.20
7.02
6.29
8.65
8.66
6.27
7.41
1.00
1.48
1.84
2.11
1.97
2.10
4.35
3.98
5.83
8.12
9.00
10.65
11.14
8.05
8.25
1.00
1.40
1.43
1.92
1.51
1.67
2.61
3.71
3.79
4.36
4.03
5.04
5.84
4.76
5.25
1.00
1.31
1.53
2.13
1.95
2.23
3.68
3.65
5.29
6.44
6.21
7.90
8.73
6.50
7.11
PANEL B
Average monthly returns and their standard deviations, average monthly excess returns and their
standard deviations, Sharpe measure and JK-Z test (Hp: Shpjp = Sh
T
g£ 300)' portfolio beta.
NON-
PORTFOLIO 1 PORTFOLIO 2 PORTFOLIO 3 PORTFOLIO 4 MANUFACTURING
Average Return (%) 1.34
Standard Deviation (%) 5.28
Average Excess Return (%) 0.42
Standard Deviation (%) 5.32
Sharpe Measure
JK-Z Test
Beta
0.08
0.53
0.73
1.35
5.55
0.43
5.60
0.08
0.55
0.84
1.46
6.24
0.53
6.31
0.09
0.80
1.01
1.15
5.55
0.22
5.61
0.04
-0.17
0.87
1.33
5.66
0.41
5.71
0.07
0.43
0.87
Note: ** Statistically significant at the 1 % level.
126
CONTROL AND PERFORMANCE: EVIDENCE FROM THE TSE 300
TABLE 10 (CONT'D)
PANEL C
Average monthly portfolio returns and their standard deviations, average monthly portfolio excess
returns and their standard deviations, Sharpe measure and JK-Z Test (H
0
: Shpq-p = SH
TSE 300
),
Portfolio Beta.
NON-
1978-1981 PORTFOLIO 1 PORTFOLIO!
Average Return (%)
Standard Deviation (%)
Average Excess Return (%)
Standard Deviation (%)
Sharpe Measure
JK-Z Test
Beta
1.86
5.33
0.80
5.42
0.15
0.39
0.68
1.63
5.59
0.57
5.69
0.10
-0.48
0.70
PORTFOLIO 3
1.59
5.86
0.53
5.97
0.09
-0.88
0.83
PORTFOLIO 4 MANUFACTURING
1.02
5.80
-0.04
5.93
-0.01
-3.1T*
0.83
1.53
5.61
0.47
5.72
0.08
-0.93
0.76
1982-1986
Average Return (%)
Standard Deviation (%)
Average Excess Return (%)
Standard Deviation (%)
Sharpe Measure
JK-Z Test
Beta
1.84
5.38
0.96
5.41
0.18
1.62
0.68
2.21
4.94
1.33
4.95
0.27
3.88**
0.81
2.60
6.67
1.73
6.73
0.26
3.58**
1.09
1.92
5.14
1.04
5.18
0.20
2.38**
0.80
2.14
5.55
1.27
5.58
0.23
2.74**
0.85
1987-1991
Average Return (%)
Standard Deviation (%)
Average Excess Return (%)
Standard Deviation (%)
Sharpe Measure
JK-Z Test
Beta
Note: ** Statistically significant
0.44
5.12
-0.41
5.14
-0.08
-0.65
0.83
at the 1% level
0.28
6.01
-0.58
6.05
-0.10
-1.06
1.03
0.21
5.98
-0.64
6.01
-0.11
-1.88**
1.14
0.48
5.76
-0.37
5.78
-0.06
-0.40
1.00
0.35
5.70
-0.50
5.72
-0.09
-0.93
1.00
127
JOG & TULPULE
FIGURE 3
VALUE OF $1 CUMULATIVELY REINVESTED IN PORTFOLIOS
MANUFACTURING, CONTROL, EQUALLY WEIGHTED
MANUFACTURING, CONTROL, WEIGHTED
128
CONTROL AND PERFORMANCE: EVIDENCE FROM THE TSE 300
FIGURE 3 (CONT'D)
NON-MANUFACTURING, CONTROL, EQUALLY WEIGHTED
NON-MANUFACTURING, CONTROL, WEIGHTED
129
H
FIGURE 4
RELATIONSHIP BETWEEN PERFORMANCE AND CONTROL/OWNERSHIP
Notes: Information documented by previous studies and as observed in the present study.
CHH Chen, Hexter & Hue (1990)
MSV Morck, Shleifer & Vishny (1988)
NS Neun & Santerre (1986)
W Wruck(l989)
MFR Manufacturing firms
NON Non-manufacturing firms
NON-PARAMETRIC AND THE PAIRED DIFFERENCE TESTS
IN ADDITION TO THE JK-Z TESTS, we conducted two tests based on the differences
in return across portfolios. First, the non-parametric test (test of proportions)
compares the number of times the monthly returns of portfolio i was greater
than that of portfolio j. Although not testing for economic significance, this
test does provide additional information about the nature of the relationship.
Second, the parametric test of differences simply subtracts, in a pairwise
comparison, returns for portfolio j from those of portfolio i for each month,
and determines whether these differences are statistically significant. Table 11
shows the results of these tests for the portfolios described in Tables 7 through
10. The results are mixed at best. Both non-parametric and parametric tests
indicate that, on a month-by-month basis, no portfolio consistently dominates
the adjacent portfolio. These results indicate not only that the number of
times any portfolio dominates the other is almost random, but also that the
difference in returns is random. The direction (i.e., the sign) is consistent with
the results shown in Tables 7 through 10.
130
JOG & TULPULE
CONTROL AND PERFORMANCE: EVIDENCE FROM THE TSE 300
TABLE 11
FURTHER PARAMETRIC AND NON-PARAMETRIC RESULTS
Non-parametric Results: the proportion (p) of the number of times the monthly return of R
i t
 > Rj
t
over the period January 1978 to December 1991 (N = 168), the proportion q being equal to 1 - p,
and the t-statistic for the null hypothesis H
0
: PJ > PJ.
HO: P2>P1 P3>P2 P4>P3 HO: P2 > PI P3>P2 P4>P3
MFR, CTRL, EQL MFR, CTRL, SIZE WTD
p 0.48
q 0.52
t -0.61
0.57
0.43
1.71*
0.47
0.53
-0.77
p 0.59 0.52 0.45
q 0.41 0.48 0.55
t 2.35** 0.30 -1.39
NON, CTRL, EQL NON, CTRL, SIZE WTD
p 0.46
q 0.54
t -1.08
0.50
0.50
0.00
0.49
0.51
-0.30
p 0.52
q 0.48
t 0.31
0.53
0.47
0.77
0.49
0.51
-0.15
Parametric Results: the difference between the mean returns (\i
i
 -u^) of two portfolios (Pj, and PJ),
the standard error (a) of the difference between mean portfolio returns, and the t statistic for the
hypothesis H
0
 : u
;
 = U:.
P2.P 1 P3.P2 P4.P3 P2,P 1 P3.P2 P4.P3
MFR, CTRL, EQL MFR, CTRL, SIZE WTD
Hi-Hj -0.11 0.57
a 0.69 0.66
t -0.16 0.86
-0.21 Uj-n, 0.09 0.29 -0.52
0.62 a 0.69 0.71 0.68
-0.34 t 0.13 0.41 -0.76
NON, CTRL, EQL NON, CTRL, SIZE WTD
Hi-u.j -0.28
a 0.56
t -0.49
0.29
0.60
0.48
-0.24
0.60
-0.41
HrHj
a
t
0.00
0.59
0.00
0.10
0.64
0.16
-0.31
0.64
-0.48
Note: ** Statistically significant at the 1% level.
131
TABLE 12
ACCOUNTING-BASED PERFORMANCE MEASURES
PORTFOLIO 1 - LOW CONTROL
PORTFOLIO 4 - HIGH CONTROL
MEAN VALUES
YEAR
1991
1990
PERFORMANCE
MEASURE
Asset Turnover
Gross Margin
Return on Assets
Return on Equity
Debt Equity
Capital Expenditure / Net Fixed Assets
Interest Coverage
Market Value / Book Value
Asset Turnover
Gross Margin
Return on Assets
Return on Equity
Debt Equity
Capital Expenditure / Net Fixed Assets
Interest Coverage
Market Value / Book Value
PORTFOLIO 1
0.78
0.11
-0.02
-0.17
1.15
-0.25
1.20
1.51
0.72
-0.10
0.02
-0.01
0.77
-0.29
1.82
1.29
PORTFOLIO 2
0.85
0.06
0.03
-0.19
1.61
-0.12
1.43
1.28
0.90
0.10
0.06
0.11
1.13
-0.16
2.57
1.12
PORTFOLIO 3
0.71
0.02
0.00
-0.15
0.92
-0.11
2.24
1.35
0.81
0.07
0.04
0.04
0.89
-0.15
1.08
1.14
PORTFOLIO 4
0.84
0.04
0.03
0.00
1.39
-0.14
6.11
1.60
0.91
0.09
0.06
0.08
1.75
-0.15
7.21
1.35
PORTFOLIO 1 - LOW CONTROL
PORTFOLIO 4 HIGH CONTROL
MEDIAN VALUES
PORTFOLIO 1
0.69
0.04
0.02
-0.01
0.72
0.17
1.14
1.45
0.77
0.10
0.06
0.10
0.60
0.20
2.32
1.30
PORTFOLIO 2
0.53
0.03
0.04
0.01
0.69
0.11
1.18
1.24
0.69
0.09
0.07
0.14
0.56
0.15
2.39
1.16
PORTFOLIO 3
0.60
0.06
0.02
-0.01
0.68
0.09
1.18
1.17
0.65
0.09
0.05
0.08
0.40
0.15
1.78
1.06
PORTFOLIO 4
0.71
0.02
0.02
0.00
0.38
0.12
0.91
1.15
0.71
0.08
0.04
0.04
0.41
0.14
1.12
1.22
TABLE 12 (CONT'D)
PORTFOLIO 1 - LOW CONTROL
PORTFOLIO 4 - HIGH CONTROL
MEAN VALUES
PERFORMANCE
YEAR MEASURE
1989 Asset Turnover
Gross Margin
Return on Assets
Return on Equity
Debt Equity
Capital Expenditure / Net Fixed Assets
Interest Coverage
Market Value / Book Value
1988 Asset Turnover
Gross Margin
Return on Assets
Return on Equity
Debt Equity
Capital Expenditure / Net Fixed Assets
Interest Coverage
Market Value / Book Value
PORTFOLIO l
0.70
0.15
0.08
0.18
0.60
-0.17
5.00
1.54
0.95
0.13
0.10
0.35
0.77
-0.20
6.14
1.76
PORTFOLIO 2
0.87
0.15
0.08
0.22
1.00
-0.16
7.90
1.42
0.75
0.16
0.10
0.27
0.83
-0.21
15.48
1.51
PORTFOLIO 3
0.89
0.12
0.08
0.18
0.86
-0.21
6.34
1.39
0.92
0.05
0.08
0.19
0.64
-0.18
8.31
1.54
PORTFOLIO 4
0.85
0.17
0.10
0.27
1.64
-0.19
10.84
1.54
0.94
0.20
0.13
0.36
2.02
-0.17
12.85
1.76
PORTFOLIO 1 - LOW CONTROL
PORTFOLIO 4 HIGH CONTROL
MEDIAN VALUES
PORTFOLIO 1
0.64
0.15
0.07
0.17
0.42
0.17
2.94
1.40
0.88
0.09
0.09
0.24
0.43
0.17
3.64
1.59
PORTFOLIO 2
0.64
0.11
0.07
0.18
0.73
0.15
3.79
1.25
0.65
0.13
0.07
0.19
0.74
0.20
4.32
1.40
PORTFOLIO 3
0.78
0.11
0.08
0.24
0.41
0.16
4.47
1.24
0.79
0.14
0.09
0.22
0.44
0.16
6.17
1.33
PORTFOLIO 4
0.65
0.14
0.08
0.20
0.36
0.18
3.82
1.54
0.74
0.17
0.11
0.30
0.29
0.19
8.87
1.62
JOG & TULPULE
TABLE 13
RANKING FOR ACCOUNTING-BASED PERFORMANCE MEASURES :
PORTFOLIOS 1 (LOW CONTROL) AND 4 (HIGH CONTROL)
PERFORMANCE MEAN RANK
YEAR MEASURE PORTFOLIO 1
1991 Asset Turnover
Gross Margin
Return on Assets
Return on Equity
Debt Equity
Capital Expenditure / Net Fixed Assets
Interest Coverage
Market Value / Book Value
1990 Asset Turnover
Gross Margin
Return on Assets
Return on Equity
Debt Equity
Capital Expenditure / Net Fixed Assets
Interest Coverage
Market Value / Book Value
1989 Asset Turnover
Gross Margin
Return on Assets
Return on Equity
Debt Equity
Capital Expenditure / Net Fixed Assets
Interest Coverage
Market Value / Book Value
1988 Asset Turnover
Gross Margin
Return on Assets
Return on Equity
Debt Equity
Capital Expenditure / Net Fixed Assets
Interest Coverage
Market Value / Book Value
27.48
28.89
28.00
28.26
23.91
31.46
24.74
27.59
27.39
25.26
26.13
26.96
24.65
31.86
24.68
26.61
22.72
24.56
24.72
26.33
21.78
22.76
25.47
23.89
18.75
23.44
22.38
22
15.91
21.67
21.07
21.06
MEAN RANK
PORTFOLIO 4
28.5
27.14
28.00
27.75
31.95
23.82
26.15
28.39
25.79
27.48
26.79
26.14
27.97
21.55
25.26
26.41
24
22.82
22.71
21.68
24-61
23.14
20.63
23.25
20.87
17.61
18.35
18.61
22.85
18.09
17.59
19.26
MANN-WHITNEY Z
-0.23
-0.40
0.00
-0.11
-1.86
-1.78
-0.34
-0.18
-0.37
-0.52
-0.15
-0.19
-0.78
-2.45
-0.14
-0.04
-0.31
-0.46
-0.49
-1.14
-0.69
-0.09
-1.21
-0.15
-0.57
-1.57
-1.08
-0.91
-1.87
-0.97
-0.95
-0.48
K-SZ
0.72
0.58
0.51
0.65
1.55
0.92
0.62
0.75
0.56
0.61
0.84
0.81
0.69
1.55
0.78
0.76
0.57
0.81
0.7
0.76
0.56
0.38
0.71
0.59
0.56
0.11
0.27
0.36
0.06
0.33
0.33
0.62
134
CONTROL AND PERFORMANCE: EVIDENCE FROM THE TSE 300
ACCOUNTING-BASED PERFORMANCE RESULTS IN THE
MANUFACTURING SECTOR
THIS SECTION HAS TWO PURPOSES: first, to supplement the results obtained
from the stock market analysis and, second, to provide a basis for comparison
with other accounting-based studies." As noted earlier, eight commonly used
accounting measures are used to perform comparisons between firms with
differential degrees of control. Table 12 shows the corresponding values of
means and medians for each of these eight measures for the four quartiles for
the years 1988 through 1991.
Several observations are in order here. First, there is a noticeable difference
between means and medians for each of the ratios; therefore, one must be
careful about drawing strong conclusions based on these results. Second,
although the mean values signify that these values are uniformly higher for
Portfolio 1 compared to Portfolio 4, no consistently increasing or decreasing
patterns can be seen for firms belonging to Portfolio 2 or Portfolio 3. For the
firms of Portfolio 1 and 4, moreover, the median values lead us to conclusions
opposite to those indicated by means. Based on median value results, there is
really no pattern which can be considered as consistent across each of the four
years and across the portfolios. These casual observations are also confirmed
through further statistical analysis. Table 13 shows the mean ranks for each of
the measures and for each of the years for the two extreme portfolios (1 and
4), the values of the Mann-Whitney Z for detecting the differences in ranking
and the Kolmogorov-Smirnov Z for detecting differences in distribution of
each. The only performance measure showing a statistically different
difference was capital expenditure-to-net fixed asset ratio and that only for
the year 1990. Thus, these results are consistent with the results based on
stock market performance as shown earlier. The absence of any statistically
significant differences between firms based on the market-to-book-value ratio
also indicates that similar conclusions could have been drawn using the
Tobin's Q values. Thus, we can conclude that, based on these accounting
measures, there is no consistent relationship between firm performance and
the associated degree of control.
SUMMARY, LIMITATIONS AND CONCLUSIONS
T
HE MAIN PURPOSE OF THIS STUDY is to provide a Canadian empirical
perspective on the issue of control and performance, using both stock-
market-based and accounting-based performance measures.
A valid criticism of our use of the stock market performance measure and
our conclusions based on the results generated by it would be that stock market
returns may not be an accurate representation of performance. The argument is
that the differences in the firm-specific environment in corporate governance
may have already been capitalized in the stock prices at the beginning of our
time period. After this capitalization, shareholders earn the normal returns
135
JOG & TULPULE
affected only by new information or surprises. Only if the rate of arrival of
information and surprises affects one group of firms (widely held) differently
from the other (closely held), would we be able to detect differences in stock-
market-based performance measures. And even if the difference were found, it
could not be attributed to the issue of control and performance.
12
There are two reasons why this criticism may not be entirely appropriate.
First, in addition to analyzing stock market performance, we also analyzed
accounting-based performance measures (for a subset of firms), including a
measure which closely resembles the Tobin's Q ratio. These results were in line
with the stock-market-based results for the period studied. Second, it is
difficult to believe that during such a long period, the stock market would
show no preference to the corporate governance structure and not adjust
expected returns accordingly, if the governance structure were deemed important
for differential performance. Our results are consistent with the notion that
alternative mechanisms are in place to mitigate agency problems. The
differences in control are of less importance due to the efficacy of disclosure
rules and other mechanisms. As a result, differences in ownership need not be
related to differential performance; individual differences in operating
procedures and other factors may of more importance.
This analysis allows us to draw some reasonably robust conclusions
based on the overall results and associated statistical tests. First, it is clear
that the relationship between control and stock market performance is sector-
specific, with the non-manufacturing sector being more sensitive to the
effects of control than the manufacturing sector. In the manufacturing sector, an
investor would have been better off simply investing in large Canadian
manufacturing firms — which did better than the overall manufacturing sector
portfolio. There was little, if any, use in investing in securities, based on the
degree of control of a firm within that sector or within a specific size group.
However, in the non-manufacturing sector, investment based on control does
have performance implications. Second, the relationship between control and
stock market performance is inconsistent with that shown in previous snap-shot-
type U.S.-based research. For example, the third quartile firms outperform other
firms in the non-manufacturing sector.
13
 Third, accounting-based measures fail
to detect any differences in firm performance based on the associated degree
of control, either for a given year or over the entire time period. Overall,
from a statistical perspective, none of our results shows any consistent differences
between widely held firms and closely held firms belonging to the TSE 300.
There are several limitations to our research design which clearly affect
both the comparability of our results with those of previous studies, and our
conclusions. These include issues about the definition of control, the choice of
our sample (since it is restricted to only TSE 300 companies), the empirical
methodology and its robustness, our ad hoc method of classifying stocks into
two sector groups, the nature of institutional ownership in Canada, our
inattention to the identity of the controlling owner (foreign, versus founder
J36
CONTROL AND PERFORMANCE: EVIDENCE FROM THE TSE 300
versus corporate versus family) and our somewhat arbitrary partition of the
sample into quartiles.
It is possible that a more powerful research methodology - using more
precise definitions of control, or enlarging the sample to non-TSE 300 firms —
may lead to different conclusions. It is also possible that a better definition of
control or ownership, or a "quantified" measure of corporate governance
(based, for example, on the ratio of independent to insider directors on the
Board) may be required to assess the effect of governance on firm performance.
It is also possible that the governance structure (through the Board of
Directors) is such that the ownership characteristic of the firm is irrelevant.
One option we are now considering is to evaluate the performance of firms
that have either a dual-class voting structure or only the restricted voting class
of shares in the public hands.
14
Notwithstanding these limitations, our results indicate that much more
longitudinal research is necessary before any definitive conclusions about the
general relationship between control, corporate governance and firm perfor-
mance can be drawn. Moreover, any categorical statement made about the
relationship between control and performance, based on the conclusions of
existing empirical research, must be viewed with great caution.
These results may also suggest that the corporate governance environ-
ment in Canada has performed remarkably well; using control as the only
distinguishing feature across firms does not seem to matter for firm perfor-
mance. Given an environment which can be characterized by better monitoring,
increased (but cost-effective) disclosure rules for publicly traded companies, a
vigilant institutional and investment research presence, a well-functioning
board of directors for each corporation, and a policy framework which ensures
that timely and relevant information is available to investors, then the actual
degree of control of a publicly traded firm by inside owner/managers may cease
to be a matter of public debate. From a policy perspective, the focus should
shift away from a concern over the degree of ownership of publicly traded
corporations and toward ensuring that the overall corporate governance
environment provides the necessary mechanisms for inside owner/manager
discipline.
ENDNOTES
1 Costs paid by the agent in order to guarantee the principal that the agent will
not take certain actions which would be detrimental to the principal's interests.
2 The dollar equivalent of the reduction in welfare experienced by the principal
due to the divergence between the agent's decision and those decisions
which would maximize the principal's interests.
3 As Jensen & Ruback (1983) point out, ownership of shares is not the only
way to align manager's interests with those of shareholders. A less than
137
JOG &. TULPULE
adequate performance by a management team results either in its replace-
ment by a better performing team, or in a takeover bid which will eventually
replace the management. This notion of external discipline induces
managers always to act in the best interest of the owners.
4 Vance found that net income was positively related to the directors' share-
holdings (of companies listed on the NYSE and AMEX). Similarly, Pfeffer
reports a positive relationship between managerial shareholding and the
Return on Equity as well as the Net Profit margin. Using the ANOVA
approach, Kim & Lyn (1988) investigate the relationship between a firm's
insider ownership and its Tobin's Q. They find that Q increases with insider
ownership.
5 The Morck & Stangeland study compares the profitability performance
differences between U.S. and Canadian firms, as well as differences
between the type of control (heir-controlled versus professional management-
controlled). They do not investigate the performance of Canadian firms
from a shareholder viewpoint. They classify firms into two groupings:
closely held (defined as those firms with 20 percent or more shares held by
a dominant shareholder) and widely held (all other firms).
6 There is no real justification for such an ad-hoc grouping, except for the
conjecture that outside investors may be in a better position to value the
hard assets found in a manufacturing firm than the soft assets (which leave
the premises at 5 p.m.) of a non-manufacturing firm. Thus, the degree of
asymmetric information may be different in these two groups, and there-
fore a different relationship between control and performance might be
expected.
7 In an earlier version of this study, we also used a control-based weighting
within a size-based weighting scheme. Since the main focus of the current
study is on control and ownership-related issues and since the results from
that complex weighting scheme do not result in any new insights, we do
not report these findings here in the interest of brevity.
8 See Kryzanowski & Zhang (1992). The Sharpe measure is given by
<|)j = Jij/Oj where jtij is the mean excess return on portfolio i, and GJ is
the estimated standard error of the excess return of portfolio i. The
Jobson-Korkie test statistic for the Sharpe measure is given by
Zj
v
 = Aj
v
 / V0j
v
 where Aj
v
 is the sample estimate of the transformed
difference between the Sharpe measures of portfolios i and v, and 0
iv
 is the
estimate of the variance of A
iv
.
9 The diversity of firms within our non-manufacturing sample makes it
impossible to make valid comparisons across the sample. Hence, we report
only the results in the manufacturing sector. It should also be noted that
our "accounting" sample contains a slightly lower number of companies
than our "stock-market" sample, due mainly to unavailability of data.
10 Note that the TSE 300 also includes sectors which are classified as neither
manufacturing nor non-manufacturing; namely banks and utilities.
138
CONTROL AND PERFORMANCE: EVIDENC E FRO M THE TSE 300
11 It could be argued that stock-market based measures account mainly for
differences, rather than level of value. If this is so, then the accounting-
based performance measures may be able to detect any systematic differences
in levels, based on measures like the market-to-book-value ratio.
12 We thank Randall Morck for encouraging us to think about this issue. We
have made liberal use of his very helpful suggestions.
13 This conclusion must be viewed with caution. In most of the U.S. studies,
the highly controlled firm may have only a 30 percent insider ownership;
this is not the case in Canada. Thus, our results cover a much wider span
of control than the U.S. studies do.
14 There are at least 40 firms with dual-class ownership structure, and a
further 100 firms which have only restricted voting shares traded on the
Toronto Stock Exchange. At present we are in the process of analyzing
their performance in relation to their peers in the industry.
ACKNOWLEDGEMENTS
W
E WOULD LIKE TO THANK A. Srivastava, W. Lawson, P. Halpern and
R. Morck for valuable comments and both the Social Sciences and
Humanities Research Council and the Financial Research Foundation of
Canada for providing financial support. This study is partially based on
Tulpule's thesis for a Masters degree in Management Studies at the School of
Business, Carleton University.
BIBLIOGRAPHY
Berle, A. and G. Means. The Modem Corporation and Private Property. New York:
Macmillan, 1932.
Chen, H., L. Hexter and M. Hue. "Ownership Structure and Firm Performance: Some New
Evidence." Second Northern Finance Association Meeting, Montreal, Canada, 1990.
Demsetz, H. "The Structure of Ownership and the Theory of the Firm." journal of Law and
Economics, 26 (1983):375-90.
Demsetz, H. and K. Lehn. "The Structure of Corporate Ownership: Causes and
Consequences. "Journal of Political Economy, 93 (1985):! 155-77.
Fama, E. "Agency Problems and the Theory of the Firm." Journal of Political Economy, 88
(1980):288-307.
Hart, O. "The Market Mechanism as an Incentive Scheme." The Bell Journal of Economics,
14(1983):366-82.
139
JOG & TULPULE
Holderness, C. G. and D. P. Sheehan. "Monitoring an Owner: The Case of Turner
Broadcasting." Journal of Financial Economics, 30 (1991):325-46.
Jensen, M. and W. Meckling. "Theory of the Firm : Managerial Behaviour, Agency Costs
and Ownership Structure." Journal of Financial Economics, 3 (1976):305-60.
Jensen, M. and R. Ruback. "The Market for Corporate Control." Journal of Financial
Economics, 11 (1983):5-50.
Jensen, M. and J. Warner. "The Distribution of Power Among Corporate Managers,
Shareholders, and Directors." Journal of Financial Economics, 20 (1988):3-24.
Jobson, J. D. and B. M. Korkie. "Performance Hypothesis Testing with the Sharpe &.
Treynor Measure." Journal of Finance. (1986):889-908.
Jog, V. and H. Schaller. "Concentrated Ownership and Market Value of the Firm."
Working Paper - Carleton University, 1989.
Kim, W. and E. Lyn. "Excess Market Value, Market Power, and Inside Ownership
Structure." Review of Industrial Organization, 3 (1988):l-25.
Kryzanowski, L. and H. Zhang. "The Contrarian Investment Strategy does not Work in
Canadian Markets." Journal of Financial and Quantitative Analysis, 27 (1992):383-96.
McConnell, J. and H. Servaes. "Additional Evidence on Equity Ownership and Corporate
Value." Journal of Financial Economics, 27, (1990):595-610.
Morck, R., A. Shleifer and R. Vishny. "Management Ownership and Market Valuation :
An Empirical Analysis." Journal of Financial Economics, 20 (1988):293-315.
Morck, R. and D. Stangeland. "Corporate Performance and Large Shareholders." Paper pre-
sented at the sixth Northern Finance Association Conference, Vancouver, September
1994.
Muellar, D. Profits in the Long Run. Cambridge: Cambridge University Press, 1986.
Neun, S. and R. Santerre. "Dominant Stock ownership and Profitability." Managerial and
Decision Economics, 7, (1986):207-10.
Oswald, S. and J. Jahera. "The Influence of Ownership on Performance: An Empirical
Study." Strategic Management Journal, 12 (1991):321-26.
Pfeffer, J. "Size and Composition of Corporate Boards of Directors: The Organization and
its Environment." Administrative Science Quarterly, 17 (1972):218-28.
Shleifer, A. and R. Vishny. "Large Shareholders and Corporate Control." Journal of Political
Economy, 94 (1986):461 -88.
Stiglitz, J. "Credit Markets and the Control of Capital." Journal of Money, Credit and
Banking, 17*2 (1985):133-52.
Stulz R. "Managerial Control of Voting Rights: Financing Policies and the Market for
Corporate Control." Journal of Financial Economics, 20, (1988):25-54.
Vance S.C. Board of Directors: Structure and Performance. Portland: University of Oregon
Press, 1964.
Wruck K. "Equity Ownership Concentration and Private Value: Evidence from Private
Equity Financing." Journal of Financial Economics, 23, (1989):3-28.
140
DISCUSSANTS' COMMENTS ON PART II
David A. Stangeland
Faculty of Management
University of Manitoba
Commentary on Part II
Governance Structure, Corporate Decision-Making
and Firm Performance in North America
T
HE STUDY BY RAO AND LEE-SlNG is essentially a broad compilation of
several measures of corporate governance structure, corporate decision-
making, and corporate performance. It is important in two ways. First, it identifies
many of the characteristics that are basic to Canadian firms and contrasts
them with those of U.S. firms. Second, it represents a first pass at assessing
some of the relationships between these characteristics and thus provides a
valuable background for the studies that follow.
Of the characteristics identified for Canadian firms that are different
from those for American firms, three stand out. Two of them relate to general
firm attributes: size and industry. The third is a difference in governance
structure - concentration of ownership.
In terms of size, American firms are generally larger than Canadian
firms. Both absolute values and relative proportions show there are more U.S.
firms than Canadian firms in the largest size category. More firms (by proportion)
fall into the smallest size category in Canada than in the United States. Total
sales of U.S. companies amount to nearly ten times the total sales of Canadian
companies during the time period studied. All of these results are consistent
with an economy that is roughly ten times the size of the Canadian economy.
The distribution of firms across industries is also different when comparing
Canada to the United States. Both countries have similar proportions of firms
in wholesale trade, transportation, labour-intensive manufacturing, construction,
agriculture, forestry, fishing, and finance. There are large differences, however,
in mining, service industries, and technology intensive manufacturing.
Relative to their U.S. counterparts, there are few Canadian firms in service
industries and high-technology manufacturing, but many in basic mining.
Probably the most striking difference between the characteristics of
Canadian and American firms is the level of ownership concentration. More
than 75 percent of the Canadian sample have a dominant shareholder (or a
small group of shareholders) who control(s) 20 percent or more of the firm's
stock. The comparable figure in the United States is less than 60 percent.
Over 50 percent of the firms in the Canadian sample have a majority share-
holder compared to less than 25 percent of the American sample. Differences
in ownership structures between Canadian and American firms are also more
pronounced for the largest-size category of firms.
141
STANGELAND/BARONE-ADESI
What makes these differences so important is that if they are ignored,
then spurious relationships between other variables may be observed and,
perhaps, interpreted inappropriately as causal. For example, Rao and Lee-Sing
show that the concentration of ownership variable is quite different depending
on the industry and size class to which a firm belongs. They also show signifi-
cant differences in industry distribution and size of Canadian firms relative to
U.S. firms. With respect to the latter, if these facts are ignored there is a risk of
attributing operating decision-making and performance effects to differences
in ownership structure whereas they are attributable to nothing more than size
and industry characteristics. Using differences in ownership structure to
explain differences in profitability between Canadian and U.S. firms is one
example of this problem. If industry effects are ignored, then Canadian firms
in more cyclical industries may appear to have performance differences related
to their different ownership structures even though there is nothing more than
a profitability-industry relationship. A similar problem occurs with research
and development. Since there are fewer Canadian firms in high-technology
industries, Canadian firms spend less on R&D. By ignoring the industry differ-
ences, however, one may conclude that it is a different corporate governance
system that determines the difference in R&D spending.
The implication of spurious relationships is serious for policy makers. If
policy is based on false relationships, then the effect of that policy is not likely
to be what policy makers had intended. For instance, continuing with the
example of research and development, if Canadian corporate governance is
seen to be responsible for unacceptably low R&D levels, it may be argued that
a policy of R&D tax breaks can correct the problem. If, however, the low R&D
level derives from Canadian firms being in non-R&D-intensive industries,
such a policy may actually precipitate a two-fold problem. On the one hand it
will reduce the tax revenue of the government; on the other, it will promote
over-investment in R&D projects whose costs are unlikely to be recovered
(that is, negative net present value research and development). To conclude
the example, such R&D tax breaks should only be targeted at those industries
where R&D is truly too low. If U.S. firms are to be used as a benchmark, then
a comparison of same-industry, same-size companies across the two countries
must be conducted to determine which particular Canadian industries have
this problem.
In conducting their analysis of the relationships between corporate gov-
ernance and corporate operations, Rao and Lee-Sing do control for size and
industry and they find these controls to be quite important. The relationships
between governance and performance are less clear. Some of the governance
variables show significant correlation with the performance variables for
American firms, but the results are weaker for Canadian firms. An interesting
extension to their work might be to match firms by industry and size and then
test whether the Canadian governance variables are important in explaining the
differences. In a recent working paper ("Large Shareholders and Corporate
142
DISCUSSANTS' COMMENTS ON PART II
Performance in Canada", March 23, 1995), Morck & Stangeland do such an
analysis (using industry matches) on a different set of governance variables with
some success. In their study they also find firm age to be an important control.
To conclude, Rao and Lee-Sing provide valuable background information
on the characteristics of Canadian firms and how they differ from U.S. firms.
These differences must be addressed and taken into account in future studies
that attempt to relate governance to aspects of corporate operations. Finally,
this study provides a first pass at analyzing some of these relationships and can
be used as a benchmark by other researchers.
Commentary on Part II
Control and Performance: Evidence from the TSE 300
I
N THEIR STUDY, JOG AND TULPULE evaluate corporate control and industry
type for the period from 1978 to 1991. They analyze how differences in these
characteristics are related to stock returns and accounting-based measures of
performance.
The relationship between their control variable and performance is of
main concern. Control is defined as the percentage of outstanding equity of a
firm that is tradable on the stock exchange. This measure gives an indication
of ownership concentration. The more stock that is tradable, the less concen-
trated is ownership; the less stock that is tradable, the more concentrated is
ownership. Ownership concentration is thought to have an important influence
on management-shareholder agency conflicts. With low ownership concen-
tration (such as in a very widely held firm) management owns few or no shares
and thus has little incentive to act in shareholders' interests. As ownership
concentration increases, either management owns more of a significant block
of shares or some other party does. In either case, management is more likely
to act more in the shareholders' interests because of its personal shareholdings
or because of monitoring by a powerful blockholder. The positive effects of
increased ownership concentration can be offset if a high ownership concen-
tration enables management to become entrenched.
Jog and Tulpule's justification for including an industry-type characteristic
is that some industries are easier to evaluate and therefore problems of
asymmetric information are less severe. They break down their sample into
manufacturing and non-manufacturing firms. They propose that manufacturing
firms have more "hard" assets in place, and thus investors will have an easier
time valuing them than the "soft" assets (such as human capital) of non-
manufacturing firms. Because of fewer information asymmetry problems with
manufacturing firms, there is less chance for management to get away with
suboptimal behaviour.
143
STANGELAND/BARONE-ADES1
What are the empirical findings of Jog and Tulpule's study? First, no
significant relationships are observed between the accounting-based measures
and the different firm characteristics. They do, however, present some interesting
observations on stock returns. They find higher stock returns for moderately
high levels of ownership concentration, a stronger effect for non-manufacturing
firms, and temporally unstable relationships.
Jog and Tulpule conclude that their results for Canadian firms are
inconsistent with previous research that does a "snapshot" (one time period)
analysis. In addition, they state that a one-time-period analysis must be viewed
with caution as their evidence shows the relationships to be unstable over time.
The conclusion of potentially unstable relationships over time appears to
be justified. As agents in the economy learn about the cause and effect nature
of relationships, they may develop various types of contracts to improve efficiency.
The nature of those new contracts may not be evident in summary measures
like ownership concentration.
It may be premature to discount previous studies based on this study's
analysis. First, Jog and Tulpule use different measures of both ownership and
performance. Second, the interpretations to be drawn from their abnormal-
return performance variable are unclear. Jog and Tulpule assert that stock
market returns reflect long-term performance from an investor's viewpoint.
This is true. However, if markets are somewhat efficient, then the effects of
ownership concentration are already capitalized in the stock price and return.
Only unexpected changes in ownership concentration should cause abnormal
returns. If a particular constant ownership structure resulted in long-run
abnormal returns, this would be powerful evidence that the market is not
semi-strong form efficient. Jog and Tulpule do not address the efficiency issue
and they do not examine changes in ownership. At best, the abnormal-return
evidence for their different portfolios can be interpreted as evidence that some
portfolios had more positive surprises while others had more negative surprises
over the time period studied. Given that surprises are unexpected (otherwise
they are not surprises) a recurring pattern would not be expected. This is consis-
tent with the temporal instability of Jog and Tulpule's result. To restate the
point, this study's return evidence does not refute evidence from other studies
because it does not test the relationship between control and performance. To
test whether such a relationship exists, Jog and Tulpule must examine changes
in control and determine how those changes affect the returns.
To conclude, Jog and Tulpule state that they find no consistent differences
between widely held and closely held firms belonging to the TSE 300. Much of
the research on the relationships between control and performance uses a single-
period cross-sectional methodology. Jog and Tulpule highlight the fact that
these relationships may indeed change over time. This is an important consid-
eration for policy makers who are often required to make decisions based on
dated studies.
H4
DISCUSSANTS' COMMENTS ON PART II
Giovanni Barone-Adesi
The Wharton School
University of Pennsylvania
Commentary on Part II
T
HE STUDIES IN THIS SESSION HIGHLIGHT the differences between Canadian
and American corporations along the lines commonly associated with
corporate governance and performance. Most current academic research on
the subject attempts to explain empirical regularities observed in the United
States. It is not surprising, therefore, that the authors generally find little
support for current theories in the Canadian markets. Differences in the
regulatory and legal framework in the two countries contribute much to the
explanation of these results.
On the governance side, Canadian firms are more frequently controlled
by one large shareholder than American firms. However, using a simple
comparison of the percentage holdings of Canadian and American large share-
holders to measure the degree of control exercised by Canadian executives
may be too restrictive. In fact, Canadian executives are much more
entrenched than their American counterparts for several reasons: the absence
of class actions by shareholders in the Canadian legal system; the consequent
difficulty of dismantling poison pills and other takeover defences in Canada;
and the coattails provisions in Canadian corporate acquisitions that tend to
benefit small shareholders but that also make the transfer of corporate control
more expensive. The widespread use of multiple classes of shares in Canada
adds further to the entrenchment of Canadian executives. Advocates of share-
holders' rights are occasionally vocal even in Canada, but they do not appear
to have a widespread influence on corporate decision-making.
Moving to the measurement of performance, it is necessary to distinguish
between accounting-based and market-based measures of corporate perfor-
mance. Accounting-based measures of performance are intended to keep track
of invested capital. If accounting were an exact science, accounting measures
would be ideally suited to measure managers' effectiveness in choosing profitable
projects. Unfortunately, the number of conventions necessary to implement
an ideal accounting system would still leave us with perplexities about
accounting measures - although accounting ambiguities do tend to be resolved
over extended time periods.
Performance measures based on stock prices are easy to compute and are
also reliable. Unfortunately, they cannot be used to measure managers' perfor-
mance in absolute terms; they can be used only in terms relative to market
expectations. The expected performance of a very good manager will be built
145
STANGELAND/BARONE-ADES1
into his/her firm's current stock price and only unexpected deviations will be
picked up by the future stock performance.
In practice, stock-based performance measures are ideally suited to reveal
the value of new information (such as takeover offers), and to shed some light
on related issues, such as the value of corporate control. However, stock-based
measures do not allow us to determine unequivocally whether a firm is over-
investing or underinvesting, whether managers are shirking, or whether
changes in strategy are called for.
To answer all of these questions, accounting measures are often the only
meaningful ones. The choice of performance measures is therefore tied to the
question being asked. However, in order to ask meaningful questions, we must
be willing to move beyond corporate statistics and be prepared to focus on the
different legal and regulatory frameworks that shape corporate decisions. In
this connection, multilateral comparisons across countries would be useful,
although different accounting and legal standards require caution in the interpre-
tation of results. In an accounting context, the secret reserves allowed under
German accounting rules is a typical example. The special responsibility given
to the controlling shareholder in France, in recognition of his/her influence is
an example in a legal context.
Before all the difficulties and ambiguities of performance evaluation
discourage us from further research, I would like to mention that the ongoing
process of integration in international capital markets may make most of these
worries obsolete. The fact is that, in a competitive international market,
Canadian firms will have to demonstrate good performance in order to attract
investors. This simple reality should provide Canadian managers with the
necessary incentives to perform — otherwise Canadian investors will buy foreign
securities.
At present, there are still some barriers to capital market integration,
mostly in the insurance and pension industries. However, the continuing
relaxation of these barriers through amended regulation and the extensive use
of derivative securities in portfolio management, is eliminating the shelters
that managers in many countries have hitherto enjoyed; everyone is now
being forced to play by similar rules. This trend will not last forever but, for as
long as it lasts, it will provide effective performance monitoring and ensure the
competitiveness of Canadian industry.
ACKNOWLEDGEMENTS
I
WISH TO EXPRESS MY THANKS to the University of Alberta and the Wharton
School, University of Pennsylvania.
146
Part III The Board and Beyond
This page intentionally left blank 
Jean-Marie Gagnon
Faculty of Administrative Studies
Universite Laval
Jose'e St- Pierre A
Department of Business and Economics \
Universite du Quebec a Trots-Rivieres
Alternative Mechanisms for Corporate
Governance and Board Composition
T
HE GROWING BODY OF LITERATURE on corporate governance now includes a
number of studies that take into account the role of the board of directors.
The main themes in these studies focus on the relationship between board
composition and corporate performance or behaviour and the board seen as a
complement or substitute for other mechanisms. In this study, we examine
empirically the determinants of board composition, drawing on a sample of
258 Canadian firms partitioned by type of ownership. Our results suggest that
board composition is a function of the distribution of voting rights. However,
the same statistical model does not apply to all three types of ownership.
Under diffuse ownership, the ratio of outside to inside members of the board
increases with the stock holdings of the important outside shareholders, and
decreases with the holdings of the inside directors. These results are consistent
with monitoring and entrenchment behaviour, respectively. We find no evidence
of substitution between board composition and financial variables or regulation.
Using a system of simultaneous equations, we explore the association between
board composition and overall performance, but we do not detect a statistically
significant relationship.
The organization of this study is that we first compile an inventory of the
various mechanisms that are available and examine the process for selecting a
subset of those mechanisms adapted to the peculiar circumstances of the firm
(taking into account their interactions). We then discuss a number of
behavioural hypotheses, and examine board composition. Presentation of the
data is then followed by tests of models of board composition and monitoring
and of board composition and performance. Following a discussion of the
implications for public policy, the study ends with a brief summary.
MECHANISMS FOR CORPORATE GOVERNANCE
MECHANISMS
THE MECHANISMS CONSIDERED IN THIS STUDY are: board composition, executive
compensation, debt policy, dividend policy and takeover defences (poison pills
149
GAGNON & ST-PIERRE
and charter, statutory and litigious defences). These appear as column headings
in Table 1. We examine the mechanisms themselves, their cost and the proba-
bility of their use.
Board Composition
Corporate governance deals with "the distribution of power among corporate
managers, shareholders and directors".
1
 Board composition, which is discussed at
length below, refers to the distribution of members according to their primary
allegiance, which may be either to the shareholders or to the managers. The for-
mer are called "outsiders" (or outside directors) and the latter, "insiders".
Ultimately, all directors are responsible to the shareholders. However, one would
not expect an inside director (employee) with no stock ownership, for example,
always to favour the same business strategies as an outside director who is also an
important shareholder. The optimal composition of the board is likely to depend
on specific circumstances within the firm, including the other mechanisms for
corporate governance.
Executive Compensation
Executive compensation is one type of contract that can be designed to align
efficiently the interests of managers and shareholders. This alignment is achieved
by relating the value of the compensation package to the market value of the
shares, either directly or indirectly, through earnings. A manager's remuneration
has two parts - variable and fixed. The variable component comprises all types of
compensation tied to earnings or to the market value of the shares (such as stock
options). The fixed component is made up of salaries, retainers and directors' fees.
Debt Policy
There are several reasons why debt policy can be seen as a mechanism for
corporate governance. First, it implies a style of governance that emphasizes rules
over discretion. Increasing debt means that a larger part of the firm's cash flow is
being returned to the bondholders and is therefore being removed from the
control of the managers: the larger the debt, the smaller the discretionary power of
the managers. Second, given the size of the firm, debt financing allows voting
rights to be concentrated more in the hands of the remaining shareholders. Third,
a relatively high debt-to-assets ratio tends to make the firm less attractive as a
takeover target, and may therefore be used as a substitute for takeover defences.
Dividend Policy
Dividend policy, i.e., the target ratio of dividends-to-earnings, can also be seen
as a mechanism for corporate governance. The higher the payout ratio, the
150
ALTERNATIVE MECHANISMS FOR CORPORATE GOVERNANCE AND BOARD COMPOSITION
smaller the amount of free cash flows. As an instrument it operates just like
debt policy, but it imposes much less severe constraints because the payment
of dividends is not mandatory.
Takeover Defences
The remaining mechanisms are direct substitutes as takeover defences. The four
categories we use are borrowed from Malatesta (1992): poison pills, charter
defences, statutory defences and litigious defences.
2
Poison pills include shareholder rights plans,
3
 as well as voting and
preferred stock plans. They are pre-emptive defences par excellence and transfer
substantial power from the shareholders to the managers, in the event that a
prospective bidder be interested in the firm. Charter defences include super-
majority and fair-price clauses as well as board schemes. They must be
approved by the shareholders but, in practice, this is also true of poison pills in
Canada.
4
 Statutory defences (control share laws, fair-price and freeze-out laws)
imply that those controlling the firm can enroll the members of the legislature
to fend off the bid. Finally, litigious defences include, besides litigation, sale of
assets, share repurchases and greenmail or standstill agreements. Although
Malatesta (1992) classifies bond issues as litigious defences, we do not.
This concludes our list of mechanisms for corporate governance. In a
competitive market we would expect each firm to select a subset adapted to its
own circumstances. The most important of the latter is the distribution of own-
ership rights, which appear under the column heading "Distribution of Voting
Rights" and make up the second dimension of Table 1. We return to this shortly.
SELECTION OF A SUBSET OF MECHANISM S
FOR CORPORATE GOVERNANCE
BOTH MANAGERS AND SHAREHOLDERS PARTICIPATE in selecting a set of tools
for corporate governance. The cost/benefit analysis of each group should take
three factors into account: the cost of each mechanism, its effectiveness and
the probability of its use.
Cost of Mechanisms
Table 2 shows the various mechanisms for corporate governance, according to
the nature of the costs involved (opportunity or out-of-pocket costs) and
ranked by ascending order of importance, the smaller ones appearing at the
top of the list. We assume these costs to be borne by the organization and
therefore out of the shareholders' resources.
Board composition is the least costly mechanism for corporate gover-
nance. Appointments to the board of directors may generally be changed at
negligible marginal cost. However, the cost of changing board composition is
151
TABLE 1
CROSS CLASSIFICATION OF THE DISTRIBUTION OF VOTING RIGHTS AND MECHANISMS FOR CORPORATE GOVERNANCE
3
DISTRIBUTION OF BOARD COMPOSITION: EXECUTIVE COMPENSATION: DEBT POLICY:
VOTING RIGHTS RATIO OF OUTSIDE RATIO OF VARIABLE RATIO OF DEBT
TO INSIDE MEMBERS TO FIXED REMUNERATION TO ASSETS
DIVIDEND POLICY:
RATIO OF DIVIDENDS
TO EARNING S
TAKEOVER DEFENCES
POISON
PILLS
CHARTER STATUTORY LITIGIOUS
DEFENCES DEFENCES DEFENCES
Inside Low ratio
Concentrated
Ownership
Outside High ratio
Concentrated
Ownership
Diffuse High ratio
Ownership (in non complex
organizations)
Low ratio for directors Low ratio
(may be high for
second-level managers)
High ratio
High ratio
Low ratio
High ratio
Unstable ratio Light Intensive Intensive Light
use use use use
Unstable ratio Light Light Light Light
use use use use
Stable ratio Intensive Intensive Intensive Intensive
use use use use
Note:
 a
 Expected level (high or low) or behaviour (stable or unstable) or intensity of use of mechanism (intensive or light) is indicated in each cell.
ALTERNATIVE MECHANISMS FOR CORPORATE GOVERNANCE AND BOARD COMPOSITION
likely to be affected by the distribution of voting rights, i.e., as we consid-
er different lines in Table 1. Mounting a proxy fight under diffuse ownership
may involve a substantial and risky investment, but may be hopeless under
concentrated ownership.
Executive compensation is the next mechanism. Again, using the
compensation package as a corporate governance mechanism may involve
only a small or even zero opportunity cost. The amount involved is the excess
of whatever remuneration the shareholders are willing to offer over the
amount that would be optimal if agency problems were negligible.
Charter defences as corporate governance mechanisms may involve only
small out-of-pockets costs, at least when used pre-emptively or initially (as
when the corporation is first organized). Subsequent modifications to the
distribution of voting rights may require the approval of shareholders.
5
Poison pills are potentially a very costly mechanism of corporate gover-
nance, since they give the board the power to veto a takeover. The opportunity
cost resides in the reduction of the expected value of a possible takeover premium.
Although the amount of an eventual premium is likely to increase when the
board has veto power, the probability that an offer will be extended decreases.
Statutory defences imply enlisting the coercive power of the State. This
may be an involved and disruptive undertaking and, therefore, is not used
frequently.
6
 So far, the stock market has generally reacted to the adoption of
poison pills (and other takeover defences) as if they are reducing the share-
holders' wealth,
7
 but the Canadian evidence is inconclusive.
8
The dividend-to-earnings ratio may be increased to return so-called "free
cash flows" to the shareholders. Such a policy may prevent takeovers, but costs
may be incurred in the process. First, changes in dividend policy may affect
the investors' expectations of future cash flows. Second, different payout ratios
will tend to attract clienteles with different tax and consumption preferences.
These investors may be deceived by unexpected policy changes, implying port-
folio rebalancing costs. The costs attached to changes in dividend policy are
also related to the distribution of ownership. For instance, one would expect
the announcement effects mentioned above to be severe with diffuse owner-
ship, but not with concentrated ownership.
Use of the debt-to-assets ratio as a governance mechanism generates
costs similar to those of dividend policy. Assuming, for instance, that individual
investors in the top tax brackets tend to buy shares of lightly leveraged
companies because they prefer to capture the tax benefit of debt financing on
personal account, increasing the debt-to-assets ratio for defence purposes is
likely to impose more financial risk on them than they are willing to bear. The
opportunity cost to the firm may also be significant if there is some validity to
Williamson's (1988) conjecture that some types of assets cannot be financed
through debt because of their low value as collateral. As a consequence, the
strategy of changing the debt-to-assets ratio might entail constraints on asset
selection and imply a suboptimal risk-return combination for the firm and
153
GAGNON & ST-PIERRE
TABLE 2
MECHANISMS FOR CORPORATE GOVERNANCE RANKED IN ORDER
OF INCREASING AGENCY COST
OPPORTUNITY COST OUT-OF-POCKET COST
Board Composition
Executive Compensation
• I. Charter Defences
Poison Pills U- JJ-
Statutory Defences
Dividend Policy
Debt Policy
Litigious Defences
portfolio rebalancing costs for the shareholders. Under concentrated ownership,
for control considerations, the latter already bears unsystematic risk and may
be reluctant to increase the debt-to-assets ratio. We suggest, therefore, that
changes in the structure of liabilities may be a costly mechanism for corporate
governance.
Finally, litigious defences appear to be a mechanism of last resort: court
proceedings are disruptive and also entail direct costs. Furthermore, they are
remedial, not pre-emptive. We doubt that management would propose ex ante, to
use them.
Effectiveness of Mechanisms
If effectiveness is defined as a relationship between the means to an end and
the achievement of that end, then shareholders and managers do not rank
mechanisms in the same order. They have conflicting interests: the managers
want to maximize the private benefits they draw from control while the
owners wish to maximize the present value of the residual cash flows. (These
hypotheses are discussed further in the next section.)
For instance, shareholders may consider that connecting managers'
remuneration packages with the market value of the firm's shares is an effec-
tive way to align the managers' interests with theirs. The managers may readily
accept or choose that type of contract, but it does not reduce their risk of
losing control, unless it provides them with a significant proportion of the
voting rights, which is not generally the effect being sought. Therefore,
although executive compensation can be an effective mechanism for corporate
governance for the shareholders, it is not for the managers. We conjecture the
opposite holds true for poison pills. They provide the managers with powers
the shareholders could retain for themselves.
9
154
ALTERNATIVE MECHANISMS FOR CORPORATE GOVERNANCE AND BOARD COMPOSITION
TABLE 3
MECHANISMS FOR CORPORATE GOVERNANCE RANKED IN ORDER OF DECREASING
EFFECTIVENESS FOR SHAREHOLDERS AND MANAGERS
SHAREHOLDERS
DECRI
EFFECT
Board Composition (prerequisite) •U
Charter Defences
Executive Compensation
Debt Policy
Dividend Policy
Statutory Defences
Poison Pills
Litigious Defences
MANAGERS
AS1NG
VENESS
li Poison Pills
Statutory Defences
Debt Policy
Dividend Policy
Executive Compensation
Board Composition
Charter Defences
Litigious Defences
Note: Diffuse ownership is assumed.
In Table 3, we list the mechanisms for corporate governance according
to decreasing effectiveness for the shareholders and the directors: the rankings
are not identical.
Board composition appears as the first item on the shareholders' side
because it is a prerequisite condition to using any of the other mechanisms on
the list. Policy decisions can be implemented only through the board.
From the shareholders' point of view, charter defences appear to be quite
effective. The initial owners may have the statutes written and modified
eventually according to their preferences. Subsequent investors for whom control
considerations are important may select companies according to the jurisdiction
under which they have been created. This provides them with control over
policy decisions and the distribution of the wealth created by the firm.
We have already suggested that poison pills are not very effective from
the shareholders' point of view because shareholders tend to defend their position
by surrendering it to the managers. That is why they appear so close to the
bottom of the list on the left of Table 3, but at the top of the list on the right.
The role and effectiveness of statutory defences, from the managers'
point of view is ambiguous. As a preventive measure a statutory defence is
generally presented as a device to protect the shareholders. For instance, the
law may require a minimum number of directors to be independent,
10
 or it may
stipulate that no shareholder may hold more than a given percentage of the
votes." However, the actual effect may differ from the intended effect.
Restricting the exercise of voting rights may favour managerial entrenchment.
Furthermore, a number of empirical studies
12
 suggest that, in the United
155
GAGNON & ST-P1ERRE
States, changes in the legislation designed to increase their effectiveness have
tended to reduce the market value of the affected firms. Therefore, such
changes were interpreted as effectively protecting the managers rather than
the shareholders. Consequently, we see statutory defences as rather effective, if
infrequent, mechanisms for corporate governance when employed by the
managers.
We then list debt and dividend policy on the managers' side. By accepting
more or less severe constraints as to the distribution of its cash flows, the firm
becomes less or more susceptible to a change in control and, therefore, the risk
of the managers of losing private benefits is reduced or increased.
Were it not for golden parachutes (one of its components), executive
compensation would appear at the bottom of the list, on the managers' side.
However, golden parachutes and similar contracts guarantee some monetary
compensation in the event that private benefits are transferred to another group.
Board composition is a mechanism that managers cannot use directly but
only through their influence on the shareholders, who may change it at will.
Therefore, it receives a rather low rank on the managers' (right) side of Table 3.
Probability of Use
The expected value of a mechanism for corporate governance is a function of its
cost, effectiveness and probability of use. We now consider this third aspect.
We suggest that the main factor affecting the probability of use is the
distribution of ownership rights. Table 1 presents three cases horizontally.
A simple (albeit an extreme) case is that of the owner-managed firm,
used as a starting point by Jensen & Meckling (1976). A manager who owns
all the voting rights, has no need of a takeover defence. Debt and dividend
policies can be selected according to portfolio-consumption needs, and
executive compensation is not relevant for the owner-manager, only for
subordinates. Finally, the directors are all insiders or affiliated outsiders selected
according to the owner's preferences, unless nominations dictated by some debt
covenant have been accepted. Specifically, it has been observed that owner-
managers prefer lower debt-to-assets ratios.
13
 Therefore, in this case, all the
constraints on the selection of mechanisms for governance are likely to be based
on operating and consumption, as opposed to control, considerations. In this
regard, for the owner-managed firms, which would be part of the first line of
Table 1, we would expect the observed mechanisms for governance to be
combined randomly because there is a very low probability that they will be
useful. This should remain so at least as long as one individual or a group of
closely related individuals owns more than 50 percent of the voting rights of
the firm.
In this regard, the opposite extreme is shown at the bottom of Table 1.
We assume that shares have been sold to a broad spectrum of outside investors
and there is no controlling shareholder. Such business entities might be
156
ALTERNATIVE MECHANISMS FOR CORPORATE GOVERNANCE AND BOARD COMPOSITION
expected to be relatively large and to be found mostly in the so-called
transparent industries. In this case the shareholders hire professional managers
who capture private benefits from control, at the owners' expense, and are moti-
vated to exert influence on the selection of corporate governance mechanisms
because they may affect their own wealth. Several consequences follow from
this action. More resources should be allocated to selecting the mechanisms
because two parties with conflicting interests are involved and the probability
that any instrument shall be used is increased. For instance, some mechanisms,
such as takeover defences, which are not useful to the owner-manager, may be
seen as effective risk-reducing devices for the managers, who will negotiate
their adoption by the shareholders. The behaviour of the shareholders is also
likely to be affected: they now need to control and monitor the agents. Board
composition and the compensation package acquire additional importance. As
the shares are now owned through diversified portfolios, financial risk (i.e.,
higher debt-to-assets ratios) may become acceptable. Finally, dividend policy
should be tailored for a specific clientele and therefore remain stable. The
bottom line of Table 1 summarizes the new situation.
Outside concentrated ownership represents an intermediate situation.
BEHAVIOURAL HYPOTHESES
As WE HAVE ALREADY SUGGESTED, corporate governance inter-relates professional
managers, shareholders and directors, whereas agency theory is concerned
primarily with the conflicts of interests between the agents (the professional
managers) and the principal (the common shareholders) who hired them to
manage the firm for his benefit. Plausible assumptions can readily be made
about managers and shareholders, but the directors' position is ambivalent.
They cannot be classified as agents or principals unless additional information
is available. We examine these questions briefly in this section. All of the
players are assumed to behave opportunistically: each attempting to maximize
his own utility function, subject to the constraints imposed by the other parties.
The principal maximizes his own wealth, part of which is represented by
the present value of the residual cash flows derived from the firm. To achieve
that objective, he attempts to monitor the managers and align their interests
with his own. Various mechanisms are available to him. He may seek member-
ship on the board of directors; he may prefer relatively high debt-to-assets and
dividend payout ratios, in order to minimize the free cash flows; or he may
allow the managers to increase their personal wealth by allowing or encouraging
them to invest in the firm. However, none of these mechanisms is without
cost. For instance, being an important shareholder increases the likelihood of
that principal being appointed to the board of directors, but it also implies
that otherwise diversifiable risk must be borne. Increasing the proportion of
equity held by the managers may, at some level, entail entrenchment and
excessive private benefits for the managers. Therefore, we expect the principal
157
GAGNON & ST-PIERRE
to substitute some mechanisms for others, until the desired equilibrium
between monitoring and control is reached.
Agents want to maximize the private benefits they can draw from the
firm. They seek retrenchment in it. Control of some voting rights, as suggested
by Stulz (1988), representation on the board and adoption of takeover
defences may enable them to achieve that objective. But, again, costs are
involved. For instance, assuming that takeover bids can hurt the managers,
but not the shareholders, the latter may object to takeover defences and "vote
with their feet", i.e., sell their shares and reduce the value of the firm.
Similarly, the relationship between managerial ownership and firm value need
not be monotonic. It has been observed empirically
14
 that, at some levels,
increasing the managers' ownership may reduce firm value. Managers might
also be expected to substitute mechanisms for corporate governance up to the
point where their control over the free cash flows produced by the firm is still
acceptable to the shareholders.
As mentioned earlier, the directors are members of a third group, whose
allegiance is somewhat ambiguous. In order to classify one as an agent or a
principal, two characteristics must be known: his status and the proportion of
voting rights he controls. Status refers to being an inside (employee) or out-
side (independent) director. The former is assumed to have a larger proportion
of his human capital than the latter invested in the firm. (This classification is
further refined below.) In principle, an outside director owes his undivided
allegiance to the shareholders and is assumed to be more likely than the inside
directors to challenge a controversial decision by the CEO. However, an inside
director with a significant proportion of his own financial wealth (which may
nonetheless represent a negligible proportion of the total value of the firm)
invested in the firm is more likely to play that role than an outside member
who owns only qualifying shares. Therefore, to assess the motives of a director
properly, those two characteristics must be taken into account.
With respect to the determinants of board composition, again, two
propositions derive from these considerations. First, two characteristics of the
distribution of voting rights are important: the proportion held by the directors
themselves, and the concentration of voting rights among the remaining share-
holders. An increase in the proportion owned by the directors might be
expected to entail an increase in the proportion of inside members of the
board, while an increase in the concentration of shares held by non-members
should entail an increase in the proportion of outside directors. In this study,
we assume the distribution of ownership to be exogenously determined.
Second, given the distribution of ownership (i.e., given a line in Table 1),
substitution might be expected to occur among the various mechanisms for
corporate governance. Alternatively, given a column in Table 1, the numerical
values on each line might be expected to differ markedly. While monitoring
and entrenchment represent behaviours attributed to outside shareholders and
managers, respectively, substitution should characterize both groups. The
158
ALTERNATIVE MECHANISMS FOR CORPORATE GOVERNANCE AND BOARD COMPOSITION
outcome is that the so-called substitution hypothesis may, in principle, be held
in two extreme forms.
The firms will select the combinations of mechanisms that enable them
to survive and prosper, given their peculiar circumstances. If the mechanisms
are literally perfect substitutes, all the possible combinations are equally
effective. To the observer of a cross-section of surviving firms, they would
appear as having been randomly selected. Alternatively, the mechanisms may
be very imperfect substitutes, being related only as complements. If the optimal
combination for each firm is unique, then, again, in a cross-section study of
surviving firms, they would appear as having been randomly assembled.
The actual universe of firms is probably best described by propositions
located between these two extremes; corporate control mechanisms are neither
perfect substitutes nor complements. Although economic Darwinism (to use
Weisbach's phrase) is felt gradually, there must be some broad generalization to
be made about them. Firms in similar circumstances should have similar
combinations of corporate governance mechanisms. Given that those common
traits or factors have been correctly identified, their effects should be discernible
empirically. The rest of this study presents such an effort with regard to the
composition of the boards of directors.
BOARD COMPOSITION
THE RELEVANCE OF BOARD COMPOSITION
A LARGE PART OF THE LITERATURE dealing with board composition is focused
on the cause-and-effect relationship illustrated below.
Each of the three links must be examined.
The proportion of outside
15
 to inside directors is generally used as the
indicator of board composition. The outsiders are seen as professional referees
who assess the managers' performance, determine their remuneration and
replace them if necessary. They are assumed to be loyal to the shareholders.
The mere existence of professional or outside directors provides evidence
of a desire to monitor managers. However, it does not per se indicate that such
monitoring is actually carried out or conducted efficiently. Persistence of a
relatively high proportion of independent directors is consistent with at least
four (not mutually exclusive) hypotheses. First, it may be a harmless practice
159
GAGNON & ST-PIERRE
that can persist for some time after it has become obsolete. However, one
would expect it eventually to disappear and be replaced by some other mechanism
for governance, unless it is actually beneficial to some of the organization's
stakeholders. Second, the presence of outside directors may be useful to the
shareholders who have selected, or at least accepted, them as monitors. This is
the hypothesis tested by most researchers and implicitly held by the lawmakers.
Third, the outside directors may, paradoxically, play a role that is useful to the
managers, who account for their presence as part of the bonding costs they
incur to guarantee they are not bound to expropriate the shareholders. In other
words, the professional directors might first exist as creatures of the managers
and continue to exist, provided their role is seen by the shareholders as indif-
ferent. Fourth, the presence of at least some outside directors may be required
by the regulatory authorities.
There is some empirical evidence in the United States that outside
directors are seen by the market as monitors of the managers. The correlation
between poor performance and CEO resignations is stronger with outsider-
dominated boards.
16
 In addition, the number of outside directors tends to
increase after unsatisfactory performance.
17
 Other results suggest that outside
directors help control managerial consumption of perquisites and monitor
banking acquisitions and are also associated with lower costs in the insurance
industry.
18
 Furthermore, the market also recognizes that the quality of monitoring
is not constant. The senior executives of companies that reduce their dividends
are less likely to receive additional outside directorships than the senior
executives of companies that do not.
19
 In this context, there appears to be a
positive correlation between holding fewer additional outside directorships
and the likelihood of being the target of a hostile takeover.
20
 However, there is
no evidence in Canada that board composition is related to the board's reaction
to a takeover bid.
21
 Another study has concluded that attention to board structure
may be misplaced because managerial incentives, such as compensation, may
be more efficient mechanisms for monitoring.
22
Establishing a link between board composition and monitoring is an
intermediate objective, which is interesting only to the extent that it enables
us ultimately to relate board composition to overall performance. This second
undertaking is more difficult because of the problem of endogeneity. It is
conceivable that board composition determines performance to some extent,
but it is also possible that performance is one of the determinants of board
composition. First, outside directors may be drafted to improve unsatisfactory
performance, as suggested above, which would be the source of measurement
problems (to which we shall return). But, fundamentally, potential agents such
as prestigious outside directors can minimize their professional nondiversifiable
risk by accepting more readily appointments to the boards of the most profitable
firms. Admittedly, on the supply side, the number of candidates for board seats
is relatively large, but we expect them to display opportunistic behaviour, as
do all other agents. The simultaneous relationship is suggested by the arrows
in the illustration above.
160
ALTERNATIVE MECHANISMS FOR CORPORATE GOVERNANCE AND BOARD COMPOSITION
These problems may explain why there is precious little empirical evidence
concerning the relationship between board composition and overall perfor-
mance. One study concludes that there is no relation;
23
 another cautiously
suggests there are some fragile nonmonotonic effects.
24
 A positive relation has
been detected in cross-country comparisons, not taking into account the
endogeneity problem.
25
To summarize, there is some evidence that independent directors do
monitor managers (i.e., the intermediate relation), but research efforts have not
shown convincingly that they succeed in improving overall performance
(i.e., the ultimate and interesting relation). In this study, we consider both aspects.
Board composition is multidimensional. Researchers have studied the
proportion of outside members, the number of directorships in other companies
held by the members, the membership of the committees of the board, and the
separation of the positions of CEO and Chairman of the Board. We take the
first three dimensions into account. Furthermore, as suggested in Table 1, the
distribution of ownership among the directors and the remaining shareholders
is a determinant of board composition and its role.
26
 We examine the determi-
nants of board composition under diffuse ownership, which is our base case,
and then make whatever adjustments are necessary to take into account
concentrated inside and outside ownership.
DIFFUSE OWNERSHIP
Ratio of Outside to Inside Members
This dimension of board composition may be seen as a proxy for the level of
monitoring of the managers by the shareholders. The concept has found its
way into Canadian corporate law and several studies
27
 have attempted to
uncover its determinants.
We partition the board into four groups: inside, outside affiliated, group
representatives and independent.
Inside directors are employees or ex-employees of the firm or of one of its
subsidiaries. Following Shivdasani (1993), we define outside affiliated directors
as members who, although not employees, have substantial business relation-
ships with the incumbent managers.
28
 Lawyers, underwriters, accountants and
consultants, as well as relatives of inside directors among others, are included
in this subgroup. The inside and outside affiliated members are collectively
designated as "insiders". The third subgroup comprises group representatives. It so
happens that companies having shareholders in common also tend to have
interlocking directors. We classify as "group directors" board members holding
directorships in other companies of the same group.
29
 Finally, we classify indepen-
dent directors as those who appear to have no relationship with the firm
except for their membership on the board. The group representatives and
independent directors are collectively designated as "outsiders".
30
161
GAGNON & ST-PIERRE
The level of monitoring provided by outsiders is often estimated through
the sheer proportion of their membership on the board. However, their own
personal characteristics are also important. The professional directors have
their reputation (human capital) at risk, and the number of directorships
31
they already hold in other companies may be seen as an indicator of the value
of that capital.
As the probability of their obtaining additional directorships is a function
of their performance as perceived by the shareholders, it is a plausible hypothesis
that outside "prestigious" directors are more likely to challenge a manager's
decision because they have more to lose from unsatisfactory orientations than
non-prestigious members. They offer high-quality monitoring. Weighting each
director by the number of additional directorships he holds may provide a better
proxy for his importance.
In the same vein, any director may have more or less influence on the
board. Weighting each of them by the number of committees on which he sits
may also provide a better estimator of his influence on corporate decisions.
Taking the ratio of outsiders to insiders as the dependent variable, what
are the explanatory variables? We hypothesize that board composition is
determined by three factors: complexity of the organization, important share-
holders, and operating conditions.
Complexity of the Organization
Following Fama & Jensen (1983), we qualify as complex those organizations
where the transfer of information from agents to principal (or even between
agents) is costly. The more costly the transfer, the more complex the organization,
and so it is more likely that decision management is separated from decision
control. Such a situation calls for a high level of monitoring and, therefore, a
high ratio of outsiders to insiders.
The notion of the board as a meeting place where the inside and outside
directors trade information implies that the optimal board cannot be made up
exclusively of insiders, even though the latter include the affiliates that small
firms may find expedient to have on the board. Neither can it be composed
exclusively of outsiders: at least, the CEO will be a board member, even
though he may be the sole insider.
32
 Furthermore, it is plausible that the mar-
ginal contribution of the outside monitors decreases, especially after their pro-
portion exceeds 50 percent.
We assume that the determinants of complexity are the type of technology
and the size of the firm. Highly developed and rapidly changing technologies
make it more difficult for the shareholders to monitor the strategic decisions of
managers. This opacity increases with the level of R&D activity, or the ratio of
scientific and professional employees to total employees.
Complexity is also a (nonlinear) function of corporate size. Large firms
are more likely to have diffuse ownership and offer "independent" monitors to
162
ALTERNATIVE MECHANISMS FOR CORPORATE GOVERNANCE AND BOARD COMPOSITION
FIGURE 1
RELATIONSHIP BETWEEN THE PROPORTION OF OUTSIDE DIRECTORS
AND FIRM SIZE OR COMPLEXITY
the capital market, as part of their bonding costs. To summarize, we expect the
relationship between the proportion of outside board members and size or
complexity to be as depicted in Figure 1.
Large Shareholders
Even under diffuse ownership, a firm may have large non-controlling share-
holders other than its own directors: families, institutional investors or some
other firm (hereafter referred to as blockholders). It is efficient for such investors
to have their preferences known. As they have to bear some unsystematic risk,
they are likely to require a relatively high level of monitoring, especially by
directors who meet with their approval. The larger the number of important
blockholders and their holdings, the higher the ratio of outside to inside
members of the board.
Shareholdings by the inside directors are also relevant. The entrench-
ment hypothesis suggests that as they acquire additional shares, inside directors
will seek additional seats on the board, thereby causing the ratio of outside to
inside directors to fall.
Regulated firms, such as financial institutions and public utilities,
represent a special case. Part of the monitoring function is taken over by the
State, usually through a regulatory agency, which then plays the role of a large
shareholder and explicitly or implicitly imposes a relatively large number of
independent directors.
33
 Given the coercive powers of the regulators, one should
expect an even stronger effect than with "ordinary" important shareholders.
163
GAGNON & ST-PIERRE
There is an alternative political reason why regulated firms may enroll prestigious
outside directors: they may carry weight with the regulators and other
representatives of the State. This is in the nature of a bonding cost incurred
by the organization. Again, this reasoning suggests that regulated firms
should be expected to retain the services of more outside directors. However,
this effect would play in the opposite direction under the substitution hypothesis
as envisioned by Demsetz & Lehn (1985). They suggest that as the monitoring
function is largely taken over by the State, there is less need for outside
directors.
Operating Conditions
The phrase "operating conditions" is meant to include both the return on
assets and the financial policies with respect to indebtedness and dividends.
We have already suggested that the relationship between board compo-
sition and financial performance is not necessarily one-sided. However, we
ignore the reciprocal character of the relationship for the moment. Even if
such a restrictive assumption allows a workable description of firm behaviour,
it may be difficult, empirically, to detect a relationship between firm perfor-
mance and board composition. For instance, assume that firms tend to replace
inside with outside directors when financial performance deteriorates.
Depending upon the length of time it takes for the firm to react and the
length of time it takes for the effects of the reaction to be felt, one may
observe in cross-section studies that, contrary to expectations, increases in the
proportion of outsiders are associated with deteriorating performance and that
no relationship between the proportion of insiders and performance is detected.
Panel data would be required to provide evidence on this mechanism.
34
Nevertheless, we still expect our preliminary tests to detect a positive
association between the rate of return on assets and the ratio of outside to
inside directors. To abstract from industry effects, we compute relative rates of
return (i.e., rates of return in excess of industry averages).
Financial policies with respect to debt and dividends may be seen as part
of the operating conditions under which board composition has to be
determined. On the one hand, we assume that outsiders are less likely to resist
a takeover bid because it generally includes a premium for the shareholders.
On the other hand, relatively high debt-to-assets
35
 and dividend payout ratios
will contribute to a reduced probability of a takeover offer. Therefore, under
the substitution hypothesis, higher debt-to-assets and payout ratios would
allow managers and inside directors to accept a higher proportion of outside
directors. However, from the shareholders' point of view, the predicted
association would be negative: the smaller the free cash flows, the less useful
the outside directors' monitoring. But then, as the debt level rises, they might
also prefer more monitoring. In addition, managers may see the various anti-
takeover devices as complements rather than substitutes and tend to increase
164
ALTERNATIVE MECHANISMS FOR CORPORATE GOVERNANCE AND BOARD COMPOSITION
the debt-to-assets and payout ratios when they believe they are vulnerable to a
takeover bid. At that point, they may also prefer to acquire additional protec-
tion
36
 by reducing, or at least not increasing, the proportion of independent
directors. This suggests that the relationship we are considering is ambiguous.
To summarize, under diffuse ownership we expect board composition to
be explained statistically in the following manner:
Board Composition = weighted or unweighted ratio of
outsiders to insiders
= / (complexity, size, independent block-
holdings, voting rights held by inside
directors, voting rights held by outside
directors, regulation, return on assets,
debt-to-assets ratio, and dividend
payout ratio) (1)
The signs in equation (1) are predicted to be positive, except for the sign
attached to voting rights held by the inside directors, which should be negative;
the signs attached to the financial variables are indeterminate.
CONCENTRATED INSIDE OWNERSHIP
IN THIS INSTANCE, THE CONTROL OF THE FIRM is assumed to be securely held by
the managers and the inside directors. Control considerations are not important:
there is little need, for instance, for takeover defences and substitution
between the various mechanisms for corporate governance. The selection of
board members is expected to be dictated entirely by the operating conditions.
This leads one to expect the mean ratio of outsiders over insiders to be low
under concentrated inside ownership. First, one would expect the inside directors
to select as directors individuals with the technical skills required by the firm.
For smaller firms, it may be more efficient to have some consultants as members
of the board rather than to hire them on a part-time basis. The number of
affiliated outside directors should then increase. Because we classify them as
insiders, the ratio of outsiders over insiders would decrease. Second, we expect
the firms in that group to be, on average, smaller in size. This entails a smaller
proportion of outsiders on the board and also a stronger size effect, as suggested
by Figure 1. Except for size and blockholders, the explanatory variables
complexity, debt-to-assets ratio" and dividend payout discussed under diffuse
ownership should not be statistically significant under concentrated inside
ownership.
Under concentrated ownership, the behaviour of two explanatory variables
may be especially interesting: regulation, and the rate of return on assets. If
the ratio of outside to inside members is increased in response to explicit or
implicit requirements of the regulatory authority, the variable regulation is
165
GAGNON & ST-PIERRE
expected to be statistically significant for the three types of ownership. But if
it is a bonding cost incurred to satisfy the shareholders, it is expected to be
significant only under diffuse ownership. Under concentrated ownership, the
large shareholders are more likely to rely on their own monitoring.
A similar proposition applies to the rate of return on assets. If the outside
directors "explain" the latter to some extent, this effect must be observed
under the three types of ownership. Otherwise, the data are consistent with
the competing hypothesis that it is the profitable firms that tend to attract the
prestigious directors, whose usefulness is then greater under diffuse than under
concentrated ownership.
CONCENTRATED OUTSIDE OWNERSHIP
UNDER CONCENTRATED OUTSIDE OWNERSHIP, one would expect the presence of
group directors, who we have classified as outsiders, because they owe their
primary allegiance to the group as a whole. That is also why the other block-
holders might insist on being able to rely on their own independent outside
directors. This entails a higher ratio of outsiders to insiders. Except for this
consideration, the relevant statistical model should be similar to the one
proposed under concentrated inside ownership.
DATA
T
HE EMPIRICAL PART OF THIS STUDY is based on a sample of 258 Canadian
firms drawn from two sets of data.
The first set is made up of 151 observations used in a previous study of
boards' reactions to takeover bids (St-Pierre et aL, 1994). The data were
extracted from the files of the Commission des valeurs mobilieres du Quebec
(QSC) for the period 1978-91.
w
 The shareholders of those firms all received
takeover offers, 51 of which were contested by the boards of directors. We
define these as hostile or disciplinary bids and classify the firms as targets.
Following Morck et al. (1988a), we assume the remaining 100 observations to
have no special characteristic worthy of mention. In fact, when we compared
targets and non-targets with regard to all the variables used in this study, we
found they were not statistically different except for the distribution of voting
rights. The inside directors of target firms own a significantly lower proportion
of voting rights than the directors of non-target firms.
The second set is a control sample of 107 observations drawn from the
Financial Post (FP) files. First, a random sample of 500 firms on FP cards was
examined in order to remove those subject to takeover offers, mergers or major
restructuring during the five-year period ending with 1992. The analysis was
based on the events as reported in the "History" section of the FP cards. A
firm was discarded if there is no information concerning the members of its
board of directors. The 1986 to 1988 financial statements were collected for
the remaining 107 firms and added to the data bank.
166
ALTERNATIVE MECHANISMS FOR CORPORATE GOVERNANCE AND BOARD COMPOSITION
BOARDS OF DIRECTORS
FOR BOTH SETS OF DATA, we examine the corporate proxy circulars, the Financial
Post's Directory of Directors and Kofmel's Who's Who in Canadian Business to trace
individual information about the directors. Each one is classified as:
• an inside director, if he is an employee or ex-employee of the firm
or of one of its subsidiaries;
• an affiliated outside director, if he appears to be a relative of some
inside director or is referred to as having some business relation-
ship with the firm;
• a group director, if he owns an additional directorship in a firm of
the same group, as identified by Statistics Canada;
• an outside (independent) director, otherwise.
The first two subgroups were designated as insiders; the last two as outsiders.
We noted the number of board committees the director may sit on, the num-
ber of voting rights he and his family own or control, as well as the total num-
ber of directorships he may hold in other companies quoted on a stock
exchange. (When restricted voting shares are outstanding, we took into
account voting rather than ownership rights.)
COMPANIES
FOR COMPANIES, WE ALSO SEARCHED the proxy circulars, the FP cards and
Statistics Canada to uncover the distribution of voting rights among the
shareholders. A company was then classified as to its type of ownership:
• inside concentrated ownership: directors own 20 percent or more
of the voting rights;
• outside concentrated ownership: three or fewer of the main share-
holders own 20 percent or more of the voting rights;
• diffuse ownership: three or fewer of the main shareholders or
directors do not own 20 percent or more of the voting rights.
The financial ratios and other data were computed as follows:
Dividends ratio of ordinary dividends over book value of common shares.
39
ROA "excess rate of return" equal to the difference between the
accounting rate of return on assets of the firm i and that of its
industry j, divided by the absolute value of industry j's average
rate. The averages are computed over the two fiscal years preceding
the observation date.
167
GAGNON & ST-PIERRE
| average operating profit | [ average operating profit |
I total assets J j v- total assets J j
average operating profit
total assets
i
Assets
Industry
Target
D/A
Total assets of company i, in Canadian 1992 dollars (i.e., inflated
with the Consumers' price index).
Standard industrial classification code of firm i, according to
Statistics Canada.
Dummy variable equal to one for firms having received takeover
bids contested by the board, equal to zero otherwise.
Total debt'to-total assets ratio, using book values.
Managers' compensation Ratio of variable to fixed compensation.
: VS + VOp
SB + DF + DC + OC
/here:
VS = Market value of the shares owned by the managers
and their relatives
VOp = Market value of stock options that may be exercised
and owned by the managers and their relatives
SB = Salary and bonus
DF = Directors' fees
DC = Deferred compensation such as fringe benefits and the
firm's contributions to management savings plans
OC = Other compensations such as financial benefits from
interest-free loans to managers.
The industry to which firm i belongs is also characterized by:
Knowledge ratio
Regulated
Ratio of scientific, technical and professional to total
personnel, as provided by Beck (1992).
Dummy variable equal to one for the firms in: public utilities,
communications, storage, transportation and financial services
except holding companies, equal to zero otherwise.
DESCRIPTIVE STATISTICS
IN TABLES 4 AND 5 WE PRESENT SOME DESCRIPTIVE STATISTICS of the main
characteristics of the 258 boards of directors we examined.
168
ALTERNATIVE MECHANISMS FOR CORPORATE GOVERNANCE AND BOARD COMPOSITION
TABLE 4
FREQUENCY DISTRIBUTION OF THE PROPORTION OF OUTSIDERS
3
ON THE BOARD OF DIRECTORS
NUMBER OF COMPANIES
PERCENT
0-10
11 -15
16-20
21 -25
26-30
31 -35
36-40
41-45
46-50
51 -55
56-60
61 -65
66-70
71-75
76-80
81 -85
86-90
91 -95
96 - 100
TOTAL
DIFFUSE
FULL SAMPLE OWNERSHIP^
8
7
8
7
17
11
28
19
22
9
33
17
25
18
12
7
6
3
1
258
MEAN (%) 51
MEDIAN (%) 54
Notes:
 a
h
c
d
3
3
2
0
4
2
7
8
4
2
11
6
9
10
5
6
1
1
0
84
55
58
CONCENTRATED
INSIDE OWNERSHIP
C
4
3
5
5
8
6
8
8
8
4
6
8
6
4
5
0
3
0
0
91
45
44
CONCENTRATED
OUTSIDE OWNERSHIP^
1
1
1
2
5
3
13
7
6
3
16
3
10
4
2
1
2
2
1
83
53
55
Includes both the group representatives and the independent directors with no connection
with the firm except for their board seats.
Neither the directors nor the three main shareholders own 20% or more of the voting rights.
The directors own 20% or more of the voting rights.
The three main shareholders own 20% or more of the voting rights.
First, we note the numerical importance of the outsiders (Table 4).
Although our definition of an outsider is somewhat restrictive, even under
concentrated inside ownership only 13 percent of the firms have a ratio of out-
siders to insiders equal to or smaller than 25 percent. The median is higher
than 50 percent, the absolute majority, under both diffuse and concentrated
outside ownership. Second, the choice of operational definitions for the four
subsets of directors is important (Table 5). Some characteristics that appear to
be statistically different between the three types of ownership are, however,
not statistically different when the inside and outside affiliated directors are
subsumed under the caption "insiders", and group and outside directors are
subsumed under "outsiders", and vice versa.
169
TABLE 5
MAIN CHARACTERISTICS OF THE BOARDS OF DIRECTORS
Average Number on Board:
Inside Directors
Outside Affiliated Directors
Insiders: Subtotal
Outside "Group" Directors
Independent Directors
Outsiders: Subtotal
DIFFUSE
OWNERSHIP
GR 1
N=84
3.08
1.20
4.29
2.15
3.80
5.95
CONCENTRATED
INSIDE
OWNERSHIP
GR2
N=91
3.99
1.15
5.14
1.36
3.56
4.92
CONCENTRATED
OUTSIDE GR 1 VS GR 2
OWNERSHIP
GR3
N=83
3.60
0.99
4.59
1.48
3.81
5.29
TTEST
IT I
3.25
0.27
2.60
1.84
0.55
1.83
PROB.
0.0014
0.7868
0.0100
0.0689
0.5831
0.0688
GR 1 vs GR 3
TTEST
IT I
1.66
1.23
0.89
1.82
0.02
1.29
PROB.
0.1010
0.2213
0.3730
0.0712
0.9822
0.1974
GR 2 vs GR 3
TTEST
IT I
1.15
1.01
1.50
0.35
0.67
0.78
PROB.
0.2537
0.3156
0.1345
0.7247
0.5063
0.4374
OVERALL
SAMPLE
WILCOXON TEST
X2
8.43
1.28
5.15
7.06
0.91
4.64
PROB.
0.0148
0.5272
0.0762
0.0293
0.6337
0.0981
Average Number of Positions on Board Committees by:
Inside Directors
Outside Affiliated Directors
Insiders: Subtotal
Outside "Group" Directors
Independent Directors
Outsiders: Subtotal
1.56
0.73
2.29
1.62
2.39
4.01
1.86
0.56
2.42
0.98
1.76
2.74
1.95
0.61
2.57
0.89
2.30
3.19
1.05
0.96
0.36
1.43
1.24
1.92
0.2949
0.3425
0.7212
0.1550
0.2168
0.0562
1.25
0.60
0.72
1.95
0.16
1.24
0.2118
0.5489
0.4702
0.0530
0.8720
0.2167
0.28
0.36
0.38
0.25
1.29
0.83
0.7763
0.7189
0.7037
0.8004
0.1994
0.4094
0.42
0.12
0.36
7.33
1.17
3.60
0.8090
0.9439
0.8348
0.0256
0.5584
0.1652
Average Number of Directorships in Other Corporations by:
Inside Directors
Outside Affiliated Directors
Insiders: Subtotal
1.67
2.20
1.87
1.25
1.37
1.34
1.28
2.42
1.46
1.42
1.85
1.80
0.1589
0.0670
0.0732
1.33
0.44
1.45
0.1867
0.6614
0.1481
0.11
2.49
0.50
0.9107
0.0147
0.6198
3.25
7.19
5.58
0.1970
0.0274
0.0613
•vj
o
TABLE 5 (CONT'D)
Outside "Group" Directors 3.68
Independent Directors 2.09
Outsiders: Subtotal 2.62
Average tenure on board (in years)
Inside Directors 8.23
Outside Affiliated Directors 6.13
Insiders: Subtotal 10.26
Outside "Group" Directors 4-47
Independent Directors 6.87
Outsiders: Subtotal 7.44
Percentage of voting rights held by:
Inside Directors 1.57
Outside Affiliated Directors 0.16
Insiders: Subtotal 1.73
Outside "Group" Directors 0.39
Independent Directors 0.67
Outsiders: Subtotal 1.06
3.37
1.19
1.73
by:
11.58
5.62
9.76
2.85
6.24
6.59
46.46
0.91
47.37
5.52
3.71
9.23
2.59
1.60
2.05
8.21
4-50
9.30
4.94
6.36
7.36
2.13
0.11
2.24
0.41
0.49
0.90
Notes: Inside directors - employees or ex-employees of the company or of one of
Outside affiliated directors - relatives of inside directors and c
0.54
3.55
2.91
3.75
0.53
0.38
2.09
0.80
0.64
15.23
1.74
15.80
3.07
3.21
4.33
its subsidiaries.
0.5905
0.0005
0.0041
0.0002
0.5984
0.7027
0.0380
0.4227
0.5316
0.0000
0.0845
0.0000
0.0025
0.0016
0.0000
2.14
1.73
1.76
0.02
1.68
0.68
0.45
0.71
0.05
0.86
0.42
0.77
0.07
0.54
0.34
0.0350
0.0851
0.0778
0.9867
0.0950
0.4979
0.6545
0.4816
0.9591
0.3904
0.6741
0.4408
0.9482
0.5895
0.7349
1.67
1.96
1.34
3.79
1.24
0.35
2.43
0.18
0.51
14.83
1.86
15.41
3.04
3.46
4.39
0.0991
0.0522
0.1813
0.0002
0.2150
0.7279
0.0163
0.8610
0.6090
0.0000
0.0644
0.0000
0.0027
0.0007
0.0000
4.97
6.78
7.61
17.07
2.85
0.47
4.77
0.74
0.20
153.37
7.89
8.60
2.05
18.72
3.31
0.0832
0.0336
0.0223
0.0002
0.2411
0.7904
0.0919
0.6893
0.9052
0.0001
0.0194
0.0136
0.3586
0.0001
0.1912
irectors having some business relationship with the firm.
Outside group directors - directors holding an additional directorship in a firm of the same sjroup.
Independent directors - all other directors.
Diffuse ownership - neither the directors nor the three main shareholders own 20% or more of the voting rights.
Concentrated inside ownership - the directors own 20% or more of the voting rights.
Concentrated outside ownership - the three main shareholders own 20% or more of the voting rights.
GAGNON & ST-PIERRE
Not only are the outside directors more numerous under diffuse owner-
ship, but also they occupy more positions on their boards' committees and
have received more directorships in other corporations. By contrast, the inside
directors enjoy longer tenures on their boards, which is related to their control-
ling a relatively high proportion of the voting rights. This is attributable to our
partitioning our sample on that basis.
In most respects (except for percentage of voting rights held) the two
groups under concentrated ownership appear not to differ significantly.
BOARD COMPOSITION AND MONITORING
W E FIRST PROPOSE TO USE the ratio of outsiders to insiders as an indicator of the level of monitoring or the distribution of power in the firm. In
other words, we suggest that, knowing the determinants of that ratio, we can
infer what groups wish to monitor the managers. Equation (1) is based on the
hypothesis that (holding some control variables constant) the ratio is deter-
mined by the distribution of property rights.
PRELIMINARY TEST
WE FIRST ESTIMATE OVER ALL OUR OBSERVATIONS a regression equation similar
to the one presented in some contributions (probabilities in parentheses).
40
Ratio of outsiders/insiders = - 1.1080 + 0.2377 x Log. assets
(0.1933) (0.0003)
+ 0.0816 x Regulated firms
(0.7632)
- 0.0101 x Tenure of insiders
(0.5780)
- 1.3016 x Voting rights held by inside directors
(0.0026)
+ 2.9065 x Voting rights held by outside directors
(0.0024)
- 0.3655 x Stockholdings by non-members
(0.3647)
Adjusted R
2
 = 0.1179
(N = 255)
This result compares with those we have just quoted and suggests that our data
are probably not substantially different from those that have been collected in
the United States. The negative sign attached to the inside directors' voting
rights variable is consistent with the entrenchment hypothesis, while the
positive one for the outside directors' is consistent with the monitoring
hypothesis. But, surprisingly, the coefficient of the blockholdings by non-
172
E FIRST PROPOSE RO USR the ratio of outsiders to insiders as an indicator
ALTERNATIVE MECHANISMS FOR CORPORATE GOVERNANCE AND BOARD COMPOSITION
members is not significant: this is not consistent with that hypothesis. One
would expect the large shareholders, although they are not directly represented
on the board, to prefer a relatively high ratio of outsiders to insiders. However,
the equation is estimated across heterogeneous groups (i.e., across three
different types of ownership) and it may capture differences between groups
rather than differences between firms. One possible technique that would take this
effect into account is to partition the data according to the type of ownership.
OWNERSHIP AND THE RATIO OF OUTSIDERS TO INSIDERS
IN TABLE 6, WE PRESENT MEASURES of central tendency for the dependent and
independent variables to be included in the regression equations. There are
some significant differences between the three groups. The firms under diffuse
ownership do display, on average, a higher ratio of outside to inside directors.
However, they are also the largest firms, in terms of total assets, and include
the highest proportion of takeover targets. In these respects, the firms with
concentrated inside ownership tend to be located at the opposite end of the
frequency distribution. They have, on average, lower outsider to insider ratios:
companies under managerial control prefer inside directors.
In Table 7, we present a multivariate test of the model designed to
explain monitoring of the managers through board composition.
First, we note that the ratio of outsiders to insiders is significantly higher
for the regulated firms under diffuse ownership, regulation being represented
by a dummy variable. This, however, does not apply to either type of concen-
trated ownership. Such a combination of results suggest that this effect is not
attributable to the regulators. It is consistent with the hypothesis that the
regulated firms under diffuse ownership take the initiative in composing their
boards with a relatively large proportion of outsiders in response to a perceived
preference of their shareholders.
The distribution of voting rights is represented by three proxies: block-
holdings of non-members of the board, voting rights of the inside directors,
and voting rights of the outside directors. The important shareholders who are
not members of the board exert influence on board composition in the direction
of increasing the ratio of outsiders to insiders. This is the effect predicted by
the monitoring hypothesis. As for the inside directors, their blockholdings are
negatively correlated with the dependent variable for the subset of firms with
diffuse ownership. However, that regression coefficient is not significant under
concentrated ownership. These results suggest that the inside directors may
seek entrenchment under diffuse ownership, but have already achieved it
under concentrated inside ownership and cannot achieve it under concentrated
outside ownership. The voting rights held by the outside directors appear to
play a significant and positive role only under concentrated inside ownership.
Size, as represented by the logarithm of total assets, has an estimated
positive and significant regression coefficient, but only under concentrated
173
TABLE 6
MEAN VALUES OF DEPENDENT AND INDEPENDENT VARIABLES TO BE USED IN THE REGRESSION EQUATIONS OF TABLES 7,
[Outsiders / Insiders]
[Outsiders * Other Directorships]/
[Insiders * Other Directorships]
[Outsiders * Board Committees]/
[Insiders * Board Committees]
Ratio of Technical and Professional to
Total Personnel, for Industry
3
Regulated Firms (%)
b
Stockholdings Non-members (%)
c
Hostile Takeover Targets (%)"
Debt'to-Assets Ratio
Total Assets (1992) ($000)
Dividends over Book Value
of Common Equity
1
^
Excess Rate of Return on Assets
DIFFUSE
OWNERSHIP
(N=84)
1.8786
3.7701
2.3287
25.0083
19.0476
0.0339
32.1429
0.5553
1,779,946
0.0241
1.2533
CONCENTRATED
INSIDE
OWNERSHIP
(N=91)
1.2952
3.3801
1.6011
24.9879
18.6813
0.0496
14.2857
0.5522
690,703
0.1376
2.4838
CONCENTRATED
OUTSIDE
OWNERSHIP
(N=83)
1.6529
3.2797
1.7330
26.6277
19.2771
0.5705
13.2530
0.5198
1,065,061
0.0232
1.1094
8 AND 9
KRUSKAL-WALLIS TEST
X
2
10.471
0.765
3.994
2.909
0.010
187.770
12.010
1.476
8.653
0.392
1.824
PROB.
0.0053
0.6823
0.1392
0.2335
0.9949
0.0001
0.0025
0.4780
0.0132
0.8221
0.4018
ALTERNATIVE MECHANISMS FOR CORPORATE GOVERNANCE AND BOARD COMPOSITION
DEFINITIONS AND NOTES TO TABLE 6
Definitions Diffuse ownership - neither the directors nor the three main shareholders own 20% or more of
the voting rights.
Concentrated inside ownership - the directors own 20% or more of the voting rights.
Concentrated outside ownership - the three main shareholders own 20% or more
of the voting rights.
Outsiders - includes both the group representatives and the independent directors.
Insiders - includes employees and ex-employees of the firm and its subsidiaries and affiliated
outside directors.
Excess rate of return on assets - difference between the firm's average rate of return on assets and its
industry's, divided by the absolute valu e of the industry's average rate.
Notes:
 a
 As provided by Beck (1992).
" As a percentage of the number of firms.
c
 Held by the three largest shareholders non-members of the board, as a percentage of total voting
rights outstanding.
The median values for the three groups are 0.0262, 0.0263 and 0.0241, respectively.
inside ownership. This is consistent with the relationship summarized in
Figure 1. The firms classified under difruse ownership are the largest, on average,
and occupy the top flat section of the curve. To a lesser extent, this is also true
for the corporations with concentrated outside ownership, while those with
concentrated inside ownership are on the rising portion of the curve and still
at the stage of increasing their ratio(s) of outsiders to insiders.
The excess of the firm's average rate of return on assets to that of the
firm's industry appears to be significant and positive under diffuse ownership,
but not under either type of concentrated ownership. This is not consistent
with the hypothesis that outside directors have a discernible favourable effect
on performance, because we would then expect that regression coefficient to
be positive and significant under all three types of ownership. Rather, it is
consistent with the joint hypotheses that relatively profitable firms find it
easier to attract professional directors and that the latter represent some form
of bonding cost incurred under difruse ownership.
Finally, we note that our tests provide no evidence in favour of some
substitution between the proportion of outsiders on the board and financial
control devices such as the debt-to-assets and dividend ratios. Negative signs
for those coefficients would have been implied by the substitution hypothesis.
The ratio of technical and professional to total personnel (knowledge
ratio) designed as a proxy for complexity is not significant. This may be due to
measurement errors: the ratio is computed for the industry, not the firm. It is
also tenable that this effect may have been captured by the size variable.
Under concentrated outside ownership, the regression equation has no
explanatory power. This is consistent with the hypothesis that ownership is so
concentrated that board composition can be adapted precisely to the share-
holders' preferences and opportunities. Therefore, no empirically detectable
regularities exist.
41
175
GAGNON & ST-P1ERRE
TABLE 7
REGRESSION OF THE RATIO OF THE NUMBER OF OUTSIDERS
TO THE NUMBER OF INSIDERS
 a
INDEPENDENT
VARIABLES
Intercept
Ratio of Technical and
Professional to Total
Personnel, for Industry
Regulation: Dummy = 1 if
Regulated, = 0 Otherwise
Blockholdings,
Non-members of Board (%)
Voting Rights of
Inside Directors (%)
Voting Rights of
Outside Directors (%)
Target:
Dummy = 1 if Target
of Hostile Takeover,
= 0 Otherwise
Debt-to- Assets Ratio
Log (Total Assets)
Dividends over Book
Value of Common Equity
Excess Rate of
Return on Assets
R2
F-value
DIFFUSE
OWNERSHIP
(N=81)
-1.6663
(0.3106)
0.0211
(0.1768)
1.1413
(0.0326)
9.1052
(0.0040)
-12.0389
(0.0307)
-5.7211
(0.4076)
0.0881
(0.8278)
0.7210
(0.3383)
0.1511
(0.2423)
-0.0727
(0.9720)
0.3031
(0.0018)
0.3983
4.700
(0.0001)
CONCENTRATED
INSIDE
OWNERSHIP
(N=89)
-1.7815
(0.1333)
-0.0219
(0.0246)
-0.5579
(0.1397)
2.6494
(0.0522)
-0.5394
(0.3710)
4-3219
(0.0001)
0.2591
(0.4742)
0.7465
(0.2458)
0.2476
(0.0075)
-0.0789
(0.5210)
-0.0008
(0.9675)
0.4636
6.828
(0.0001)
CONCENTRATED
OUTSIDE
OWNERSHIP
(N=79)
4-4529
(0.0693)
-0.0141
(0.4359)
-0.2796
(0.6390)
-2.1819
(0.0704)
-2.5981
(0.5593)
-5.2642
(0.4910)
-0.5911
(0.3577)
-0.6783
(0.5760)
-0.0386
(0.8109)
-3.0291
(0.2114)
0.0087
(0.8475)
0.1078
0.834
(0.5978)
Notes:
 a
 Probabilities in parentheses.
For definitions, see Definitions and Notes to Table 6.
176
ALTERNATIVE MECHANISMS FOR CORPORATE GOVERNANCE AND BOARD COMPOSITION
Tentatively, the main lesson to be drawn from Table 7 is that, under
diffuse and concentrated inside ownership, board composition is determined
by the distribution of property rights as represented by blockholdings. As the
most interested parties persist in appointing directors of their own kind, they
must believe they are deriving net benefits from that course of action.
42
However, this does not necessarily imply that such perceived advantages can
induce a detectable positive correlation between the rate of return on assets
and board composition. We now turn to that question.
BOARD COMPOSITION AND PERFORMANCE
T
HE QUESTION OF INTEREST IS: can corporate performance, as measured by
the rate of return on assets, be explained partially by its governance
mechanisms in general and board composition in particular? To answer this
question, we use the excess rate of return on assets as our dependent variable.
ONE'EQUATION MODEL
WE FIRST HYPOTHESIZE THAT THE excess rate of return is associated statistically
with three factors: corporate governance mechanisms, financial variables and
a number of control variables. All the variables already appear in Table 1.
The first governance mechanism we consider is the distribution of
ownership. We expect large shareholders to be involved in monitoring activities
(in addition to influencing board composition) that result in improved financial
performance. On the other hand, managers or inside shareholders are expected
to capture private benefits that depress performance as seen by the residual
owners. In the same vein, the regulatory authorities are seen as monitors
whose allegiance is to the consumers, rather than to the owners, unless they
are captured by the industry. Whether or not regulation improves performance
is a moot point. The type of managerial compensation is expected to be a powerful
tool to improve performance because it is designed to tie the managers' interests
to the shareholders' interests. Board composition is the last element of the
governance mechanisms.
The second factor includes financial variables. The negative relationship
between the debt-to-assets ratio and the accounting rate of return is well
documented:
43
 the larger the earnings, the more likely the firm is to reduce its
debt. Therefore, we need the debt-to-assets ratio as a control variable and
expect a negative sign. However, the notion of indebtedness as a mechanism
for governance suggests that the managers left with smaller free cash flows are
forced to concentrate on capital expenditures whose net present value is positive.
This would imply a positive relation between the excess rate of return and the
debt-to-assets ratio. By similar reasoning, one also expects a positive relation
between the excess rate of return and the ratio of dividends to the book value
of common equity.
177
GAGNON & ST-PIERRE
TABLES
REGRESSION OF THE RATIO OF THE FIRM'S RATE OF RETURN ON ASSETS
 a
INDEPENDENT
VARIABLES
Intercept
Ratio of Technical and
Professional to total
Personnel, for Industry
Regulation: Dummy = 1 if
Regulated, = 0 Otherwise
Blockholdings,
Non-members of Board (%)
Voting Rights of
Inside Directors (%)
Target: Dummy = 1 if
Target of a Hostile
Takeover, = 0 Otherwise
Debt-to- Assets Ratio
Log (Total Assets)
Dividends Over Book
Value of Common Equity
Board Composition: Ratio
of Outsiders to Insiders
Managers' Compensation:
Ratio of Variable to Fixed
Compensation
R2
F-value
DIFFUSE
OWNERSHIP
(N=81)
-1.6282
(0.3671)
0.0032
(0.8649)
-0.0222
(0.9724)
-6.1296
(0.0943)
0.5631
(0.9323)
-0.1755
(0.7032)
-0.2254
(0.7967)
-0.0609
(0.6648)
0.3658
(0.8806)
0.3888
(0.0043)
0.0003
(0.1575)
0.1836
1.597
(0.1253)
CONCENTRATED
INSIDE
OWNERSHIP
(N=89)
9.4191
(0.1871)
0.0784
(0.1808)
2.9404
(0.1792)
-19.0639
(0.0183)
0.0355
(0.9913)
1.0836
(0.6157)
-1.2302
(0.7492)
-0.6819
(0.2529)
-0.3334
(0.6505)
0.4452
(0.4688)
-0.0046
(0.7748)
0.1228
1.105
(0.3686)
CONCENTRATED
OUTSIDE
OWNERSHIP
(N=79)
-3.9241
(0.5316)
0.0177
(0.7107)
-1.1286
(0.4829)
3.0404
(0.3395)
-6.7114
(0.6157)
-0.5079
(0.7703)
-1.7163
(0.5919)
0.2785
(0.5061)
13.4372
(0.0599)
0.0678
(0.8314)
0.0933
(0.1645)
0.1144
0.879
(0.5572)
Notes:
 a
 Probabilities in parentheses.
For definitions, see Definitions and Notes to Table 6.
178
ALTERNATIVE MECHANISMS FOR CORPORATE GOVERNANCE AND BOARD COMPOSITION
The third factor controls for operating conditions and complexity, as
represented by size and the knowledge ratio, and vulnerability to takeovers.
One expects the firms with lower accounting rates of return to be more likely
to become targets of hostile takeovers.
To summarize, we propose to test the following statistical model:
Excess rate of return = / (insiders'blockholdings, outsiders'block
holdings, type of managers' compensation,
board composition, regulation,
debt-to-assets ratio, dividend-to-book-
value ratio, size, knowledge ratio, and
takeover target) (2)
The results of this regression, by type of ownership, are presented in
Table 8.
As in some recent studies,
44
 this ordinary least squares estimate suggests a
significant relation between board composition and performance, albeit for
only the diffuse ownership case. This is consistent with the proposition that
the role played by the outside directors is a function of the type of ownership.
However, if the outsiders did explain performance, we would also expect board
composition to be significant under concentrated outside ownership. In fact,
we may not be observing a cause-and-effect relationship because ownership
and board composition may be endogenous to performance. That problem is
considered in Table 9.
SIMULTANEOUS-EQUATIONS MODEL
IN TABLE 9, WE ESTIMATE SIMULTANEOUSLY EQUATIONS (1) AND (2). Using the
two-stage least squares method, we estimate simultaneously the determinants
of board composition and of the excess rate of return. The results are as follows.
Again, we find that board composition, under diffuse ownership, is
determined by the distribution of voting rights: the ratio of outsiders to insiders
rises with the blockholdings of non-members and falls with those of the
directors. Such behaviour is consistent with the monitoring and entrench-
ment hypotheses, respectively. However, the excess rate of return on assets
variable becomes non-significant. Moreover, board composition also becomes
a non-significant explanatory variable of the return on assets. These new fig-
ures provide no reason to alter the tenor of our earlier conclusions. Our data
suggest that board composition is seen as important to the large shareholders,
who attempt to monitor the managers, but do not display any detectable
association between performance
45
 and board composition.
179
TABLE 9
SIMULTANEOUS (DOUBLE-STAGE LEAST SQUARES) REGRESSIONS OF THE RATIO OF OUTSIDERS TO INSIDERS AND THE
EXCESS RATE OF RETURN ON ASSETS
 a
DIFFUSE OWNERSHIP
(N=81)
INDEPENDENT VARIABLES
Intercept
Ratio of Technical and Professional
to Total Personnel, for Industry
Regulation: Dummy = 1 if Regulated, = 0 Otherwise
Blockholdings, Non-members of Boards (%)
Voting Rights of Inside Directors (%)
Voting Rights of Outside Directors (%)
Target: Dummy = 1 if Target of a
Hostile Takeover, = 0 Otherwise
Debt-to- Assets Ratio
RATIO OF
OUTSIDERS
TO INSIDERS
ON BOARD
-1.7024
(0.3356)
0.0171
(0.3395)
1.0012
(0.1012)
10.2764
(0.0078)
-10.7583
(0.0866)
-8.1186
(0.3286)
0.2050
(0.6640)
0.7851
(0.3360)
EXCESS RATE
OF RETURN
ON ASSETS
-2.9421
(0.7888)
0.0625
(0.6520)
3.1212
(0.6611)
11.7309
(0.7717)
-30.4106
(0.6654)
—
-0.2642
(0.8134)
1.3229
(0.7417)
CONCENTRATED INSIDE
OWNERSHIP (N=89)
RATIO OF
OUTSIDERS
TO INSIDERS
ON BOARD
3.9958
(0.7992)
0.0272
(0.8376)
0.9769
(0.8190)
-10.1524
(0.7613)
0.9208
(0.8404)
10.2022
(.5147)
0.9220
(0.6909)
0.6263
(0.8231)
EXCESS RATE
OF RETURN
ON ASSETS
10.0311
(0.1807)
0.1038
(0.1076)
3.2788
(0.1556)
-23.3915
(0.0104)
3.4416
(0.4238)
—
0.3377
(0.8847)
-2.2451
(0.5845)
CONCENTRATED OUTSIDE
OWNERSHIP (N=78)
RATIO OF
OUTSIDERS
TO INSIDERS
ON BOARD
4.4566
(0.1198)
-0.0143
(0.4435)
-0.2709
(0.6736)
-2.1846
(0.1225)
-2.5750
(0.5687)
-5.2470
(.4980)
-0.5869
(0.3658)
-0.6781
(0.5977)
EXCESS RATE
OF RETURN
ON ASSETS
-7.0268
(0.6704)
0.0279
(0.6921)
-0.8965
(0.6585)
4.7040
(0.5921)
-4.5890
(0.7921)
—
-0.0647
(0.9817)
-1.0792
(0.8134)
TABLE 9 (CONT'D)
DIFFUSE OWNERSHIP
(N=81)
INDEPENDENT VARIABLE
Log (Total Assets)
Dividends over Book Value of Common Equity
Excess Rate of Return on Assets
Board Composition: Ratio of Outsiders to Insiders
Managers' Compensation: Ratio of
Variable to Fixed Compensation
R2
F-value
RATIO OF
OUTSIDERS
TO INSIDERS
ON BOARD
0.1236
(0.3942)
-0.2274
(0.9192)
0.6154
(0.2133)
—
—
0.3182
3.313
(0.0014)
EXCESS RATE
OF RETURN
ON ASSETS
0.4104
(0.7079)
0.8971
(0.8800)
-2.0234
(0.7050)
0.0010
(0.5220)
0.0195
0.141
(0.9990)
CONCENTRATED INSIDE
OWNERSHIP (N=89)
RATIO OF
OUTSIDERS
TO INSIDERS
ON BOARD
-0.2525
(0.8510)
-0.3021
(0.6999)
-0.7181
(0.6973)
—
—
0.0455
0.377
(0.9533)
EXCESS RATE
OF RETURN
ON ASSETS
-0.9935
(0.1384)
-0.2171
(0.7796)
—
2.1401
(0.1403)
-0.0158
(0.4022)
0.1301
1.181
(0.3162)
CONCENTRATED OUTSIDE
OWNERSHIP (N=78)
RATIO OF
OUTSIDERS
TO INSIDERS
ON BOARD
-0.0388
(0.8454)
-3.1733
(0.4919)
0.0108
(0.9681)
—
—
0.1046
0.794
(0.6343)
EXCESS RATE
OF RETURN
ON ASSETS
0.2760
(0.5294)
15.8957
(0.2613)
—
0.8846
(0.8249)
0.0933
(0.1844)
0.1054
0.801
(0.6278)
Notes:
 a
 Probabilities in parentheses.
For Definitions, see Notes to Table 6.
GAGNON & ST-PIERRE
PUBLIC POLICY
W
HAT LESSONS CAN WE LEARN from this discussion? To determine that,
compare the objectives of the Canada Business Corporations Act
(CBCA) with the empirical evidence assembled here.
We take it that the objective of the Act is to create an environment
such that efficient governance mechanisms are available and corporate
resources are allocated to the best alternative use. As we have been unable to
uncover a strong association between overall performance and board composition,
we cannot suggest stringent regulation in that area.
46
 However, we do have
evidence that under diffuse ownership, both inside and outside blockholders
seek to secure representation on the boards of the companies in which they
have invested. That is evidence of a desire to monitor managers on the part of
those who are best motivated to do so efficiently. We suggest that lawmakers
might facilitate this in two ways.
First, we have observed that the apparent ratio of outside to inside directors
may differ markedly depending on the more-or-less-restrictive definition of an
outside director. The objective being that at least some members of the board
be truly independent of the managers, we suggest that the definitions and
regulations of the insurance legislation
47
 be also made part of the CBCA: that
affiliated outside directors not be counted as independent directors. Second,
our empirical work takes into account the distribution of voting, as opposed to
ownership, rights. Those two distributions differ to the extent that restricted
voting shares have been issued. In order to increase the owners' monitoring
ability, it would be desirable for cumulative voting to become mandatory when
some outstanding shares have restricted voting rights.
On a purely intuitive basis, it might be considered desirable that the pro-
portion of outside directors be increased, and that the directors' remuneration
be tied more closely to the value of the firm and to the directors' attendance at
board meetings. As trends in these directions have already been observed in
the United States,
48
 there is probably little risk and equally little gain in
putting them on the statute books. Finally, one effective but controversial rule
would be to require directors to have a non-negligible part of their personal
wealth invested in the firm they monitor. Such a measure would have to be based
on the a priori belief that self-interest provides powerful incentives. This would not
necessarily be seen as a feasible or even desirable piece of legislation, however.
CONCLUSIONS
T
HE EMPIRICAL RESULTS OF THIS STUDY can be summarized as follows. Our
data are consistent with the proposition that board composition is a function
of the distribution of voting rights. First, our simple statistical model partly
explains the differences in board composition under diffuse and concentrated
inside ownership, but not under concentrated outside ownership. Second, the
182
ALTERNATIVE MECHANISMS FOR CORPORATE GOVERNANCE AND BOARD COMPOSITION
influence of the blockholders who are not members of the board moves in the
direction of increasing the ratio of outsiders to insiders. This is consistent with
the hypothesis of a desire to monitor the managers. The influence of the block-
holders who are also inside directors moves in the opposite direction, suggesting
that they seek entrenchment.
These results also suggest that board composition is perceived as important
and that large shareholders attempt to monitor managers - but they do not
necessarily succeed in doing so efficiently. However, to the extent that it is not
costless, such behaviour would not have persisted unless it is of some benefit
to those who engage in it. In that sense, our data provide some indirect
evidence of a relationship between board composition and performance, but
we have been unable to uncover any direct evidence of that relationship.
Our statistical model does not explain board composition under concen-
trated outside ownership. The firms in our subsample behave as if the
combination of mechanisms for corporate governance were unique to each
one of them, or, paradoxically, as if they are unimportant.
Only under diffuse ownership is the ratio of outsiders to insiders higher
for the regulated firms. This is not consistent with the hypothesis of substitution
between regulation and board composition, but rather suggests that managers
seek the services of outside directors as a bonding expenditure to assure the
shareholders that they are bound not to expropriate them.
We find no evidence that debt and dividend policies are substitutes for
board composition, or that the latter is an effective takeover defence.
Overall, the evidence provided in this study suggests that both monitoring
and entrenchment forces are at work in this sample.
ENDNOTES
1 Title of an essay by Jensen & Warner (1988).
2 See Stangeland (n.d., Appendix) for a description of 11 anti-takeover
devices.
3 The behaviour of Inco Limited (1988) is a good example of the simultaneous
use of several governance mechanisms. In 1988, after an especially
profitable year, it adopted a shareholder rights plan (poison pill), paid a
special dividend of $10.00 (as opposed to a regular dividend of $0.20) and
almost doubled its long-term debt. However, this restored it only to its
former level, where it remained afterwards. The out-of-pocket cost of the
operation was $10 million.
4 Not all takeover defences have to be approved by the shareholders (see
Stangeland).
183
GAGNON & ST-PIERRE
5 If the plan calls for a special resolution, under the Canada Business
Corporations Act, it must be passed by a majority of not less than two-thirds
of the votes cast (Canada, 1985).
6 The best-known case in Canada is Bill C-131 drafted by the federal
government to prevent accumulation of shares of Canadian Pacific
Company.
7 Malatesta(1992,p. 636).
8 Halpern(1990).
9 In the Inco case, summarized in note 3, the Caisse de depot et placement
du Quebec, a large institutional investor, went to court to have the share-
holders cast two, rather than one, vote: one on the special dividend and
one on the poison pill. In at least one other case, the regulation was even-
tually diluted (Alcan). In another case (John Labatt) the resolution was
defeated by the shareholders.
10 For instance, the Canada Business Corporations Act requires corporations
with shares issued to the public to have at least two outside directors. (See
Canada 1985, par. 23-427.)
11 For instance, the Canada Bank Act sets to 10 percent the upper limit of the
proportion of voting rights that may be held by any given shareholder in a
Schedule A bank.
12 See Brickley etai. (1988), Pound (1992), Wahal etal. (1993).
13 Friend & Lang (1988).
14 Morck et al. (1988b) and McConnell & Servaes (1990).
15 The operational notion of outside director is discussed below.
16 Weisbach(1988).
17 Hermalin& Weisbach (1988).
18 Brickley & James (1987) for the banking, and Mayers et al. (n.d.) for
the insurance industries.
19 Kaplan & Reishus (1990).
20 Shivdasani(1993).
21 St-Pierre etal. (1994).
22 Beatty&Zajac(1994).
23 Hermalin & Weisbach (1991).
24 Barnhart et of. (1994).
25 Li (1994).
26 Thain & Leighton (1991).
27 For this section, the most useful papers have been: Baysinger & Butler
(1985), Brickley & James (1987); Weisbach (1988); Morck, Shleifer &
Vishny (1988b); Beatty & Zajac (1994); and Occasio (1994).
28 The notion of affiliated members has also been retained in laws such as
those legislating insurance companies. Teolis et al. (1992; section 172
relates to inside directors, who may represent no more than 15 percent of
the directors, and section 171 relates to affiliated persons, who may represent
no more than two-thirds).
184
ALTERNATIVE MECHANISMS FOR CORPORATE GOVERNANC E AND BOARD COMPOSITION
29 Groups are identified in Statistics Canada, Inter-Corporate Ownership, cat.
61-517.
30 In several studies, "grey" directors include all the board members who
could not be classified as inside or independent directors. Amoako-Adu &
Smith in their study of Canadian boards (n.d.) make use of the concept of
outside financial directors.
31 See, for instance, Cotter et al (1994), and Kaplan & Reishus (1990).
32 Jensen (1993) argues that the only inside board member should be the
CEO and that boards should not get beyond seven or eight people.
33 Compare note 10 with note 28, for instance.
34 Hermalin & Weisbach (1988), and Weisbach (1988).
35 Chenchuramaiah et al. (1994) have examined the inverse relationship
between debt and managerial ownership.
36 For evidence that firms may adopt that behaviour, see Pound (1992) and
Wahaletal.(1993).
37 We do mean to imply that debt-to-assets and dividend payout ratios are
independent of the distribution of ownership. Empirical studies suggest
that both are lower under concentrated inside ownership (Friend & Lang,
1988, and Eckbo &. Verma, 1994). For tax reasons, the payout ratio should
be higher when the main shareholder is a taxable Canadian corporation.
38 Note that these files cover all the firms that have sold securities to
Quebec investors and are not restricted to firms based in Quebec.
39 We use as the denominator the book value of the common stocks rather
than current earnings in order to remove the random component of the
latter. (Book value of common stock includes the cumulative value of
reinvested earnings. The ratio we use is an accounting estimate of dividend
return.)
40 See Weisbach (1988, p. 448) or Beatty & Zajac (1994, p. 328).
41 We also ran the regressions using the ratio of outsiders to insiders weighted
by the number of committees or directorships in other corporations as the
dependent variables. Under diffuse ownership, total assets and block-
holdings, but not regulation, remain as statistically significant explanatory
variables. These variables are not significant for the two groups with
concentrated ownership. As the models appear to have no explanatory
power, they are not presented here.
42 Not only do the main shareholders persist in nominating outsiders, panel
data suggest that the trend is increasing over time (SpencerStuart, 1993).
Admittedly, this evidence from the United States is relevant to the diffuse
ownership case.
43 Harris & Raviv (1991), Gagnon et al. (1987). Harris & Raviv argue that
there is no relation between debt and control.
44 See Li (1994).
45 This statement also applies to alternative measures of performance, such as
Tobin's Q, with which we have experimented.
185
GAGNON & ST-P1ERRE
46 For a similar conclusion, see Baysinger &. Butler (1985).
47 The relevant sections are referred to in note 28.
48 SpencerStuart(1993).
ACKNOWLEDGEMENTS
W
E HAVE RECEIVED MOST USEFUL COMMENTS from Randall Morck, Lee Gill
and an anonymous referee. However, they cannot be held responsible
for whatever shortcomings that remain.
BIBLIOGRAPHY
Amoako-Adu, B. and B. F. Smith. "Outside Financial Directors and Corporate
Performance: An Aspect of Corporate Governance." Manuscript, Wilfrid Laurier
University, undated.
Barnhart, S. W., M. W. Marr and S. Rosenstein. "Firm Performance and Board
Composition: Some New Evidence." Managerial and Decision Economics, 15 (1994):329-40.
Baysinger, B. D. and H. N. Butler. "Corporate Governance and the Board of Directors:
Performance Effects of Changes in Board Composition." The Journal of Law, Economics
and Organizations, 1,1 (Fall 1985):101-24.
Beatty, R. W. and E. J. Zajac. "Managerial Incentives, Monitoring and Risk Bearing: A
Study of Executive Compensation, Ownership, and Board Structure in Initial Public
Offerings." Administrative Science Quarterly, 39 (June 1994):313-35.
Beck, N. Shifting Gears: Thriving in the New Economy. Toronto: Harper Collins Publishers
Ltd., 1992.
Brickley, J. A. and C. M. James. "The Takeover Market, Corporate Board Composition,
and Ownership Structure: The Case of Banking." Journal of Law and Economics, 30, 1
(April 1987):161-80.
Brickley, J., R. Lease and C. W. Smith. "Ownership Structure and Voting on Antitakeover
Amendments." Journal of Financial Economics, 20 (1988):267-91.
Canada. Business Corporations Act with Regulations. Don Mills, ON: CCH Canadian
Limited, 1985 6th ed.
Chenchuramaiah, T. B., K. P. Moon and R. P. Rao. "Managerial Ownership, Debt Policy,
and the Impact of Institutional Holdings: An Agency Perspective." Financial
Management, 23, 3 (Autumn 1994):38-50.
Commission des valeurs mobilieres du Quebec. Bulletin de statistiques. Montreal, biannual.
Cotter, J. F., A. Shivdasani and M. Zenner. "The Effect of Board Composition and
Incentives on the Tender Offer Process." University of Iowa, March 1994.
Demsetz, H. and K. Lehn. "The Structure of Corporate Ownership: Causes and
Consequences." Journal of Political Economy, 93 (1985):! 155-77.
186
ALTERNATIVE MECHANISMS FOR CORPORATE GOVERNANC E AND BOARD COMPOSITIO N
Eckbo, B. E. and S. Verma. "Managerial Shareownership, Voting Power, and Cash
Dividend Policy." journal of Corporate Finance, 1, 1 (1994):33-62.
Fama, E. and M. C. Jensen. "Separation of Ownership and Control." Journal of Law and
Economics, 26 (June 1983):301-25.
Financial Post Directory of Directors, The Financial Post Information Service, Toronto,
updated annually.
Financial Post Survey of Industrials (on cards), The Financial Post Information Service,
Toronto, updated continually.
Friend, I. and L. H. P. Lang. "An Empirical Test of the Impact of Managerial Self-Interest
on Corporate Capital Structure." The Journal of Finance, 43, 2 (June 1988):271-81.
Gagnon, J-M., Suret, J-M. and J. St-Pierre. "Asymetrie de 1'information, fiscalite et endette-
ment au Canada." Finance, 8, 1 (1987):75-103.
Halpern, P. "Poison Pills: Whose Interests Do They Serve?" Canadian Investment Review, 3,
1 (Spring 1990):57-66.
Harris, M. and A. Raviv. "The Theory of Capital Structure." The journal of Finance, 46, 1
(March 1991):297-355.
Hermalin, B. E. and M. S. Weisbach."The Determinants of Board Composition." The Rand
Journal of Economics, 19, 4 (Winter 1988):589-606.
. "The Effects of Board Composition and Direct Incentives on Firm Performance."
Financial Management, 20, 4 (Winter 1991):101-12.
Inco Limited. "Notice of Special Meeting." Toronto, November 7, 1988.
Jensen, M. C. "The Modern Industrial Revolution, Exit and the Failure of Internal Control
Systems." The Journal of Finance, 48, 3 (July 1993): 831-80.
Jensen, M. C. and W. H. Meckling. "Theory of the Firm: Managerial Behavior, Agency
Costs and Ownership Structure." Journal o/Financial Economics, 3, 4 (1976):305-60.
Jensen, M. and J. B. Warner. "The Distribution of Power Among Corporate Managers,
Shareholders and Directors." Journal o/Financial Economics, 20 (1988):3-24.
Kaplan, S. N. and D. Reishus. "Outside Directorships and Corporate Performance." Journal
o/Financial Economics, 27 (1990):389-410.
Kofmel, K. G. (ed.). Who's Who in Canadian Business. Toronto: Trans-Canada Press,
Toronto, updated annually.
Leighton, D. and D. Thain. "The Role of the Corporate Director." Business Quarterly
(Autumn 1990):20-24.
Li, J. "Ownership Structure and Board Composition: A Multi-Country Test of Agency
Theory Predictions." Managerial and Decision Economics, 15 (1994):359-68.
Malatesta, P. "Takeover Defences," in The New Palgrave Dictionary of Money and Finance.
Edited by P. Newman, M. Milgate and J. Eatwell. New York: The Stockton Press, 1992,
Vol. Ill, pp. 633-36.
Mayers, D., A. Shivdasani and C. W. Smith. "Board Composition in the Life Insurance
Industry", undated manuscript.
McConnell, J. J. and H. Servaes. "Additional Evidence on Equity Ownership and
Corporate Value." Journal o/Financial Economics, 27 (1990):595-612.
Morck, R., A. Shleifer and R. W. Vishny. "Characteristics of Targets of Hostile and
Friendly Takeovers," in Corporate Takeovers: Causes and Consequences. Edited by A. J.
Auerback. Chicago: The University of Chicago Press, 1988a.
. "Management Ownership and Market Valuation, an Empirical Analysis." Journal
o/Financial Economics, 20 (1988b):293-315.
187
GAGNON & ST-PIERRE
Occasio, W. "Political Dynamics and the Circulation of Power: CEO Succession in U.S.
Industrial Corporations, 1960-1990." Administrative Science Quarterly, 39, (June 1994):
285-312.
Pound, J. "On the Motives for Choosing a Corporate Governance Structure: A Study of
Corporate Reaction to the Pennsylvania Takeover Law." Journal of Law, Economics and
Organization, 8, 3 (1992):656-72.
Rosenstein, S. and J. G. Wyatt. "Outside directors, Board Independence, and Shareholder
Wealth." Journal of Financial Economics, 26 (1990): 175-91.
Salant, R. "Board Expansions: Inside vs Outside Directors Firm Size and Market Reaction",
Manuscript, Mankato State University, March 1993.
Shivdasani, A. "Board Composition, Ownership Structure and Hostile Takeovers." Journal
of Accounting and Economics, 16 (1993):167-98.
SpencerStuart. SpencerStuart Board Index, 1993 Proxy Report. Board trends and practices at
100 major companies, Montreal, 1993.
St-Pierre, J., J-M. Gagnon and J. Saint-Pierre. "Takeover Bids, Structure of Ownership and
Board Composition," Working Paper #DAE-1994-02, Universite du Quebec a Trois-
Rivieres.
Stangeland, D. A. "Issues in Corporate Control and the Performance of Corporations",
Unpublished Ph.D. dissertation, University of Alberta, Edmonton, Alberta.
Statistics Canada. Inter-Corporate Ownership. Ottawa: Supply and Services Canada, cat. 61-
517, occasional.
Stulz, R. M. "Managerial Control of Voting Rights." Journal of Financial Economics, 20
(1988):25-54-
Teolis, J. W., J. S. Graham and J-P. Bernier. Financial Institutions Reform Package: Phase
Three, New Insurance Legislation Annotated. Don Mills: CCH Canadian Limited, 1992.
Thain, D. H. and D. S. R. Leighton. "Ownership Structure and The Board." Canadian
Investment Review, 4, 2 (Fall 1991):61-65.
Wahal, S., K. W. Wiles and M. Zenner. "Who Opts Out of State Antitakeover Protection."
University of North Carolina at Chapel Hill, October 1993.
Weisbach, M. S. "Outside Directors and CEO Turnover." Journal of Financial Economics, 20
(1988):431-60.
Williamson, O.E. "Corporate Finance and Corporate Governance." The Journal of Finance,
43,3(Julyl988):588-91.
188
Ramy Elitzur & Paul Halpern
Faculty of Management
University of Toronto
5
Executive Compensation and Firm Value
INTRODUCTION
T
HE NEW REGULATION REQUIRING REPORTING of dollar compensation for
CEOs has become a catalyst for discussions about the extravagance of
corporations and at some social gatherings a subject for lively conversation.
However, on its own, this development is not very useful in understanding the
role of compensation in corporate activity. The issue is usually framed in the
context of U.S. capital markets in which the widely held corporation is the
norm. In that scenario, the growth of U.S. corporations has led in many cases
to a separation of managers and owners and to difficulties in the effective
monitoring of managers by owners and their representatives, including boards
of directors. It is also argued that employees, executives, directors and owners
usually do not share the same goals. This conflict of interest occurs when the
interests of the managers of a firm (who are interested in their own compensation)
are not aligned with the interests of the owners (who are interested in the value
of their investment); this is referred to as the "agency problem". Compensation
systems in which managers' rewards are related to performance (which, in turn, is
of value to shareholders) can be used to ameliorate the agency problem. If
the system for executive compensation is not well designed, it will cause the
corporation to deviate from its shareholder value-maximization goals and subse-
quently reduce the value of the company. The debate in the U.S. literature does
not argue with the use of compensation incentives to elicit appropriate performance.
Rather, it questions whether executive compensation is sufficiently sensitive to
changes in shareholder wealth to elicit the correct behaviour by management.
The debate in the United States, and its framing in the context of the
widely held corporation, leaves a Canadian audience bemused because in
Canada, the widely held corporation is the exception rather than the rule.
Many have commented on the number of widely held companies in Canada.
A company is defined as widely held if there are no blocks of voting shares in
excess of 15 percent of the outstanding equity.
1
 In 1993 there were 90 companies
defined as widely held on the TSE 300 Composite Index.
189
ELITZUR & HALPERN
In fact, the typical Canadian company can be described as closely held.
At one extreme, the owner of a firm has a significant equity interest in the
company, either directly or indirectly. In some cases, the majority owner is a
member of the founding family of the company. In these circumstances the
owner may have direct managerial input as the CEO or indirect control
though his or her position on the board of directors. In either case, manage-
ment is entrenched. Although it could be argued that their large equity interest
will lead these individuals to maximize share value and that compensation
schemes to align interests are unnecessary, there is empirical evidence consistent
with entrenched management making decisions that will maximize personal
utility, not share value.
2
Another more serious form of entrenchment is the dual-class share
structure in which an individual or group of individuals, most frequently the
members of the founding family, have a majority of the voting (or superior
voting) shares and a small number of the non-voting (or restricted voting)
shares. This share structure usually arises when the founder wishes to cash out
a major portion of his or her position yet still maintain control. The resulting
structure, where the individual has a controlling position in the voting equity
but a small percentage of the overall equity of the firm, has the potential for
serious agency costs of equity and protection from the discipline of the
takeover market. The agency cost of equity arises when management can
undertake non-wealth-maximizing decisions for which they bear only a small
part of the cost through the effect on the share price.
Unlike the case of the widely held company, entrenched management
may have a very different view of executive compensation, its form as cash or
share-price related, and its role in generating wealth for investors. For example,
in a closely held company, in which management already has a large proportion
of the total equity, decisions affecting the value of the equity will be reflected
in personal wealth. Thus, compensation schemes that are based on share-price
performance may not be necessary; a scheme using more cash through salaries
and even less cash through bonus payments related to performance, could be
more beneficial to these individuals. Similarly, in the dual-class share structure,
owners may have a small stake in the overall equity of the firm. By providing
more compensation linked to equity, appropriate incentives may be provided
to maximize share price. However, since owners control the firm through their
control of the voting shares, there is no incentive for them to introduce these
equity-based schemes, since this would reduce their ability to gain through the
agency cost of equity based on their small holdings.
The importance of the relationship between compensation and firm
performance, as well as the amount of compensation to the CEO and top
management, recently became an important issue in Canada with the
amendment to the Ontario Securities Act governing disclosure of executive
compensation for Ontario issuers. The new disclosure requirements increase
both the breadth and the depth of the disclosure obligation. Under the new
190
EXECUTIVE COMPENSATION AND FIRM VALUE
disclosure rules, the issuer must disclose compensation in detailed form for the
CEO and the four most highly paid executives, in addition to the CEO. This
disclosure must include annual compensation (salary, bonus, and other forms
of short-term rewards) and long-term compensation (options, stock appreciation
rights, restricted shares, etc.). However, for a number of years, companies
interlisted on U.S. exchanges have been required to provide similar disclosure
on executive compensation. When the Multi-Jurisdiction Disclosure System
(MJDS) was implemented in 1990, Canadian interlisted companies could
choose to continue reporting under U.S. regulations or to use their home-
country documents, which had fewer onerous executive compensation disclosure
requirements.
3
 With the new Ontario regulations, those companies choosing
to use their home-country documents will now have generally to satisfy the
same disclosure requirements concerning executive compensation as found in
U.S. jurisdictions.
Since the change in Ontario is quite recent, there is limited scope to
undertake empirical research on the relationship between compensation and
firm-specific variables, and compensation and wealth creation. Based on our
analysis of cash compensation and firm-specific variables, however, our results
are similar to those found in other studies with respect to the influence of
company size but they differ significantly when the focus is on the influence
of firm-specific performance variables. Unlike other studies, we have broken
the sample into two sub-samples based on ownership concentration; a sample
of companies that are widely held and a sample of companies that are closely
held. Regardless of the ownership concentration, there is a strong positive
relationship between salary and total compensation and firm size. However,
contrary to findings in other studies, performance measured either by accounting-,
cash-flow- or market-based variables has no effect on the level of bonus, salary
or total compensation for either closely held or widely held firms. We also
investigated the determinants of the percentage change in compensation and
found no relationship to performance variables.
Compensation was positively influenced by the existence of a poison pill
variable and the interlisting status of the firm. There is some evidence that the
effect of these variables is stronger for the closely held sample. Finally, a persistent
difference in the relationship of closely held and widely held firms is the
importance of persistence in the percentage change in cash compensation. For
the closely held companies, the constant term is positive and significant for
both bonus and salary, whereas for the widely held firms, it is significant for
only the salary component.
In the first two sections of this study we review the effects of executive
compensation on the incentives of managers with particular emphasis on the
incentive effects associated with different compensation practices. The linkage
between the theory and testable hypotheses is presented in the next section.
Then we identify the actual compensation practices of the sample of
Canadian companies with respect to cash compensation - salary and bonus.
4
191
ELITZUR & HALPERN
The relationships among cash compensation, total and bonus and salary
components, and firm-specific characteristics are presented in the next section,
followed by our conclusions.
THE EFFECTS OF COMPENSATION PLANS ON
MANAGERS' INCENTIVES
T
HE LITERATURE IN THIS AREA has expanded greatly since the early 1980s,
coinciding with the growth in the takeover market and the recognition
that these transactions are a way to discipline management whose interests are
not aligned with those of shareholders.
5
 There are a number of studies that
identify a positive relationship between the introduction of short- and long-
term incentive-based compensation schemes and announcement period share-
price increases (Bhagat et al, 1985; Brickley et al, 1985; Larcker, 1983; and
Tehranian & Waegelin, 1986). The firm-specific performance measures used
are both accounting- and market-based.
Of particular interest for the present study is the relationship between
compensation and the financial performance of companies. The existence and
strength of this relationship provides information on the managerial incentives
in place to assist in the alignment of shareholder and manager interests. The
literature identifies two general questions related to compensation and
performance. The first considers the relationship of compensation to underlying
financial characteristics of the firm, with "compensation" defined as cash compen-
sation, including salary and bonus, plus deferred compensation, such as
options, (although many of the articles consider only the cash compensation).
The financial characteristics are related to size of company and performance,
measured from both accounting- and market-based perspectives. These papers
find that compensation is related to both size and market-based performance
of the equity of the firm. Some researchers (Murphy, 1984; Gibbons &
Murphy, 1990; and Kaplan, 1994) investigate compensation for upper level
executives of the firm, either the CEO alone or the top executives. Other
researchers, such as Leonard (1990) and Abowd (1990), consider a larger
group of managers, and different factors have been introduced to explain
executive compensation such as equity ownership by executives and the
existence of major blockholders, including institutional holdings (Mehran,
1995). In addition, if the definition of compensation is broadened sufficiently
to include top executive turnover, many studies have noted that turnover is
significantly related to return on assets, stock returns, and operating income,
and this relationship exists across a number of countries (Gibbons & Murphy,
1990; Kaplan, 1994; and Kang et al, 1995).
The second set of studies considers the effect of incentive compensation
schemes on management behaviour. At one extreme, the studies consider the
influence of incentive compensation on firm performance in subsequent periods.
192
EXECUTIVE COMPENSATION AND FIRM VALUE
If incentive compensation is intended to align the interests of shareholders
and managers, the performance of the firm should be related to incentives
introduced in the compensation scheme. The research to this point is consistent
with a positive relationship between incentive compensation and future firm
performance. Abowd (1990) identifies in a large sample of managers the
sensitivity of compensation to specific performance measures and determines
the relationship between this sensitivity and the future period performance of
the particular variable. He finds that accounting-based performance measures
yield weaker results than performance measures based on the ratio of cash flow
to replacement cost of assets and on stock rate of return. Mehran (1995) finds
that firm performance, measured as either Tobin's Q or return on assets, is
related to the percentage of compensation that is equity based and inversely to
cash compensation.
At the other extreme there are studies that examine the relationship
between incentive compensation plans and non-shareholder value maximizing
managerial behaviour to improve their wealth. Watts (1977) and Watts &
Zimmerman (1978) argue that bonus schemes create an incentive for managers
to select accounting policies that boost the present value of their awards.
Examples of this behaviour are noted by Dhaliwal et al (1982) with respect to
the choice of depreciation accounting, and by Hunt (1985) in the choice of
inventory valuation models when considering owner-managed firms compared
to non-owner-managed firms. Healy (1985) detects a strong association
between accruals and managers' income-based incentives under a bonus contract.
Kamin & Ronen (1978) report that owner-managers are less likely to smooth
income than non-owner-managers. Finally, Elitzur 6k Yaari (1994) suggest that
the choice of an executive incentive compensation plan by owners affects earnings
manipulation undertaken by managers.
Based on the literature on incentive compensation, such plans should
meet the following objectives.
• Enhance goal congruence between shareholders and executives,
leading to the enhancement of shareholder wealth.
• Reinforce informative reporting, i.e, ameliorate earnings manipu-
lation.
• Strike a balance between the long-range strategy and the short-
term goals of the firm.
The pay instruments commonly used in the design of incentive compen-
sation plans are as follows:
• Cash Bonuses These take four forms: profit sharing, profit sharing
with a hurdle, target plans, and target plans with a threshold.
193
ELITZUR & HALPERN
Under a profit-sharing plan the bonus is a percentage of divisional
profit after a deduction of a capital charge. Sometimes the cash
bonus is calculated as a percentage of divisional profit in excess
of the budgeted income, denoted as profit sharing with a hurdle.
Under a target plan the executive is paid cash as a function of
pre-set targets; under a target plan with a threshold, the cash is
paid only in a certain interval of the performance measure.
• Stock Options A stock option is a right to buy a number of
shares in the company at a given price at some future period.
• Stock Purchase Plans A very popular form of incentive com-
pensation in which executives can buy shares in the company at
a discount.
• Phantom Shares Sometimes the company awards executives
shares for bookkeeping purposes only, i.e., the executives do not
actually own shares, but for the purpose of calculation of incen-
tive payments they are viewed as if they do. At the end of a
specified period of time the executive is paid on the basis of
stock performance. This payment can be in cash, in shares, or a
combination of both.
• Performance Shares Shares are awarded to executives when
specific long-term goals have been attained. The goals can be
either corporate, divisional, or individual. Performance shares are
rarely observed in Canada.
• Performance Unit Plan An arrangement similar to performance
shares, except that, on achievement of given targets, the executive
receives specially valued units and not shares. Consequently, this
manner of compensation has the same advantages and drawbacks
as performance shares.
• Formula Value Stock Plan Under this alternative, the executive
receives some shares that are not traded publicly, the value of
which is calculated according to a formula. This formula can be
based on accounting variables or other long-term measures of
performance. In this way, a formula value stock plan very much
resembles performance shares or a performance unit plan.
• Restricted Stock In this case, the executive receives shares in
the company at a discount or, sometimes, at no cost. The shares
awarded cannot be transferred until certain conditions are met.
194
EXECUTIVE COMPENSATION AND FIRM VALUE
When the conditions are met, the restrictions are lifted and
executives can do with the shares whatever they please. (For
example, an executive can transfer these shares only after a
specific period of continuous employment with the company.)
Restricted stocks are not common in Canada because they are
not tax effective (for either the company or the executive), and
they are restricted by the Toronto Stock Exchange (TSE).
• Stock Appreciation Rights (SARs) With SARs, the executive is
paid based on the appreciation of a specified number of shares.
SARs have characteristics similar to phantom shares.
The attributes of these incentive-based compensation instruments are summarized
in Table 1.
6
THE THEORY OF EXECUTIVE INCENTIVE
COMPENSATION DESIGN
T
HE DESIGN OF THE OPTIMAL incentive compensation scheme stems from
the Principal-Agent model (see Appendix 1 for a description of the theory
behind optimal compensation schemes).
Under this model a principal hires an agent to run the firm. The principal's
payoff is a function of some outcome less the cost of compensating the agent.
The outcome and, in turn, the principal's payoff, depend on the agent's effort
and some random state of nature (which can describe, for example, exogenous
economic conditions). According to the model, the observation of both out-
come and the state of nature is not enough to reveal how much effort the
agent has expended.
In the model the agent's payoff is derived from compensation less the
disutility of effort. In this sense, effort is costly from the agent's viewpoint because
it involves the opportunity cost due to the sacrifice of leisure time and so forth.
The choice of effort is made in a manner that optimizes the agent's payoff.
The principal decision on the optimal compensation scheme must take
into account both the agent's choice of effort, as described in the paragraph
above, and the motivation of the agent to stay with the firm.
The model implies that optimal compensation should follow this formula:
Compensation = Fixed Compensation + Variable Compensation
The fixed component of compensation should depend on the agent's reservation
payoff, i.e., how much the agent can receive elsewhere. The variable component
of compensation should be non-decreasing in outcome. If the agent is risk-
neutral the variable component of compensation is a constant multiplied by
outcome, i.e., compensation becomes a linear equation.
195
TABLE 1
ATTRIBUTES OF VARIOUS PAY INSTRUMENTS
PAY FOR
INSTRUMENT PERFORMANCE
Cash Bonus +
Stock Options +/-
Stock Purchase Plan -
Phantom Shares +/—
Performance Shares + +
Performance Unit Plan + +
Formula Value Stock Plan ++
Restricted Stock +/-
Stock Appreciation Rights +/-
Note: ++ Effective
+ Somewhat Effective
0 Unclear
/ In-between
- Weak
ALIGNING INTERESTS
INFORMATIVE OF EXECUTIVES WITH
REPORTING SHAREHOLDERS
— +
+/- +
+ +
+/- +
0 ++
0 ++
0 ++
0 +
+/- +
LEAD TO CASH OUTFLOW
SHORT-TERM FOR THE DILUTION TAX
ORIENTATION COMPANY OF EQUITY EFFECTIVENESS
— — ++ —
++ ++ — +
++ ++ — +
++ — ++ —
++ + + 0
++ + + 0
++ + + 0
++ + - -
++ - ++ -
EXECUTIVE COMPENSATION AND FIRM VALUE
LINKING THE THEORY OF EXECUTIVE COMPENSATION
DESIGN TO TESTABLE HYPOTHESES
T
HE THEORY OF EXECUTIVE COMPENSATION DESIGN provides some interesting
insights. First, it implies that there should be a positive correlation
between compensation and outcome. Hence, it follows that compensation
should increase with performance. Second, the theory implies that optimal
compensation in the setting above involves a constant, w, which relates to the
manager's reservation utility.
While quite elegant and providing many intuitive implications for
compensation, the theory is remarkably silent on the definition of compensation
and how performance is to be measured. Considering compensation, the
possibilities include salary, cash bonus, and equity awards as described in the
previous section. Mehran (1995) presents evidence that is consistent with the
board of directors taking into consideration executives' total incentives in
designing pay packages. Thus, executives who have large holdings of the equity
of the company that they manage (more frequently the case in Canada than in
the United States), will likely have less equity-based compensation and more
cash compensation. Therefore, all forms of compensation should be used. By
omitting equity-based compensation, an observation that cash compensation
is unrelated to performance does not lead to the conclusion that compensation
is unrelated to firm performance, since the influence may be found in the
omitted variable. However, data restrictions are often particularly important
in determining the measure of compensation actually used. In this study we
use cash compensation, in total and separately for bonus and salary, and do not
use option-related compensation (due to data problems).
There are a great many definitions of performance, based on both market
and accounting information. The measures can be based on various forms of
reported income (net income, operating income and so forth), sales, assets,
cash flow (operating cash flow, for example) in either levels, changes, or ratios.
Also, stock market returns can be used. In this study we use both accounting-
and market-based performance measures, since one of our purposes is to
identify the variables that appear to be relevant in the Canadian context. The
choice of variables used in this study is conditioned by the variables used in
the literature.
According to the theory of optimal incentive contracts, w is related to a
manager's reservation utility. We would therefore expect that heads of large
companies make a high marginal contribution to production. Their ascent to
the top echelons of the firm as successful winners of an internal corporate
labour-market competition, suggests that top executives will demonstrate
clearly the talents necessary to direct a large corporation and that these scarce
talents are best utilized by having them near the top of the organization. The
talent of senior managers is magnified by spreading their responsibility and
control over long chains of command and scales of operation.
197
ELITZUR & HALPERN
The existence of this scarce factor suggests a high reservation utility to
top executives and with this high opportunity cost comes a high wage. Also,
since the success of a top executive can be readily observed in the performance
of the company, it is possible that competing companies may be willing to pay
the executive based on the company's current performance. Thus, not only
should cash (salary) compensation be related to the size of the organization
but also there may be a standard to which the reservation payoff, w, is related.
This standard could be related to whether or not the company is
interlisted on U.S. markets. Since we control for industry and size, any effect
of interlisting could reflect the added responsibility of managing a company
which trades securities in the United States. Alternatively, managers (and
their compensation consultants) may use U.S. firms as their comparison group
and thus have higher salaries on average. The interlisting effect is measured
by using a dummy variable, which takes on the value of unity if the company
is interlisted over the period of consideration and zero otherwise.
We also control for other exogenous factors that relate to the existence
of entrenchment. The first variable identifies whether the company had a
poison pill over the period of interest and takes the value of unity if a poison
pill is in existence and zero otherwise. Typically, a poison pill is introduced
into firms in which shares are widely held and management ownership in the
equity of the firm is small. The result of the poison pill, although open to
debate (Comment & Schwert, 1993), is to provide management with the
power to decide on the success of a takeover bid, thus leading to potential
entrenchment. In our sample, a number of firms that meet the closely held
definition also have poison pills. This reflects the fact that, although technically
closely held, the ownership structure does not eliminate the threat of a takeover.
In terms of compensation, there can be offsetting effects. If the poison pill is not
an entrenchment device and managers' interests are to be aligned with share-
holders' interests, the low levels of managerial equity ownership suggest that
total compensation would be composed of greater equity linked instruments. If,
on the other hand, the pill does result in entrenchment, management can
increase its salary and bonus without undue concern to the incentive effects.
This influence would lead to a positive influence in the relationship of cash
compensation and poison pills. Unfortunately, without the equity linked
compensation, we cannot test for the first implication of the poison pill.
The second entrenchment influence is the dual-class share structure, and
this is identified as a dummy variable with the value of unity if dual-class
shares are present. This variable is found in the closely held sample. The
typical company with dual-class shares is one in which the founder retains
some equity ownership through voting shares, but need not be active in the
management of the firm. If the founder/controlling shareholder, along with
other members of the family, is in the top management team, there may be
no need to provide incentives through equity-based compensation since
management already has equity ownership, although this conclusion will depend
/98
EXECUTIVE COMPENSATION AND FIRM VALUE
upon the size of management's equity position. However, holders of superior
class voting shares may not want to increase their equity exposure. Also, with
entrenchment, the cash part of compensation will be higher. The result would
be an increase in cash compensation in the presence of dual-class shares.
EXECUTIVE COMPENSATION IN PRACTICE
T
HIS SECTION CONSIDERS the current compensation practices of Canadian
companies based on a sample of 180 companies on the TSE 300 Index. The
period of analysis relates to the companies' fiscal year which occurred between
January 1993 and March 1994- The description of the database and the data
sources are presented below; the characteristics of the compensation information
follow; and the section ends with the regression results and analysis.
DATA
UNDER THE 1993 REVISION to the Ontario Securities Act (OSA), companies
with securities traded in Ontario must complete Form 40, "Statement of
Executive Compensation", which requires the company to report the compen-
sation for the CEO and for each of the company's four executive officers (other
than the CEO) who receive the highest compensation. The only exception to
the reporting requirement is for an executive officer whose cash compensation,
both salary and bonus, does not exceed $100,000. The company must supply
information on the salary, bonus and long-term and deferred compensation for
each of the company's last three financial years. The compensation covered
includes options, stock appreciation rights and long-term incentive plans.
In addition, the executive compensation report requires filing the five-
year rate of return on the common equity of the company, with dividends
reinvested and the rate of return on a broad index, with dividends included,
over the same period.
Finally, and probably most interesting, there is the requirement under
Item ix that the policies used by the compensation committee to determine
the compensation of executive officers must be described for the most recent
financial year. The report must include a discussion of the specific relationship
of corporate performance to executive compensation. Also, if an award was
made to an executive officer whose compensation must be reported under a
performance-based plan, despite failure to meet the relevant performance
criteria, the company must disclose the bases for the decision to waive or
adjust the relevant performance criteria.
While these disclosures identify a company's rationale for compensation
payments, they are not required for any compensation decisions prior to
January 1, 1994- This covers the bulk of our sample of 180 companies. Also,
there is some evidence that these disclosures are not very helpful. The same
disclosure requirements must be met by companies listed in the United States.
199
ELITZUR & HALPERN
Form 40 is included in the proxy material that is released approximately
three months after the fiscal year end. The basic data was provided to us by
KPMG who entered the numerical contents of the proxy statements they
received. In addition we obtained balance-sheet and income-statement data
from the Financial Post files and share-price and dividend data from both the
Toronto Stock Exchange Review and the Financial Post files.
We calculated the average value of compensation for the top five
executives for each company in the sample for the two most recent financial
years. If the company did not report compensation for five executives we
removed it from the sample. The companies covered 14 industries, but one
industry was removed since it had only one company. Consequently, we were
left with 180 companies in 13 industries.
Our data set includes cash compensation, salary and bonus, and does not
at this time include the deferred compensation and long-term incentive plans.
As noted above there are some problems with the quality of the reported
deferred compensation numbers, but as this data is improved it will be used in
future research. Certainly, the omission of this part of executive compensation
results in an incomplete picture of a company's compensation philosophy and
of the tradeoffs that may be made between cash and deferred compensation.
The use of the cash component of compensation is not unique to this study (see
Abowd, 1990; Gibbons & Murphy, 1990; and Leonard, 1990, among others).
In order to take into consideration the unique characteristics of the
Canadian market, we break the sample into two parts: companies that are
widely held and all others. To be widely held, a company should not have any
blocks of shares owned that are greater than 15 percent of the outstanding
equity. In the case of dual-class shares, we look at the concentration of holdings
of the voting shares. Using this definition we obtain 131 companies that were
not widely held and 49 that were widely held. The former group is called
"closely held" in our study, but there need not be a control position in place.
Table 2 presents some comparative statistics for the two groups of companies.
Table 2 shows that, as a percentage of the number of companies in each
group, widely held companies are interlisted and have poison pills more often
than the closely held companies. In addition, except in one case, dual-class
share structures are found in closely held companies. In our sample, of the 131
companies in the closely held sample, 24 percent had dual-class share structures.
There was no difference in the existence of negative earnings in the prior fiscal
year, which was 24 percent of the sample for both groups.
COMPENSATION CHARACTERISTICS
IN TABLE 3, PANEL A, WE SHOW SUMMARY STATISTICS on the levels and changes
in cash compensation by bonus and salary and the sum of the two elements for
the total sample, the closely held and the widely held sub-samples separately.
Considering levels first, it is clear that for the total sample, cash compensation
200
EXECUTIVE COMPENSATION AND FIRM VALUE
TABLE 2
CHARACTERISTICS OF CLOSELY HELD AND WIDELY HELD COMPANIES
CLOSELY HELD WIDELY HELD
(131 COMPANIES) (49 COMPANIES)
NUMBER PERCENTAGE NUMBER PERCENTAGE
Negative Earnings
Interlisted
Poison Pill
Dual-Class Shares
31 24 12 24
36 27 21 72
8 6 10 2 0
3 1 24 1 2
through salary is higher and has less variability than compensation through
bonus. The average value for salaries is $258,000 whereas the average value for
bonus is $87,000. We also observe that the distribution for bonus is more highly
skewed than the distribution for salaries. These observations are consistent
with the use of salaries to reflect reservation utility and bonus to be related in
some way to performance, although salaries may be related to longer-term
performance. We also compare the two sub-samples and observe that the
means and standard deviations for all compensation categories are greater for
the closely held sample than for the widely held sample. Also, the distributions
for all compensation forms are more skewed for the closely held companies.
The changes in the components of cash compensation are presented in
Panel B of Table 3 and show a consistent pattern. While the average value of
the changes in bonus and salary for the total sample is approximately the same,
roughly $19,000, the variability in the distribution of changes is much larger for
the bonus than for the salary categories. Also, the cash bonus distribution is
negatively skewed, whereas for the change in salaries it is positively skewed.
These observations are again consistent with the bonus reflecting some element
of performance whereas salaries appear to be less sensitive to this effect.
Looking at the two sub-samples, there are some interesting differences
between the changes for the closely held and widely held sub-samples. For the
former, the pattern of means, standard deviations, and skew is very similar to
the total sample results. However, the widely held sample is different. The
mean change for bonus is less than the mean change for salary. The standard
deviation for change in bonus is greater than the standard deviation of the
change in salary just as in the closely held sample, but the values of the
standard deviations for both compensation categories are smaller. Also, there
is positive skewness in the change in the bonus, unlike the case of the closely
held company in which the skew is negative.
201
ELITZUR & HALPERN
TABLE 3
SUMMARY STATISTICS FOR CASH COMPENSATION BY COMPONENT
FOR TOTAL SAMPLE AND Two SUB-SAMPLES
PANEL A - LEVELS
TOTAL SAMPLE
Bonus
Salary
Total Compensation
CLOSELY HELD
Bonus
Salary
Total Compensation
WIDELY HELD
Bonus
Salary
Total Compensation
MEAN
(000)
86.7
258.4
345.1
93.3
261.9
355.2
69.1
249.3
318.4
STANDARD
DEVIATION
(000)
143.0
129.9
220.0
158.0
138.0
236.0
90.7
106.0
170.4
SKEW
4.98
2.21
2.72
4.86
2.34
2.77
2.08
0.96
1.494
MINIMUM
(000)
0
100.0
107.5
0
100.0
107.5
0
109.9
109.9
MAXIMUM
(000)
1,312.0
956.1
1,541-0
1,312.0
956.1
1,541-0
388.9
516.9
886.5
PANEL B - CHANGES
TOTAL SAMPLE
Bonus
Salary
Total Compensation
CLOSELY HELD
Bonus
Salary
Total Compensation
WIDELY HELD
Bonus
Salary
Total Compensation
18.9
19.0
37.9
22.2
21.0
43.2
10.1
13.7
23.9
103.7
34.6
112.1
110.0
39.0
119.0
84-8
15.1
89.2
-0.20
6.86
0.34
-0.44
6.26
0.14
1.03
1.10
1.19
-484.5
-16.8
-465.0
-484.5
-16.8
-465.5
-270.8
-13.1
-239.6
531.0
376.0
551.5
531.0
376.0
551.5
360.1
61.7
389.3
Table 4 presents the means and standard deviations of the cash compensation
elements by industry class. The industry classification with the largest number of
companies in our sample is Oil and Gas with 39 firms, and the classifications with
the smallest number of firms are Transportation and Environmental Services
and Real Estate, with three firms each. For every industry classification, the
average salary in the industry was greater than the average bonus value. The
ratio of average salary to bonus ranged from a high value of 8.8 times for the
202
EXECUTIVE COMPENSATION AND FIRM VALUE
TABLE 4
SUMMARY STATISTICS FOR CASH COMPENSATION BY INDUSTRY CATEGORY
INDUSTRY
(TSE 300 CATEGORY)
MEAN COMPENSATION
(STANDARD DEVIATION IN BRACKETS)
NUMBER
OF FIRMS
1 . Metals & Minerals
2. Gold & Silver
3. Oil & Gas
4. Paper & Forest Products
5. Consumer Products
6. Industrial Products
7. Real Estate
8. Transportation &
Environmental Services
9. Utilities
10. Communications & Media
11. Merchandising
12. Financial Services
13. Conglomerates
7
13
39
13
16
28
3
3
14
18
12
18
6
BONUS
($000)
75.2
(111.2)
80.9
(103.8)
39.4
(47.6)
73.5
(97.8)
118.0
(215.1)
75.0
(79.1)
33.1
(39.9)
82.4
(35.3)
84.1
(59.6)
102.9
(54.9)
75.1
(73.4)
170.8
(315.1)
204.6
(236.6)
SALARY
($000)
259.2
(99.1)
225.8
(110.0)
191.4
(67.7)
240.3
(97.8)
258.3
(150.8)
250.7
(86.5)
289.8
(86.4)
255.8
(33.6)
214.8
(84.7)
452.1
(277.7)
320.1
(153.4)
302.7
(114.2)
413.4
(165.0)
RATIO TOTAL
(Salary/Bonus) ($000)
3.4 334.4
(165.6)
2.8 306.7
(163.4)
4.9 230.8
(94.1)
3.3 313.8
(97.8)
2.2 376.3
(356.3)
3.3 325.7
(123.6)
8.8 322.9
(125.5)
3.1 338.1
(62.4)
2.6 298.9
(138.8)
4.4 555.0
(305.2)
4.3 395.2
(164.9)
1.8 473.5
(328.6)
2.0 618.0
(345.4)
Real Estate industry to a low of 1.8 times for the Financial Services industry.
Since Real Estate has few observations, the next highest ratio is 4.9 times for
the Oil and Gas industry. This ratio could depend upon a number of factors,
including the viability of the industry (e.g., Real Estate) and the extent to
which there are other forms of performance-related compensation such as
203
ELITZUR & HALPERN
SARs, and options and top management ownership of equity securities, which
would result in a low ratio. In this study we have not controlled for these factors.
Average total cash compensation by industry ranges from a low of
$230,000 in the Oil and Gas Industry to a high of $618,000 in the Conglomerate
segment. Next we investigate the relationship of compensation to size and
performance variables.
RELATIONSHIP OF COMPENSATION AND FIRM-SPECIFIC VARIABLES
C
OMPENSATION SHOULD BE RELATED to many factors, such as responsibility,
size of firm and company performance, among others. As we have noted,
managerial talent and productivity and the ability to manage a large organi-
zation are scarce factors which should receive compensation related to the size
of the organization. This compensation is typically in the form of salaries.
However, top executives should also be compensated through incentive
schemes, since their decisions have an effect throughout the organization.
This implies that cash bonuses will be related not to size but to company per-
formance. This relationship is confused somewhat when the ownership
structure of the firm is taken into consideration. For example, when a firm is
widely held the normal relationship between performance and compensation
is as described. However, if the firm is closely held and the owner/manager has
substantial personal holdings in the equity of the firm, the salary compensation
may be higher to reflect the fact that incentives based on share-price perfor-
mance are not needed and may, in fact, increase the owner/manager's exposure
to stock price variability, thereby generating more risk-averse behaviour.
We now consider the relationship of cash compensation, measured as
bonus, salary and total compensation, with both performance and firm-size
variables, along with other firm-specific variables including entrenchment
variables such as the existence of poison pills and dual-class share structures.
Since this is the first analysis of Canadian data, we will relate our observations
to those observed in studies using data from other countries.
COMPENSATION AND FIRM SIZE
GIVEN THE DIFFERENCES IN OWNERSHIP STRUCTURE found in the sample of firms,
the analysis will consider closely held and widely held firms separately; the
results of the widely held sample will be most relevant in comparisons with
studies undertaken on U.S. data.
Table 5 presents the results of regressing the components of compensation
and total compensation on corporate size variables, measured by assets and
sales and other firm-specific variables, both for widely held and closely held
samples. In the regressions all dollar variables are measured as the natural
logarithm of the specific variable. With this specification, the regression
coefficient on the size variable is interpreted as an elasticity measure where an
204
EXECUTIVE COMPENSATION AND FIRM VALUE
increase of, say, ten percent in the size variable is associated with a specific
percentage increase in compensation.
We also consider the influence of additional firm-specific variables on
cash compensation. If the bonus is related to performance, we would expect
the size of the bonus to be related to the existence of negative earnings
measured as net income during the fiscal year. This variable is measured as a
dummy variable, taking the value 1 if net income is negative and zero otherwise.
There are 12 companies in the widely held sample and 31 in the closely held
sample with negative net income in the fiscal year. If salaries reflect opportunity
cost of managers or long-term (not short-term) performance, there should be
no significant relationship with this variable. We also include dummy variables
to identify interlisting status, the existence of a poison pill, and the existence of
dual class shares in the closely held company sample. The dummy variables take
the value of 1 if the company has the particular characteristic and zero otherwise.
A number of studies have investigated the relationship between
compensation and size and the results are summarized in Rosen (1990) and in
Milgrom & Roberts (1992). These studies find that the elasticity for total cash
compensation ranges in the small interval .2 to .25, using sales as the size
variable. This result is sufficiently robust over countries and different samples
that the result has been called a compensation "constant". In Table 5, Panel
A, the elasticity of total compensation for widely held companies is .164 when
size is measured by sales and .161 when size is measured by assets. This result is
slightly lower than noted in other studies and is statistically significant. Thus
for a 10 percent increase in size in our sample, measured either as sales or
assets, there is an increase of approximately 1.6 percent in total compensation.
For the closely held sample in Panel B, the elasticity measure is higher at .19
for sales and . 17 for assets.
We also consider the effect of firm size on the components of total cash
compensation. If, as noted above, the salary component is related to the
opportunity cost of managing large firms, then the coefficient on the size vari-
able for salary compensation should be positive. Considering the bonus, if
there is a performance-based element, then the relationship to firm size may
still be positive - a successful firm based on shareholder value criteria may
have growth in sales and/or assets — but it is likely to be an indirect and hence
less significant effect. In columns 2 and 5 of Panel A in Table 5, we observe
that for widely held companies, salary is positively and significantly related to
firm size with a coefficient of approximately .14. Notice that the constant term
is statistically significant, suggesting that there is a fixed element in the salary
relationship which is independent of size effects. The observed relationship
with bonus found in rows (1) and (4) while positive, is not statistically signifi-
cant using either assets or sales as the measure of firm size. For the closely held
firms in Panel B, the elasticity of salary with respect to firm size is greater than
that observed in the widely held sample; the values range between .17 and .18.
The constant term remains positive and statistically significant. For the bonus
205
ELITZUR &. HALPERN
TABLE 5
COMPENSATION RELATED TO SIZE AND OTHER FIRM-SPECIFIC
Constant
Sales
Assets
Negative
Earnings
Interlisted
Poison Pill
R2
Constant
Sales
Assets
Negative
Earnings
Interlisted
Poison Pill
R2
(1)
BONUS
2.370
(0.550)
0.462
(1.425)
—
-3.1 19
b
(-1.974)
1.111
(0.814)
1.472
(0.880)
0.046
3.783
(1.124)
0.424
C
(1.705)
—
-1.376
(-1.560)
0.026
(0.031)
2.293
(1.489)
0.031
PANEL A
(2)
SALARY
10.485
a
(39.827)
0.138
3
(6.984)
—
0.077
(0.802)
0.056
(0.667)
0.000
(0.001)
0.521
PANEL B -
9.92 l
a
(36.957)
0.181
3
(9.145)
—
0.059
(0.833)
0.133
h
(1.991)
0.031
(0.251)
0.386
VARIABLES
- WIDELY HELD COMPANIES
(3)
TOTAL
COMPENSATION
10.363
3
(32.046)
0.1 64
a
(6.745)
—
-0.027
(-0.230)
0.106
(1.039)
-0.023
(-0.182)
0.479
(4)
BONUS
7.686
(1.633)
—
0.047
(0.147)
-2.619
C
(-1.653)
0.866
(0.623)
1.781
(1.050)
0.003
(5)
SALARY
10.204
3
(40.220)
—
0.145
a
(8.354)
0.140
(1.639)
0.039
(0.518)
0.082
(0.900)
0.610
(6)
TOTAL
COMPENSATION
10.199
3
(29.739)
—
0.161
a
(6.829)
0.055
(0.475)
0.081
(0.804)
0.076
(0.611)
0.486
CLOSELY HELD COMPANIES
10.098
3
(29.557)
0.1 89
a
(7.479)
—
-0.076
(-0.845)
0.165
h
(1.937)
0.146
(0.932)
0.300
4.840
(1.259)
—
0.334
(1.212)
-1.488
C
(-1.682)
-0.048
(-0.057)
2.271
(1.464)
0.021
10.003
a
(30.122)
—
0.1 69
a
(7.128)
0.012
(0.151)
0.104
(1.433)
0.033
(0.243)
0.271
10.280
3
(24.710)
—
0.1 69
a
(5.688)
-0.125
(-1.304)
0.134
(1-472)
0.145
(0.862)
0.196
Notes:
 a
 Statistically significant at 1% level.
Statistically significant at 5% level.
c
 Statistically significant at 10% level.
206
b
EXECUTIVE COMPENSATION AND FIRM VALUE
relationship, the elasticity measures for sales (while approximately the same as
in the widely held firm) are significant at the 10 percent level. However, using
assets, the coefficient relationship is not statistically significant. A statistical
test could not reject the hypothesis that the overall relationship for compen-
sation for the closely held and widely held samples were the same.
7
The regressions also present information on the effect of other firm-
specific information on the level of compensation. For the widely held sample,
neither the interlisted dummy variable nor the existence of a poison pill has
an effect on the level of cash compensation measured as salary, bonus, and the
sum of the two. As expected, the existence of negative earnings has a
statistically significant negative effect on the bonus relationship but no effect
on salary or total compensation when size is measured by sales. When size is
measured by assets, the coefficient of the negative earnings variable is negative
but only marginally significant in the bonus relationship.
In Panel B of Table 5 the relationship for the closely held firms presents
a somewhat different picture. When size is measured by sales, the interlisted
dummy variable is positive and significant for salaries; thus, closely held firms
which are interlisted have higher salaries; this relationship is also found in
total cash compensation but not for the bonus. The result for salaries in the
closely held company is consistent with the opportunity cost argument and
perhaps the added skills necessary to manage a firm which has shares listed on
U.S. exchanges. However, this latter argument is muted since the interlisted
variable is not significant for the widely held shares. The relationship for the
interlisted variable (when size is measured by assets) while positive for salaries
and total compensation, is not significant. For bonus compensation, while the
relationship with the presence of negative earnings is negative as found in the
widely held sample, it is marginally significant. Finally, entrenchment as
measured by the poison pill variable has a positive influence, as observed in
the widely held sample, but it is not significant. The interpretation of this last
variable is difficult for the closely held sample because it includes companies
in which the ownership of the voting shares is greater than 15 percent. Thus
there will be some companies for which control is precarious and the poison
pill is an entrenchment tool. In other firms there may be dual-class shares
which act as a very effective entrenchment device and a poison pill is not
needed. The effect of dual-class shares on the relationship of compensation
and firm-specific variables is considered in the following section.
Therefore, for the widely held sample we observe bonuses to be related
to the existence of negative earnings and salaries to be influenced only by size
and having a significant fixed element. For the closely held sample, there is
evidence that salaries, along with a significant fixed element, are influenced
by size and interlisted status and bonuses are marginally negatively affected by
the existence of negative earnings.
As noted above, the closely held sample includes companies with very
different ownership characteristics, but we do not know the actual equity
207
EL1TZUR & HALPERN
ownership of the CEO and top executives. However, we do know whether the
firm had a dual-class share structure. This share structure typically can induce
two types of behaviour. First, the major shareholder owns a large proportion of
the voting shares but not of the total equity since the voting shares are typically
a small proportion of the total equity. This can result in agency costs of equity
since owner/managers bear only a small proportion of the cost of opportunistic,
non-shareholder wealth maximization. Second, the dual class shares are an
effective anti-takeover device that could result in poor decision-making and
compensation payments unrelated to performance. Further, it could also lead
to poor performance not followed by disciplinary action through the takeover
market. To investigate the effect of dual-class shares, we have re-estimated
the regression for the closely held sample, introducing a dummy variable with
a value of unity if the company has dual-class shares. The regression results are
found in Table 6 where the results for firm size measured by sales are presented.
When size is measured by assets the results are unchanged in terms of signs
and significance of the variables.
Considering salaries first, we observe that both interlisted status and
dual-class share structures have a positive and significant effect. The result for
the dual-class variable is consistent with the entrenchment/agency cost
argument presented earlier. A counter argument to this conclusion is that in
closely held companies the owner/manager already has a large equity interest,
and cash compensation (whether in bonus or in salary) is needed more than
performance-related compensation. While this may be true, with dual-class
shares, it is not necessarily the case that the owner/manager owns a large portion
of the total equity. Thus, the observation of higher salary in dual-class shares
still appears to be an entrenchment problem. Looking at the bonus relation-
ship, dual-class shares do not have a significant effect and the poison pill
variable is positive and almost significant at the 10 percent level. If the poison
pill is introduced in those companies in which owners do not have a large
ownership position or dual-class shares, then the observation of higher bonus
compensation is consistent with an entrenchment story.
Therefore, it appears that for these relationships, unlike the situation in
widely held companies, salary and bonus compensation are both related to
some entrenchment variables, either poison pill or the existence of dual-class
share structures. This is different from the widely held companies where the
poison pill variable is not significant and where negative earnings have a
significantly negative effect on bonus compensation.
The relationships identified to this point reflect cross-sectional dispersion
across firms but do not take into consideration the effect of changes in variables
over time. The compensation decision may be framed in terms of changes in
compensation in relation to certain firm-specific variables. We introduce this
aspect by considering the effect on the percentage change in cash compensation,
measured as the change in the natural logarithm of these variables over two
adjacent fiscal year ends, based on changes in firm size and the other firm-
208
EXECUTIVE COMPENSATION AND FIRM VALUE
TABLE 6
THE RELATIONSHIPS BETWEEN COMPENSATION AND SALES, FIRM-SPECIFIC VARIABLES
AND DUAL-CLASS VARIABLES IN THE CLOSELY HELD SAMPLE
CLOSELY HELD COMPANIES
Constant
Sales
Negative Earnings
Intel-listed
Poison Pill
Dual Class Shares
Adjusted R
2
BONUS
3.673
(1.090)
0.4 14
C
(1.663)
-1.344
(-1.522)
0.122
(0.144)
2.490
(1.600)
0.819
(.930)
.030
SALARIES
9.899
a
(37.514)
0.1 79
a
(9.200)
0.065
(0.940)
0.1 52
b
(2.298)
0.070
(0.573)
0.1627
b
(2.358)
.407
TOTAL COMPENSATION
10.074
3
(29.857)
0.187
a
(7.482)
-0.068
(-0.773)
0.1865k
(2.202)
0.190
(1.219)
0.1 84
b
(2.085)
.318
Notes:
 a
 Statistically significant at 1% level.
Statistically significant at 5% level.
c
 Statistically significant at 10% level.
specific variables identified in the previous section. The change in firm size is
measured using the same formula as the change in compensation. Table 6 presents
the relationships of change in cash compensation to change in firm size and
other firm-specific variables. Panel A shows the results for widely held firms
and Panel B shows the results for closely held firms. Based on the observed
performance of the dual-class variable in Table 6, we show the set of regressions
for closely held firms including the dual-class variable so that the results in
Panel B of Table 7 are not exactly comparable with those in Panel A.
However, the introduction of the dual-class variable does not have an important
effect on the sign or significance of the regression coefficients of the other
firm-specific variables in the closely held sample.
8
For widely held companies (Panel A) with the size variable measured by
change in sales, the only significant variables for the change in bonus compen-
sation are the existence of negative earnings and the poison pill; both vari-
ables are positive and significant at the 10 percent level. The negative earn-
ings result is difficult to interpret since it suggests that the change in bonus is
209
ELITZUR & HALPERN
TABLE 7
CHANGES IN COMPENSATION RELATED TO CHANGES IN SIZE AND
OTHER FIRM-SPECIFIC VARIABLES
PANEL A - WIDELY HELD COMPANIES
CHANGE IN
Constant
Change in
Sales
Change in
Assets
Negative
Earnings
Interlisted
Poison Pill
Adjusted R
2
(1)
BONUS
-0.720
(-0.702)
1.733
(0.896)
—
2.752
C
(1.650)
-0.385
(-0.280)
2.866
C
(1.707)
0.042
(2)
SALARY
0.048
a
(2.845)
-0.004
(-0.121)
—
-0.002
(-0.061)
0.049
b
(2.168)
-0.027
(-0.981)
0.024
(3)
TOTAL
COMPENSATION
0.019
(0.419)
0.051
(0.597)
—
0.095
(1.240)
0.026
(0.426)
0.088
(1.202)
-0.007
(4)
BONUS
-0.474
(-0.429)
—
0.293
(0.123)
2.279
(1.368)
-0.509
(-0.364)
2.932
C
(1.711)
0.024
(5)
SALARY
0.033
C
(1.861)
—
0.069
C
(1.833)
0.016
(0.603)
0.044
b
(1.967)
-0.020
(-0.727)
0.093
(6)
TOTAL
COMPENSATION
0.000
(0.016)
—
0.124
(1.215)
0.104
(1.458)
0.012
(0.198)
0.103
(1.403)
0.018
greater for companies in which there are negative earnings. The poison pill
variable is consistent with the entrenchment argument. Considering change
in salaries, the interlisted variable is positive and significant suggesting that
changes in salaries are greater if the firm is interlisted. This may be a result of
compensation practices in the United States or the difficulty of dealing with
shareholders in both countries. Finally, the constant term for change in salary
is positive and statistically significant. The interpretation of this result is that
there is some persistence in salary compensation. When we use assets to measure
size, the results are similar. The persistence effect is observed for salaries only
and the percentage change in salaries is positively and significantly related to
the percentage increase in assets over the fiscal year. This result is consistent
with the span of control argument in which the increasing asset size suggests
more responsibility and hence more compensation. Note that increases in
sales can occur without increases in assets, and thus the results for the two size
variables can be different. Finally, there are no significant variables in the
change in total compensation relationships.
210
EXECUTIVE COMPENSATION AN D FIRM VALUE
TABLE 7 (CONT'D)
PANEL B - CLOSELY HELD COMPANIES
CHANGE IN
Constant
Change in
Sales
Change in
Assets
Negative
Earnings
Interlisted
Poison Pill
Dual Class
Adjusted R
2
(1)
BONUS
1.177
b
(2.259)
0.332
(0.254)
—
-1.264
(-1.581)
-1.266
C
(-1.679)
2.867
b
(2.075)
0.298
(0.381)
0.048
(2)
SALARY
0.090
a
(4.641)
0.071
(1.455)
—
-0.017
(-0.556)
-0.002
(-0.079)
-0.019
(-0.362)
-0.033
(-1.145)
-0.007
(3)
TOTAL
COMPENSATION
0.152
3
(4.382)
0.095
(1.091)
—
-0.086
(-1.630)
-0.069
(-1.378)
0.206
b
(2.249)
-0.032
(-0.621)
0.061
(4)
BONUS
1.332
a
(2.780)
—
-1.493
(-1.492)
-1.084
b
(-2.060)
-1.058
(-1-402)
3.111
b
(2.255)
0.369
(0.474)
0.064
(5)
SALARY
0.099
3
(5.471)
—
.032
(0.853)
-0.017
(-0.548)
-0.002
(-0.080)
-0.024
(-0.461)
-0.035
(-1.207)
-0.018
(6)
TOTAL
COMPENSATION
0.1 59
a
(4.986)
—
0.108
(1.619)
-0.070
(-1.295)
-0.077
(-1.538)
0.1 88
b
(2.051)
-0.038
(-0.738)
0.071
Notes:
 a
 Statistically significant at 1% level.
Statistically significant at 5% level.
c
 Statistically significant at 10% level.
Panel B of Table 7 presents the closely held sample results, including a
dual-class share variable. Using the change in sales variable, for the change in
bonus compensation relationship, the constant and poison pill variables are
significant. The negative earnings and interlisted status variables have a negative
coefficient and the latter is marginally significant. When assets are used, the
negative earnings variable has a negative and significant coefficient and the
poison pill remains positive and significant. Therefore, the bonus relationship
shows a persistence effect, an entrenchment effect through the poison pill
relationship, and a performance effect in the wrong direction observed in the
negative earnings variable. When salaries are considered, regardless of whether
assets or sales are used, the constant is the only significant variable. (Recall
that for the widely held sample, the interlisted variable has a positive and
211
ELITZUR &. HALPERN
significant effect as does change in assets.) Finally, for total compensation, in
the closely held sample the two significant variables are the constant and
the poison pill, and the change in assets is almost significant at the 10 percent
level. However, it may be incorrect to infer some relationships from total
compensation since we observe that the significance of this variable in total
compensation comes from the bonus relationship.
There are some important observations in these results. First, in both the
widely held and closely held samples there is an entrenchment effect noted in
the bonus relationship through the significance of the poison pill variable.
Thus, the change in bonus payments is higher for firms that have poison pill
plans in place. Second, in both samples there is a persistence effect in the
change in salary relationship. There are some important differences in the two
samples, as well. First, closely held firms have a persistence effect in the
change in bonus. Thus, managers in these firms will have a bias toward
positive changes in bonus payments compared to managers in widely held
firms. Second, there is no effect on salary of interlisted status in the closely
held firms, although this relationship does exist in the widely held sample.
Third, the closely held firm sample displays a significant relationship between
change in bonus and the existence of negative earnings. Unlike the result for
the widely held sample, the performance relationship is in the correct direction;
the presence of negative earnings is related to a smaller change in the bonus.
COMPENSATION AND PERFORMANCE
IF COMPENSATION CONTRACTS are designed to motivate executive officers to
make decisions consistent with shareholder value, then the performance variables
on which compensation is based should be related to shareholder wealth. We
consider two sets of performance measures: accounting-based and market-
based. The former relates to conventional financial ratios measured over the
most recent fiscal year. The accounting ratios include return on equity defined
as net income divided by book equity, and return on assets defined as operating
income divided by the book value of total assets; all variables are measured at
the end of the fiscal year. These variables do not reflect actual cash flows, so we
also included the ratio of cash flows from operations divided by the book value
of total assets. The market-based performance measure is the rate of return on
the common equity over the fiscal year and is measured as the natural
logarithm of the value relative of the equity security including dividends.
Compensation by component and its total value is measured first by the
log of the compensation value and second by the change in the logarithm.
Many studies have considered the relationship of levels of compensation and
performance. By and large the semi-elasticity of compensation with accounting
performance measures is positive. As noted by Rosen (1990), the semi-elasticity
is estimated to be about 1.0 to 1.25 and statistically significant. These result
hold in the United States as well as in Britain. Other studies estimate the
2/2
EXECUTIVE COMPENSATION AND FIRM VALUE
TABLE 8
INFLUENCE OF MARKET-BASED PERFORMANCE ON LEVELS OF COMPENSATION
WIDELY HELD
TOTAL
BONUS SALARY COMPENSATION BONUS
Constant 10.158
b
 10.284
3
(1.946) (36.221)
Assets -0.072 0.1 42
a
(-0.212) (7.643)
Stock -1.687 -0.054
Return (-1.084) (-0.643)
Negative -2.279 0.151
C
Earnings (-1.413) (1.722)
Intel-listed 0.928 0.041
(0.669) (0.541)
Poison Pill 1.952 0.088
(1.148) (0.950)
Dual Class — —
Adjusted R
2
 0.007 0.604
10.314
3
(26.892)
0.155
3
(6.201)
-0.078
(-0.686)
0.071
(0.597)
0.084
(0.827)
0.084
(0.669)
—
0.480
4.715
(1.201)
0.343
(1.239)
-0.813
(-0.776)
-1.621
C
(-1.773)
0.099
(0.116)
2.681
C
(1.693)
0.990
(1-114)
0.019
CLOSELY HELD
SALARY
9.925a
(30.322)
0.1 73
a
(7.493)
-0.104
(-1.189)
-0.001
(-0.012)
0.136
C
(1.915)
0.112
(0.846)
0.232
a
(3.128)
0.319
TOTAL
COMPENSATION
10.118
a
(24.329)
0.1 75
a
(5.975)
0.027
(0.246)
-0.108
(-1.115)
0.1 64
C
(1.816)
0.202
(1.207)
0.249
a
(2.648)
0.227
Notes:
 a
 Statistically significant at 1% level.
Statistically significant at 5% level.
c
 Statistically significant at 10% level.
semi-elasticity with respect to market rate of return and find the values positive
and range between .10 and .16. The implication of these findings is that
compensation is based on performance, measured either by accounting returns
or market returns. The influence of accounting performance measures should
not surprise anyone, even if they can be manipulated by managers, since
accounting information is readily available, typically audited, and does provide
some information on the performance of the firm and ultimately on the
performance of the share price. It would be comforting to those who believe
that incentive compensation should maximize shareholder value if accounting
performance measures were related to cash flow.
Table 8 presents the results of the regressions of the level of compensation
on market performance, measured as the annual rate of return on equity
including reinvestment of dividends, and other firm-specific variables for both
widely held and closely held companies. The size variable used is assets since
213
ELITZUR & HALPERN
this provides slightly higher adjusted R
2
 values, but the size, sign and significance
of the individual regression coefficients are not affected dramatically if sales
are used instead. For widely held companies, the only significant variable in
the bonus relationship is the constant; the stock return performance variable
is negative and not significant. The salary and total compensation relation-
ships also do not show any relationship to share-price performance; the only
significant variables are the constant and the size of the firm and, for the salary
component, an anomalous positive relationship of negative earnings and
salary. For the closely held sample, compensation is also not related to any
form of performance. As observed in similar regressions in previous sections,
the dual-class and interlisted variables are positive and significant for salaries
and total compensation and the poison pill is positive and significant for the
bonus relationship.
9
However, instead of levels, the change in compensation may be related to
firm performance over the period. To investigate this possibility, the relationship
of the percentage change in cash compensation, bonus, and total is related to
the percentage change in sales, the stock market performance variable, and
other firm-specific variables. The results are presented in Table 9 for both
widely held and closely held companies. Looking first at the widely held sample,
for the percentage change in bonus relationship the poison pill variable is positive
and almost significant at the 10 percent level, and the stock return perfor-
mance variable has an insignificant influence on bonus as does the existence
of negative earnings. The poison pill relationship is consistent with
entrenchment by existing management. For the change in salary variable, the
interlisted variable is positive, reflecting a greater responsibility and hence
higher salaries. Also, the stock-return variable has a positive and significant
effect. This is somewhat puzzling since the salary and the relationship should
be related to long-term and not to short-term performance; the one-year stock
return reflects the latter.
Considering the closely held companies, performance does not affect
either the bonus or the salary relationship. For the bonus relationship, interlisting
status is negative and marginally significant and the poison pill, just as in the
widely held sample, is positive but is now statistically significant. The negative
interlisting effect is puzzling and may reflect the fact that interlisted companies
were reporting executive compensation before the requirement was imposed in
Ontario and this may have had a constraining effect on the bonus payments. For
the salary equation, only the constant is positive and significant. The dual-class
variable has no significant effect in any of the regressions. Finally, stock market
performance has a positive and statistically significant effect on total compen-
sation. This strong and positive effect is surprising, given the poor performance
observed in the bonus and salary regressions and the weak influence of stock-
market performance in the widely held sample.
The major difference between the widely held and closely held firms is
the significance of the constant for the closely held sample. This reflects a
214
EXECUTIVE COMPENSATION AN D FIRM VALUE
TABLE 9
CHANGE IN COMPENSATION RELATED TO CHANGE IN SIZE,
PERFORMANCE, AND FIRM-SPECIFIC VARIABLES
WIDELY HELD
BONUS
Constant -0.986
(-0.802)
Sales 1.717
(0.878)
Stock 0.591
Return (.401)
Negative 2.655
Earnings (1.561)
Interlisted -0.426
(-0.300)
Poison Pill 2.810
(1.652)
Dual Class —
Adjusted R
2
 0.023
SALARY
0.030
(1.528)
-0.005
(-0.159)
0.040
(1.706)
-0.088
(-0.304)
0.047
(2.084)
-0.031
(-1.140)
—
0.065
TOTAL
COMPENSATION
-0.026
(-0.487)
0.048
(0.574)
0.099
(1.569)
0.074
(1.026)
0.019
(0.317)
0.079
(1.088)
—
0.025
BONUS
0.956
(1.762)
0.220
(0.168)
0.831
(0.899)
-1.110
(-1.358)
-1.323
(-1.772)
2.626
(1.906)
0.335
(0.430)
0.053
STOCK PRICE
CLOSELY HELD
SALARY
0.078
(3.697)
0.070
(1.436)
0.007
(0.215)
-0.014
(-0.441)
0.002
(0.055)
-0.012
(-0.231)
-0.036
(-1.218)
-0.017
TOTAL
COMPENSATION
0.073
(2.044)
0.069
(0.829)
0.206
(3.494)
-0.044
(-0.841)
-0.071
(-1.485)
0.173
(1.970)
-0.046
(-0.934)
0.151
Notes:
 a
 Statistically significant at 1% level.
" Statistically significant at 5% level.
c
 Statistically significant at 10% level.
persistence element that was observed in previous regressions using change
in compensation. Also, the poison pill effect is stronger in the closely held
sample, while the interlisted variable gives different signs in the two samples
- positive and significant in the salary equation and negative and marginall y
significant in the bonus equation for the closely held sample.
Since accounting numbers are so pervasive, perhaps the limited success
we observe in relating levels and/or changes in compensation to market-based
performance will be improved by introducing accounting performance measures
since compensation committees may actually use these numbers. Table 10
summarizes the relationships of compensation to accounting performance
variables. In estimating the relationships, we have removed the variable
identifying the existence of negative income since it would likely be highly
215
EL1TZUR & HALPERN
TABLE 10
RELATIONSHIP OF COMPENSATION TO ACCOUNTING-BASED PERFORMANCE VARIABLES
COMPENSATION
ACCOUNTING
PERFORMANCE
VARIABLE
WIDELY HELD
BONUS SALARY TOTAL
CLOSELY HELD
BONUS SALARY TOTAL
Return on Equity
Return on Assets
> 0
n.s.
>0
significant
< 0
n.s.
< 0
n.s.
>0
n.s.
>0
n.s.
> 0
n.s.
> 0
n.s.
< 0
significant
< 0
significant
< 0
n.s.
< 0
n.s.
Cash Flow
Divided by Assets
< 0
n.s.
< 0
significant
< 0
significant
< 0
n.s.
< 0
significant
< 0
n.s.
Notes: Accounting performance is measured by Return on Equity, Return on Assets and Cash Row Divided
by Assets. The entries identify the sign of the regression coefficient and its significance in a relationship
of compensation to accounting performance. Size variable in regression is Sales.
n.s. = not significant
significant = statistical significance at least at 5% level
marginal = statistical significance at 10% level
correlated with the accounting-based performance variable. The accounting-
based performance variables include the return on equity and return on assets,
both of which reflect accounting profits and not cash flows. The third perfor-
mance variable attempts to remedy this problem by using cash flow divided by
assets. For the widely held sample, only return on asset for bonus is significant
and positive. For the closely held sample, both accounting-based performance
variables are significant only for salaries but the sign is in the wrong direction.
Although not reported in the table, when assets are used as the size variable
the coefficients on the salary relationships remain negative but become
insignificant. This result is consistent with that found for the annual stock
return. Note that although not presented, the sign and significance of the
dual-class, poison pill and interlisted variables remain unchanged under all
specifications of the performance variables and generally similar to the results
found when performance is measured by rate of return on equity in the market.
For salaries, cash flow divided by assets is significant in both samples, but the
sign is in the wrong direction.
Finally, a similar analysis is undertaken using change in cash compensation
and performance variables; the results are presented in Table 11. In the closely
held sample, there is no significant relationship of compensation in any of its
forms with accounting performance. For the widely held sample, the return on
equity has a negative and significant effect on the change in bonus and total
compensation. This is counter to our expectations of a positive sign. The other
216
EXECUTIVE COMPENSATION AND FIRM VALUE
TABLE 11
RELATIONSHIP OF CHANGE IN COMPENSATION TO
ACCOUNTING-BASED PERFORMANCE VARIABLES
CHANGE IN COMPENSATION
ACCOUNTING
PERFORMANCE
VARIABLE
WIDELY HELD
BONUS SALARY
TOTAL
COMPENSATION BONUS
CLOSELY HELD
TOTAL
SALARY COMPENSATION
Return on Equity
Return on Assets
Cash Flow
Divided by Assets
< 0
significant
< 0
n.s.
<0
< 0
n.s.
> 0
n.s.
<0
< 0
significant
< 0
n.s.
<0
n.s. marginal n.s.
>0
n.s.
>0
n.s.
>0
n.s.
>0
n.s.
>0
n.s.
> 0
n.s.
> 0
n.s.
> 0
n.s.
> 0
n.s.
Notes: Accounting performance is measured by Return on Equity, Return on Assets and Cash Flow Divided
by Assets. The entries identify the sign of the regression coefficient and its significance in a relationship
of change in compensation to accounting performance. Size variable in regression is Sales.
n.s. = not significant
significant = statistical significance at least at 5% level
marginal = statistical significance at 10% level
performance variables, although also having a negative sign (except return on
assets for salaries), provide no significant relationships. For total compensation,
return on equity is negative and significant.
Although not shown in Table 11, we have investigated the effect of
other variables on the compensation relationship. For the widely held sample,
regardless of the accounting performance variable, the bonus equation displays
a positive and significant effect for the poison pill variable. Using change in
salary, the constant and the interlisting variables are positive and significant.
These results are somewhat different from those found in the closely held sample.
Here, regardless of the accounting-based performance measure, the constant is
positive and significant, interlisted status is negative and significant, and the
poison pill remains positive and significant. For the change in salary, the constant
is the only significant variable. The results using the performance-based variables
suggest that there is an entrenchment effect through the poison pill influence
on bonuses. The differences between closely and widely held companies arise
in the influence of the interlisted variable in the bonus relationship where it is
insignificant for the widely held companies and negative and significant for
the closely held companies. Also, the interlisted variable is positive and signif-
icant for the salary relationship only in the widely held sample.
217
EL1TZUR &. HALPERN
CONCLUSIONS
T
HE PURPOSE OF THIS STUDY HAS BEEN TO IDENTIFY the factors that are
important in the determination of executive compensation in Canada and
to ascertain whether these variables are consistent with shareholder wealth
maximization. Also, we are interested in determining whether the ownership
structure and direct entrenchment variables such as poison pills and dual-class
shares have any effect on compensation.
The results of our analysis are not very encouraging when the effect of
firm performance on compensation is considered. When performance is
measured by stock-market return, regardless of the ownership structure, there
is no relationship between it and the level of compensation either in terms of
components or in total. With changes in compensation, salaries are positively
related to stock return for the widely held (sample) and total compensation in
the closely held companies. The results using the accounting-based performance
variables are confusing. Regardless of the ownership structure of the firms and
the form of compensation, most of the performance variables are insignificant
and, where they are significant, they have the wrong sign. When performance
is considered as the presence of negative earnings, the effect is to have lower
bonus compensation for both widely held and closely held companies. Thus,
there appears to be some performance-related effect on the bonus payments
through the negative earnings influence.
Finally, we consider the effect of entrenchment variables on compensation.
Regardless of the ownership structure, the poison pill variable has no effect on
the salary equation. However, it is positive and significant for the bonus relation-
ship for both widely held and closely held companies when both the change in
bonus and (in some instances) total compensation are considered. Thus
entrenchment seems to have a positive effect on the change in bonus. The
dual-class share variables are significant only in the levels equation for salary
and total compensation. Therefore, there appears to be some entrenchment
behaviour in compensation for both widely held and closely held companies.
The final difference is the importance of the constant term in the relation-
ships, especially in the changes relationships. For the levels, while the salary
relationship always has a positive and significant constant term, it is most likely
that for closely held companies, the bonus relationship tends to have a positive
and significant constant term. Therefore, they have a bias toward positive
compensation. This observation, however, is probably consistent with the high
ownership proportions of managers and their desire to obtain compensation
through salaries and bonuses which are independent of performance since they
already own a substantial portion of the equity.
The results, while disheartening in some respects, should be interpreted
with two important caveats. First, we did not measure non-cash compensation
and the performance-related aspects that may be found in these areas. Second,
the reporting of compensation for non-interlisted companies is relatively new
218
EXECUTIVE COMPENSATION AND FIRM VALUE
in Canada. As the compensation information becomes more accessible and
information on the relationship (or the lack thereof) between compensation
and performance is recognized, there will be changes in compensation practices
in firms in which the takeover and capital markets can have an influence.
Thus, companies with entrenched owners/managers, closely held companies,
and those with dual-class shares and, to a lesser extent, those with poison pills,
may lag behind other companies which must make changes in compensation
or in management due to market pressures. It is our expectation that, using
data for future periods, subsequent studies of compensation and performance
will find stronger relationships than we observed, at least for the widely held
companies.
ENDNOTES
1 Daniels &. Macintosh (1991) report that in 1990 only 14 percent of the
companies in the TSE 300 were widely held, compared to 63 percent of the
companies in the American Fortune 500. Of the remainder, 60.3 percent
are owned by a single shareholder with legal control, 25.4 percent by one
shareholder with effective control (20 percent to 49.9 percent of voting
shares) or by two or three shareholders having the ability to combine and
establish joint legal or effective control. As of the end of 1993, approxi-
mately 38 percent of the companies on the TSE 300 were widely held,
based on the 15 percent holding of voting shares definition.
2 For a summary of the literature in this area see Daniels & Halpern (1995).
3 An analysis of the costs and benefits of these instruments is found in Elitzur
(1995).
4 Although the database does include deferred compensation such as stock
options, we do not use it in this study since the reliability of the data is
questionable and there are inconsistencies within the data. For a discussion
of problems with the reporting of Long Term Incentive Plans see OSC Staff
Report, February 1995.
5 We are in the process of collecting data on deferred compensation and
intend to include this variable in subsequent research.
6 For formal analysis of the optimal incentive scheme see Varian, 1992,
Ch. 25, pp. 448-452; Kreps, 1990, Ch. 16, pp. 586-608; and Laffont, 1990,
Ch. 10, pp. 159-164.
7 We used a Chow test in which regressions are run separately for the widely
held and closely held samples and for the two samples combined. The
calculated test statistic has an F distribution.
8 We also tested whether the widely held firms could be as different from the
specification for the closely held firms. The Chow test found that the null
2/9
ELITZUR & HALPERN
hypothesis (which assumed) that the widely held firms had the same
relationship as the closely held firms could not be rejected.
9 We also ran the regressions eliminating the negative earnings dummy
variable in case there was multicollinearity between this variable and the
share return variable. The regression fit is slightly poorer only for the
bonus relationship in the widely held sample and the results for the other
variables are unchanged.
APPENDIX 1
THE THEORY OF EXECUTIVE COMPENSATION SCHEMES
T
HE MODEL FOR THE DESIGN of executive compensation packages is the
Principal-Agent model. This Appendix utilizes such a model which
combines features of models from Rasmusen (1989, Ch. 6), Kreps (1990, Ch.
16), Laffont (1990, Ch. 11) and Varian (1992, Ch. 25). In this model an agent
is hired by a principal to run a firm. The principal's value function, V, depends
on some outcome, x, and the cost of compensation to the agent. The outcome,
x, is a function of effort, e, and a state of nature, 6, occurring according to a
probability density, f(0). The agent's utility, U, depends on compensation, s,
which, in turn, is a function of x. The agent chooses an effort level that
optimizes her payoff. In addition, the agent must be paid enough to take the
job. The minimal utility of the agent which will induce the agent to stay is
denoted as U
0
. Denoting by E the expectations operator, the following
program depicts the principal's problem:
MaxEV (x(e , 9)-C(e))
2 (1)
where C (e) = Min Es(x(e, 9) )
s(. ) (2)
subject to
e= ArgMax EU(e ,s(x( e , 9))) (2.1)
e
EU(.)>U
0
 (2.2)
Constraints (2.1) and (2.2) are known as the Incentive Compatibility Constraint
and the Participation Constraint respectively.
220
EXECUTIVE COMPENSATION AND FIRM VALUE
This problem is solved through the "two steps approach" based on
Grossman & Hart (1983). The solution procedure is as follows:
Step 1 Find for each level of effort, e, the cheapest contract to
induce the agent to (a) take the job and (b) choose that
effort level.
Step 2 Find the optimal level of effort for the principal.
Step 1 is achieved through the solution of Program (2) to (2.2). Step 2 is
obtained by maximizing problem (1), given our solution to Step 1.
From this model, the optimal compensation is of the following form:
s(J* = w + g(x ) :g'(x)>0 (3)
w is a constant and related to the Participation Constraint and, thus, depends
on UQ, the agent's reservation utility (which is exogenous). g(x) denotes a
function which is nondecreasing in outcome, x. If the agent is risk-neutral, the
optimal compensation package will take the following form:
s(.) = w + bx (4)
b is a constant and, thus, Equation (4) describes a linear incentive scheme.
ACKNOWLEDGEMENTS
W
E WISH TO THANK KPMG for access to their executive compensation
database, and Frans Harts and Steven Hadjiyannakis for data collection
and database construction.
BIBLIOGRAPHY
Abowd, J. "Does Performance-Based Manageria l Compensation Affect Corporate
Performance." Industrial & Labor Relations Review, Supplement (1990):52S-73S.
Bhagat, S., J. A. Brickley and R. Lease. "Incentive Effects of Stock Purchase Plans." Journal
of Accounting and Economics. (1983):195-215.
Brickley, J. A., S. Bhagat and R. Lease. "The Impact of Long-range Managerial
Compensation Plans on Shareholder Wealth." Journal of Accounting and Economics.
(1985):115-29.
221
EL1TZUR & HALPERN
Daniels, R. and J. Macintosh. "Toward a Distinctive Canadian Corporate Law Regime."
Osgoode Hall Law Journal. (1991).
Daniels, R. and P. Halpern. "Too Close for Comfort: The Role of The Closely Held Public
Corporation in the Canadian Economy and the Implications for Public Policy."
Canadian Business Law Journal (Forthcoming, Fall 1995).
Elitzur, R., and Yaari. "Executive Incentive Compensation and Earnings Manipulation in a
Multi-period Setting." Journal of Economic Behavior and Organization. (Forthcoming).
Gibbons, R. and K. Murphy. "Relative Performance Evaluation for Chief Executive
Officers." industrial & Labor Relations Review, Supplement (1990):30S-51S.
Jensen M. and K. Murphy. "Performance Pay and Top-Management Incentives." Journal of
Political Economy. (1990):225-64.
Kang, J. and A. Shivdasani. "Firm performance, corporate governance, and top executive
turnover in Japan." Manuscript, University of Rhode Island, 1994.
Kaplan, S. "Top Executive Rewards and Firm Performance: A Comparison of Japan and the
United States." Journal of Political Economy. (1994):510-45.
. "Top Executives,Turnover, and Firm Performance in Germany." Journal of Law,
Economics and Organizations. (1994):142-59.
Kostiuk, P. "Firm Size and Executive Compensation." The Journal of Human Resources.
(1989):90-105.
Kreps, D. M. A Course in Microeconomic Theory. Princeton, N J: Princeton University Press, 1990.
Lambert, R., D. Larcker, and K. Weigelt. "The Structure of Organizational Incentives."
Administrative Sciences Quarterly. (1993):438-61.
Laffont, J. J. The Economics of Uncertainty and Information. Cambridge, MA: The MIT Press,
Second Printing, 1990.
Larcker, D. F. "The Association Between Performance Plan Adoption and Corporate
Capital Investment." Journal of Accounting and Economics. (1983):3-30.
Leonard, J. "Executive Pay and Firm Performance." Industrial & Labor Relations Review,
Supplement. (1990):13S-29S.
Mehran, H. "Executive Compensation Structure, Ownership and Firm Performance."
Journal of Financial Economics. (Forthcoming 1995).
Milgrom, P. and J. Roberts. Economics, Organisation, and Management. Englewood Cliffs,
N.J.: Prentice Hall, 1992.
Murphy, K. "Corporate Performance and Managerial Remuneration." Journal of Accounting
and Economics. (1985):ll-42.
Rasmusen, E. Games and Information: An Introduction to Game Theory. London: Basil
Blackwell Ltd., 1989.
Staff Report. "Executive Compensation and Indebtedness Disclosure." Ontario Securities
Commission. February 17, 1995, 1-86 to 1-95.
Rosen S. "Contracts and the Market for Executives," NBER, Working Paper 3542,
December 1990.
Tehranian, H. and J. Waegelin "Short Term Bonus Plan Adoption and Stock Market
Performance - Proxy and Industry Effects: A Note." Financial Review. (1986):345-53.
Tosi, H. and L. Gomez-Mejia. "The Decoupling of CEO Pay and Performance: An Agency
Theory Perspective." Administrative Sciences Quarterly. (1989):169-89.
Varian, H. R. Microeconomic Analysis. New York: Norton, Third Edition, 1992.
Zenger, T. "Why Do Employers Only Reward Extreme Performance? Examinin g the
Relationships among Performance, Pay and Turnover." Administrative Sciences Quarterly.
(1992):198-219.
222
David Stangeland Ronald]. Daniels Randall Morclc /T
Faculty of Management Dean of Law Faculty of Business v_)
University of Manitoba University of Toronto University of Alberta
In High Gear: A Case Study of the
Hees-Edper Corporate Group
INTRODUCTION
T
HIS STUDY COMPARES FIRMS IN THE HEES-EDPER GROUP with a number of
other independent firms of similar size and in the same industries over a
four-year period from 1988 to 1992, just prior to the first release of news that
the Hees-Edper group was in financial trouble. During that period, Hees-
Edper firms recorded profitability levels comparable to (or below) those of the
matched firms. The Hees-Edper firms were also shown to have been much
higher risk investments well before the group's financial position began to dete-
riorate. They were more highly levered, but even after risk levels are adjusted
for this, the risk levels of Hees-Edper firms remain much higher.
Our study shows that the extreme incentive-based compensation
schemes used by Hees-Edper firms encouraged managers to adopt high-risk
strategies, and that the intercorporate co-insurance (allowed by the interlock-
ing ownership structure of the firms) made this possible by increasing the group's
apparent debt capacity. Since this higher risk did not improve overall perfor-
mance, it was arguably at an economically inefficient higher level. The higher
leverage of Hees-Edper companies should have produced a sizable tax advan-
tage because of the deductibility of interest at the corporate level. The
mediocre performance of the companies thus raises the possibility that abnor-
mally poor performance was masked by tax breaks.
THE ECONOMICS OF CONGLOMERATES
D
URING THE 1960s AND 1970s CONGLOMERATES WERE "the glamour invest-
ment on the stock market" (Firth, 1980) and financial markets reacted to
the news of diversifying acquisitions by sending stock prices of acquiring firms
skyward (Matsusaki, 1993). The resulting mob psychology infected managers
and investors alike, and diversifying acquisitions did enormous damage to
many firms that would otherwise have remained healthy and prosperous.
Eventually, the corporate world realized the diversification was excessive and
223
STANGELAND, DANIELS & MORCK
there was a return to "core" lines of business. Today, the conglomerate merger
wave of the 1960s seems like a mania.
IN DEFENCE OF CONGLOMERATES
ARE THESE JUDGEMENTS PRECIPITOUS? There are arguments in favour of
conglomerates that make economic sense. Indeed, some of them are reasonably
persuasive, at least superficially.
First, Caves (1982) and Rugman (1994) argue that certain intangible
assets have higher returns when used on a larger scale. These are thought to
include R&D, marketing expertise, and good management. The intuition is
that a new product or advertising campaign has fixed up-front costs, but its
return depends on the size of the operation to which it is applied. Similarly, a
good manager in charge of a large operation generates more wealth than the
same good manager in charge of a small operation. The implication is that
good managers should be put in charge of operations that are as large in scope
and scale as possible.
These arguments are a widely accepted justification for international
horizontal expansion, but they also appear to have some applicability to
domestic firms. Montgomery & Wernerfelt (1988) and Panzar & Willig
(1981) utilize them analogously to explain why the wave of corporate diversi-
fication made by conglomerates in the '60s and '70s make sense.
Second, to some extent a conglomerate structure is a substitute for capital
markets. If capital markets were hopelessly myopic or otherwise grossly inef-
ficient, it would make sense to circumvent them. However, most of the
academic work on this issue suggests that markets are not that inefficient and
those, including Keynes (1933), who do argue for such a degree of inefficiency
also argue frequently that corporate managers are afflicted by the same mood
swings that affect investors. Nonetheless, even in an economy with efficient
capital markets, there are reasons to circumvent them. Two such reasons (that
also dovetail into an argument in support of conglomerates) are the "lemon"
problem and the "free-cash" problem.
The lemon problem is characterized by a firm that has good investment
projects but no spare cash; it must therefore raise funds by issuing securities.
Myers & Majluf (1984) point out that this is not costless. Firms should issue
new shares when their outstanding shares are overpriced. Securities, like used
cars, are difficult to value, and buyers are always inclined to suspect that there
is something wrong with the product - otherwise, (they ask) why is it for sale
now? Is it a lemon? Investors might rationally view the news of a new securi-
ties issues as a signal that outstanding securities are overvalued. In fact, share
prices do tend to fall when firms announce they are issuing more shares. Lesser
analogous effects are also observed for bond issues. This lemon problem in capi-
tal markets means that firms should use a "pecking order" approach when
financing new projects - that is, they should use funds obtained from internal
224
IN HIGH GEAR: A CASE STUDY OF THE HEES-EDPER CORPORATE GROUP
cash flow whenever possible, and raise external capital only when internal
funds are not available and when the benefits of the new project outweigh the
cost of depressing the prices of the firm's outstanding securities. By transferrin g
funds between divisions, a conglomerate structure side-steps the lemon problem;
it can act like a financial intermediary in the sense of Diamond (1991).
Jensen (1986) argues that firms in stable, low-growth industries often
invest in money-losing projects. The free-cash problem is characterized by a
firm in a low-growth industry with no profitable investment projects; it should
pay out its free cash to shareholders as dividends. But, retaining funds within
the firm often serves managers in other ways; it enables firms to expand and
thus to build up the size of managers' empires and/or it allows for labour peace
or it cements ties with politicians. Such "over-investment" by cash cows is
called the free-cash problem. Conglomerates that span both low-growth cash-
rich industries and high-growth cash-starved industries neatly solve both the
free-cash problem and the lemon problem in one easy step. A conglomerate
can invest internal funds in the best of all its divisions' projects, and thus better
serve shareholders.
Third, diversification reduces risk at the corporate level (Gahlon &
Stover, 1979). Financial academics never tire of arguing that diversification
brings no benefits to shareholders because shareholders could achieve the
same risk reduction by holding a more diversified portfolio themselves. This
argument is suspect because it assumes that diversification at the corporate
level and at the individual investor's portfolio level are perfect substitutes.
They are not. Reducing risk at the corporate level might allow for more
credible long-term commitments to workers, suppliers and customers. It might
also reduce the need to forego a return on part of the firm's capital in order to
maintain the financial slack necessary to insure liquidity. Lower level corporate
risk might also attract better workers and managers at lower wages, since a risk
premium need not accompany any investment in firm-specific skills (Aron,
1988). It might also encourage managers to undertake more risky corporate
investments than their innate aversion to risk would otherwise preclude,
thereby encouraging a greater alignment of managers' interests with those of
shareholders. Diversification at the corporate level may well not benefit share-
holders, but the case is not as open-and-shut as many believe.
Fourth, diversification reduces corporate taxes by making a more highly
levered capital structure optimal. By insuring each other through intercorporate
transfers of earnings, the divisions of a conglomerate each lower the other's
probability of defaulting on its debt relative to that of a free-standing one-
industry firm. This makes a higher over-all leverage more feasible for a
conglomerate than for a portfolio of independent one-industry firms. Of
course, if the firm elects to lever up in order to take advantage of the co-insurance
to increase its debt-related tax deductions, the risk reduction that corporate
diversification can provide is limited. The benefits of lower corporate taxes to
investors as a whole are mitigated by the higher personal taxes on debt, but in
225
STANGELAND, DANIELS & MORCK
a world where tax-free investors — like pension funds - are playing an ever
greater role, it is not clear that there would be a wash in securities prices in
general. Moreover, shareholders are paying for bailouts they would otherwise
walk away from because of the limited liability granted the owners of stock.
The size of this reduction in dividends, in the absence of tax gains, would
exactly compensate for the better terms the firm could get from creditors and
it would be a wash. The tax deductibility of interest, but not dividends, shifts
the balance in favour of diversification.
THE FAILURE OF THE CONGLOMERATE FORM
T
HE POOR PERFORMANCE OF CONGLOMERATES casts doubt on the universal
validity of the arguments noted above. Berger & Ofek (1995) find a 13
percent to 15 percent discount in the values of conglomerates relative to
comparable portfolios of stand-alone firms. Comment &. Jarrell (1995) find a
positive link between firm focus increases and stock returns. John & Ofek
(1995) find that asset sales improve firm performance when they also increase
the firm's focus. In the 1980s, firms that announced acquisitions in their own
lines of business saw their stock prices rise, while those that announced
takeovers in other industries saw their stock prices decline (Morck et a/.,
1990). Wernerfelt & Montgomery (1988) find a "positive focus effect" in an
empirical study of the determinants of firms' values. Many firms that diversi-
fied aggressively in earlier years spent the 1980s shedding unrelated operations
and re-establishing their commitments to core businesses (Donaldson, 1990).
WHY DID CONGLOMERATES FALL so FAR OUT OF FAVOUR?
FIRST, THE IDEA THAT CONGLOMERATES COULD EXPLOIT the intangible assets of
their component firms on a large(r) scale was always more strained than the
analogous theory justifying multinationals. Arguably, R&.D and marketing
skills are considerably less transferable to operations in unrelated industries
than they are to operations in the same industry but in another country. Thus,
attention was centred on management skills as the intangible asset that would
increase the values of all the assets combined with the conglomerate.
Conglomerates, it was believed, had "... dynamic, entrepreneurial manage-
ment ... [which were] injected into firms which were taken over, [and] greatly
increased efficiency and profits ... which would be reflected in higher share
price performance" (Firth, 1980). Management skills are now viewed as much
less portable; today, managers who are acknowledged experts at finding oil are
not as likely to be considered to have an advantage in running a brewery too.
Moreover, even good managers can be guilty of hubris.
Second, conglomerates were seen as being plagued by corporate gover-
nance problems. They were over-centralized (Baker, 1992). Many degenerated
into little more than exercises in empire-building. Shareholders who would
226
IN HIGH GEAR: A CASE STUDY OF THE HEES-EDPER CORPORATE GROUP
have had considerably more influence and better information about the financial
decisions being made in smaller, one-industry firms were unable to monitor or
discipline the managers of large complex conglomerates who had more oppor-
tunity to run amok. Amihud & Lev (1981) find that manager-run firms are
much more likely to establish diversified conglomerate structures than owner-
run firms; they argue that conglomerates themselves may be a manifestation of
corporate governance problems.
1
 Certainly, the reduced risk in a conglomerate
should be attractive to managers. Rose &. Shepard (1994) find that salaries of
managers of conglomerates are 10 percent to 12 percent higher and total
compensation is 13 percent to 17 percent greater than that of their peers in
otherwise comparable one-industry firms of similar size. However, they also
find that this premium is not related to tenure and argue that this implies it
might be due to the fact that a higher level of skill may be required to manage
a conglomerate.
Although the gains to be made through better use of internal funds,
lower corporate risk, and higher debt capacity may be real, it appears they are
largely swamped by the corporate governance problems that emerge in
conglomerates.
AN AMALGAM OF CONGLOMERATES AND FREE-STANDING FIRMS?
DESPITE THESE FINDINGS, CONGLOMERATES MIGHT STILL be a valid corporate
form, useful in some circumstances. Roe (1994) makes the case that the failure
of the U.S. conglomerate, despite its potential advantages, was due (in part) to
the fact that in conglomerates owning 100 percent of their subsidiaries,
managers were deprived of market signals that provided valuable feedback to
managers in free-standing companies. Instead, conglomerate managers
received feedback through a command-and-control system based mainly on
accounting information. Roe goes on to say that "an amalgam of partial control,
market signalling and partial integration of finance and industry (or of different
levels of Industry)" might have been superior to both the conglomerate form
and the market-disciplined free-standing firms.
However, in the United States, conglomerates with large numbers of
partially owned subsidiaries are discouraged by the Investment Company Act of
1940.
2
 Once 40 percent of the portfolio of a U.S. conglomerate is devoted to the
partial ownership of other firms, the company is presumed to be an investment
company and must therefore pay taxes on dividends it receives from its partially
owned subsidiaries. Since one of the main reasons underlying the existence of a
conglomerate is its ability to reallocate capital efficiently, this is a serious barrier.
The only escape is for the conglomerate to become a mutual fund, but this
entails restrictions on portfolio composition and on intercompany dealings.
In Canada, the federal Investment Companies Act
3
 does not constitute a
barrier to conglomerate formulation comparable to that posed by the U.S.
legislation. This act requires federally incorporated companies that use debt
227
STANGELAND, DANIELS & MORCK
capital to finance equity or debt investments to comply with certain reporting
obligations (administered by the Office of Superintendent of Financial
Institutions - the regulator of federal financial institutions). The Act also
requires companies to comply with restrictions on sundry related-party trans-
actions. The embrace of this statute can be easily avoided through provincial
incorporation or reincorporation, because there are no comparable legislative
schemes in the provinces.
Federal tax legislation has a more important regulatory influence on the
structure and performance of Canadian conglomerates. In contrast to the
United States, Canada permits tax-free dividends to be paid within a corporate
group, thereby permitting internal capital transfers to be effected on a more
tax-efficient basis. However, the comparative benefits of this difference in
dividend treatment should not be overstated. In contrast to the scope permitted
by consolidated reporting of conglomerate earnings in the United States, the
Canadian tax statute does not allow consolidation. Presumably, this makes it
more difficult for Canadian conglomerates to maximize the tax-avoidance
value of losses incurred by member corporations. The more liberal availability
of the deduction for interest payments in Canada further complicates matters,
particularly in respect of debt incurred on foreign assets. Interest deductibility
provides an implicit subsidy for debt, and this encourages corporate managers
to use high levels of debt to finance asset acquisitions. Thus, in tandem, a
cursory review of tax legislation in Canada and the United States does not
supply unequivocal evidence that the size and durability of the Canadian
conglomerate is necessarily related to differential taxation standards. Which
effects dominate is an empirical question.
Looking beyond tax policy, however, there is a range of distinctive
regulatory policies in Canada that, while not providing targeted incentives for
conglomerate formation, create scope for Canadian controlling shareholders
and their appointed managers to engage in opportunistic behaviour via the
conglomerate vehicle. For instance, it is arguable, generally, that the commit-
ment of successive Canadian governments to mercantalist industrial policies
has reduced the bargaining power of Canadian shareholders who invest(ed) in
securities of Canadian corporations. The foreign-property rule of the Income
Tax Act is an example. This rule caps the permissible level of tax favoured
retirement investments at 20 percent of the value of the portfolio (10 percent
until recently). Thus, Canadian investors have fewer alternative investments
to choose from when they wish to move their money because they disagree
with the policies of corporate managers. This may have allowed inefficient
conglomerate holding structures to survive, and may also have prolonged
wealth-reducing redistribution from investors to Canadian corporate insiders.
4
The same arguments can be made in the context of Canadian corporate
and securities law. Here it is arguable that the lack of a vigorous, privately
enforced securities disclosure regime in Canada reduces the transparency of
internal corporate transactions to external shareholders and heightens the
228
IN HIGH GEAR: A CASE STUDY OF THE HEES-EDPER CORPORATE GROUP
attractiveness of the conglomerate form of organization to opportunistic
corporate insiders.
5
 Similarly, the lack of a clearly articulated corporate law
fiduciary duty from majority to minority shareholders in Canada is also significant
— at least historically - in explaining the attraction of conglomerates to oppor-
tunistic managers and shareholders (Daniels & Macintosh, 1991). In the
absence of legislated fiduciary duties, controlling shareholders in Canada and
their appointed management enjoy much greater scope for unfair self-dealing
transactions than if their companies were incorporated in the United States.
6
We also believe that the mercantalist industrial policies adopted by
successive Canadian governments encouraged conglomerate formation. High
levels of external trade protection, restrictions on the export of domestic capital,
and favourable tax treatment of certain types of domestic equity investment
all contributed to an inward-looking industrial economy in which Canadian
corporations focused on producing a broad range of goods and services for the
protected Canadian market rather than on a narrow range of competitive
products for the international market. In this setting, the diversified conglom-
erate served as a natural vehicle to achieve corporate growth.
In sharp contrast to the United States, in Canada a more congenial
political environment for the concentration of economic power provided
further support for the formation of conglomerates. Whereas American political
traditions have coalesced around a deep and abiding mistrust of concentrated
economic power, the Canadian political environment has been much more
sanguine. In Canada, the development and preservation of a fragile national
identity easily outweighs concerns over the concentration of corporate power.
So, to the extent that economic concentration is the inexorable result of state
protectionism, Canadians regard this as a price worth paying to promote
collect!vist goals (Benidickson, 1993).
A respectable case can be made, therefore, that Canadian laws and
customs do encourage just the sort of amalgam Roe (1994) visualizes. The
largest example of corporate concentration in recent years is the Hees-Edper
group, controlled by Edward and Peter Bronfman, to which we now turn.
THE HEES-EDPER GROUP
IN 1952 SAM BRONFMAN, THE ENTREPRENEUR WHO BUILT Seagram's into a
liquor empire during Prohibition in the United States, informed his nephews,
Edward and Peter, that, while his sons would inherit the family business, a trust
would be established to provide for them. By the early 1990s, with the help of
South African financial strategist Jack Cockwell and ignoring the trend against
diversification, that trust - the brothers' nest egg — had grown into a corporate
empire of more than 100 companies spanning industries from merchant banking
to forestry. At its apex, the group of Bronfman companies made up 15 percent of
the total capitalization of the Toronto Stock Exchange. When it was eventually
liquidated the trust yielded more than C$ 100 million in Seagram's stock.
229
STANGELAND, DANIELS & MORCK
Cockwell's strategy was based on pyramids of control. A privately held
company would own a controlling stake in a firm that would hold a control-
ling stake in another firm that would hold a controlling stake in yet another
firm, and so on. Using this strategy, control could be leveraged. The
Bronfmans could fully control a firm in which they held only 51 percent of 51
percent of ... of 51 percent of the stock. By crossing the layers of the pyramid
and liberally using restricted-voting or non-voting shares for outsiders and
super-voting shares for Bronfman insiders, the equity stakes needed to exert
control were further reduced.
This pyramid ownership structure meant that publicly traded rumps of
stock existed throughout the group. Thus, the group's organizational structure
was an amalgam of a conglomerate (with decisions coordinated by the central,
privately held companies) and public ownership (with traded stock, share-
holder meetings, boards of directors, financial statements, and institutional
ownership).
A number of the Hees-Edper companies were added to the group via
workouts organized by the brothers' merchant bank, Hees International
Bancorp Inc. A typical example was the takeover by Hees of Versatile
Corporation in May 1987. Versatile, a farm equipment maker, had expanded
into the energy sector, and then into ship building through the purchase of
Davie in 1985. By 1987 the firm was bankrupt and Hees assumed control in a
workout, eventually holding an equity stake of over 40 percent. Another
example is the 1989 workout of National Business Systems, in which Hees
bought $80 million of the failed firm's debt from U.S. institutional investors
and assumed control. Critics may now refer to this as "vulture capital", but
Hees was arguably acquiring an expertise in organizing the affairs of troubled
firms. This falls into the category of special management skills similar to those
claimed for the managers of U.S. conglomerates in the 1960s and 1970s.
In other cases, the Hees-Edper group expanded by acquiring major
players in specific industries, such as Noranda Forests and MacMillan Bloedel.
The group's real estate firms, Carena Development and Bramalea, the energy
firm Norcen, and the publishing company Pagurian, all played dominant roles
in their respective industries throughout the 1980s.
Another feature of the Hees-Edper group that deserves comment is its
practice of using exaggerated incentive-based compensation schemes to pay
managers and (in some cases) employees. Top managers received salaries that
were low by industry standards, often in the neighbourhood of only $100,000
per year, but were allowed (and expected) to borrow up to ten times their
annual salary from the group — interest free — to buy stock in its member firms.
7
However, based on anecdotal information describing the compensation
arrangements used in one group firm, Royal Trust, it appears that there was
asymmetric sharing of risk and return by management and shareholders;
implicit promises were allegedly made to key managers that they would be
protected from any downside losses resulting from leveraged equity investments,
230
IN HIGH GEAR: A CASE STUDY OF THE HEES-EDPER CORPORATE GROUP
but that they would retain all upside gains. This system was in place for the
better part of a decade. Certainly, these arrangements provided strong incen-
tives for conglomerate managers to take risks. However, the existence of
incentive-based compensation does not in itself mean that agency problems in
the design of these schemes were obviated.
How did the Hees-Edper group take advantage of its hybrid structure? Did
it reduce risk by setting up a network of co-insurance between group firms? Or
did it use this co-insurance to lever up to a higher debt level and convert the risk
reduction into a tax advantage? Did the publicly traded rumps of stock lead to
better corporate governance than would have been the case in a pure con-
glomerate? We now turn to these issues by comparing various financial measures
for Hees-Edper group firms with those for comparable independent firms.
DATA AND METHODOLOGY
P
UBLICLY TRADED COMPANIES IN THE HEES-EDPER GROUP were identified each
year using Statistics Canada's Directory of Intercorporate Ownership. The
period from 1988 through 1991 was selected because these years saw the
group's largest extent. The sample period begins with 1988 to avoid including
the October 1987 crash in the data; the end of the sample period just predates
the real estate problems that triggered the decline of the Bronfman group.
Total debt and total assets were taken from the CD-ROM Canadian
Compustat. Daily stock returns were taken from the TSE-Western CD-ROM.
Companies are classified by industry using three- and four-digit standard
industrial classification (SIC) codes. Size is measured using 1990 total assets.
Each Bronfman company is matched with an independent control
company in the same industry and of roughly the same size. However, because
of the lack of suitable control companies (i.e., companies that are not member
firms of other corporate groups such as the Reichmann brothers' Olympia and
York), most large real estate companies and some financial firms had to be
dropped from the study. This left 19 companies spanning four years - a total of
76 firm-year observations. Six firm-year observations were deleted as outliers,
defined as having beta (B) or variance estimates more than three standard
errors from the mean estimates for that company or its control match, or
having a debt-to-assets ratio greater than one. This left 70 firm-year observa-
tions. The distribution of the data over time is as follows: 19 observations in
1988, 19 in 1989, 17 in 1990 and 15 in 1991.
FINDINGS
T
HE MAIN RESULTS ARE DISPLAYED IN TABLE 1. Tables 2 through 5 contain
statistical test results that determine the reliability of the differences
between the results obtained for the Hees-Edper firms and those obtained for
the control firms shown in Table 1.
231
STANGELAND, DANIELS & MORCK
TABLE 1
UNIVARIATE STATISTICS FOR ALL VARIABLES FOR ALL FIRM-YEARS STUDIED
Row VARIABLE
1)
2)
3)
4)
5)
6)
7)
8)
Note
Operating
Income/Assets (%)
Levered Equity
Beta
Leverage (%)
Unlevered Asset
Beta
SAMPLE MEAN MEDIAN
Edper Firms
Control Firms
Edper Firms
Control Firms
Edper Firms
Control Firms
Edper Firms
Control Firms
7.1
7.7
0.694
0.303
33.1
26.3
0.473
0.215
8.7
9.5
0.645
0.043
32.6
18.5
0.436
0.025
MINIMUM
-95.0
-49.7
-0.22
-0.048
0
0
-0.129
-0.005
MAXIMUM
22.5
25.7
2.03
1.35
70.8
95.1
1.74
1.07
Sample size is 70 Bronfman firm-years and 70 control firm-years, except for operating income,
are only available for 65 firm-year pairs.
STANDARD
DEVIATION
14.7
9.9
0.454
0.402
17.2
24.6
0.337
0.293
where data
Rows 1) and 2) of Table 1 give a measure of overall corporate profitability,
operating income-per-dollar of assets, expressed as a percentage return. Hees-
Edper firms have slightly worse performance by this measure than do comparable
independent firms. However, these differences are not sufficiently clear-cut to
pass the statistical tests shown in Table 2. Analogous tests using other accounting-
performance ratios yield similar results. We conclude that Hees-Edper firms did
not perform better than comparable independent firms. Their performance
was, at best, comparable to that of the matched control firms.
Rows 3) and 4) of Table 1 compare levered equity 6s for the two groups
of firms. A firm's B is a standard measure of risk used by portfolio managers.
8
A high B indicates a high-risk investment, while a low B indicates a relatively
safe investment. The Bs for Bronfman firms are substantially higher than those
of the control firms. Table 3 shows that these differences are statistically highly
significant. We conclude that the stock of Bronfman firms is much riskier than
that of comparable independent firms.
Rows 5) and 6) of Table 1 compare the leverage of Hees-Edper firms
with those of the matched control firms. Bronfman firms have much higher
financial leverage than comparable independent firms, and Table 4 shows
that this difference is statistically highly significant. We conclude that
Bronfman firms have taken on much higher debt loads than comparable
independent firms.
Rows 7) and 8) of Table 1 show a comparison of unlevered Bs. Unlevered
asset Bs are theoretical Bs that companies would have if they had no debt.
9
This risk measure is used by financial economists as a measure of the underlying
risk in the firm's operations. Even after making this adjustment, Hees-Edper
companies continue to have higher risk levels than the control firms. We
conclude that the higher risk of the Bronfman companies is not due solely to
232
IN HIGH GEAR: A CASE STUDY OF THE HEES-EDPER CORPORATE GROU P
TABLE 2
STATISTICAL TESTS COMPARING OPERATING INCOME PER DOLLAR OF ASSETS OF
BRONFMAN FIRMS (DEBT/ASSETS) TO THAT OF MATCHED CONTROL FIRMS
MEAN MEDIAN
Bronfman Companies (%) 7.11 8.72
Industry/Size-Match Companies (%) 7.74 9.48
Difference between Bronfman
and Matching Companies (%) 0.63 0.100
p-Value for Differences (t-Test for Means,
Wilcoxon Signed Rank Test for Medians) 0.468 0.811
Number of Observations (Firm Years) 65 65
TABLE 3
STATISTICAL TESTS COMPARING LEVERED EQUITY BETAS OF BRONFMAN FIRMS WITH
THOSE OF MATCHED CONTROL FIRMS
MEAN MEDIAN
Bronfman Companies 0.694 0.645
Industry/Size-Match Companies 0.303 0.043
Difference between Bronfman
and Matching Companies 0.391 0.360
p-Value for Differences (t-Test for Means,
Wilcoxon Signed Rank Test for Medians) 0.0001 0.0001
p-Value for Weighted t-Test 0.0001
Number of Observations (Firm Years) 70 70
high debt loads. The underlying business operations of the Hees-Edper firms
appear to entail more risk than other Canadian companies of similar size in
the same industries. They have higher operating leverage as well as higher
financial leverage.
Missing from Table 1 are stock market performance measures. A proper
analysis of stock price performance is complicated for firms like the Hees-Edper
group. Their involved and interlocking ownership structure with multiple classes
of differential voting stock, some privately held, make it difficult to apply the
standard tools of firm valuation and stock return measurement. We are pursuing
further research in this area, examining a broader range of performance
measures, including some that are market-value based.
233
STANGELAND, DANIELS & MORCK
TABLE 4
STATISTICAL TESTS COMPARING LEVERAGE IN BRONFMAN FIRMS (DEBT/ASSETS) TO
LEVERAGE IN MATCHED CONTROL FIRMS
MEAN
Bronfman Companies (%) 33.1
Industry/Size-Match Companies (%) 26.3
Difference between Bronfman
and Matching Companies (%) 6.8
p-Value for Differences (t-Test for Means,
Wilcoxon Signed Rank Test for Medians) .0033
Number of Observations (Firm Years) 70
MEDIAN
32.6
18.5
6.6
.0009
70
TABLE 5
STATISTICAL TESTS COMPARING UNLEVERED ASSET BETAS OF BRONFMAN FIRMS
WITH THOSE OF MATCHED CONTROL FIRMS
MEAN
Bronfman Companies .473
Industry /Size-Match Companies .215
Difference between Bronfman
and Matching Companies .259
p-Value for Differences (t-Test for Means,
Wilcoxon Signed Rank Test for Medians) .0001
p-Value for Weighted t-Test .0001
Number of Observations (Firm Years) 70
MEDIAN
.436
.025
.261
.0001
70
CONCLUSIONS
O
UR OVERALL RESULTS SUPPORT THE FOLLOWING CONCLUSIONS. First, our
data show no clearly superior performance in terms of average return on
assets for Hees-Edper firms over comparable independent firms. This suggests
that the conglomerate structure did not improve overall economic efficiency
by enabling member firms to exploit each other's intangible assets and thereby
achieve new synergies. For example, employing superior management techniques
developed at one firm to invigorate another should have produced higher
234
IN HIGH GEAR: A CASE STUDY OF THE HEES-EDPER CORPORATE GROUP
performance. Therefore Canadian public policy that directly or indirectly
encourages the formation of conglomerates cannot be justified on the grounds
of increased economies of scale or scope in applying such assets, at least in this
case, or if such advantages were achieved, they were offset by other negative
factors.
Second, our data do not provide evidence that the Hees-Edper group
allocated capital internally in ways superior to those accomplished by financial
markets or financial institutions. This should also have produced evidence of
better performance in group firms than in comparable independent firms. The
fact that it did not casts doubt on the benefits of centralized managerial control
over a diverse range of industries. Quite simply, senior Hees-Edper managers
failed to confer tangible economic gains on member firms through superior
capital allocation. Indeed, quite the opposite may be true. Hees-Edper manage-
ment may have used a small stable of cash cows to support earlier investments
in chronically under-performing firms. In this respect, the lack of vigorous
market pressure may have allowed the conglomerate's management system-
atically to persist in maintaining irrational and idiosyncratic commitments to
dog companies.
Third, Hees-Edper management did not use the co-insurance their
conglomerate structure allowed to reduce overall risk levels in the corporation.
Instead, they used the group's risk-sharing potential to increase overall levels
of risk beyond what would have been permitted by debt markets for comparable
independent firms. This was accomplished in part through increased financial
leverage, and in part it appears to stem from higher operating leverage, i.e.,
riskier overall business practices. To the extent that Canadian business is
hampered by excessive innate risk aversion on the part of Canadian managers,
encouraging conglomerates may have an invigorating effect.
This line of argument is valid only if the managers of the conglomerate
use the risk-sharing potential of intercorporate co-insurance to justify invest-
ments that would otherwise be regarded as too speculative. The riskier
management decisions made in the Hees-Edper group were facilitated by the
group's conglomerate structure, but they might not have occurred without the
extreme incentive-based compensation schemes Bronfman managers and
employees were given. Furthermore, high relative levels of firm risk should
have been accompanied by higher levels of relative returns if the risk-taking
was of an economically efficient sort. Perhaps the incentive-based compensation
scheme the group used actually encouraged excessive and overly speculative
risk-taking.
Fourth, Hees-Edper companies were more highly levered than comparable
independent firms, and this likely produced a tax advantage. The fact that this
is not reflected in higher earnings casts a somewhat harsher light on the
mediocre accounting performance of the firms in the group. Also, since the
higher debt in group firms was accompanied by risk-sharing, there need be no
improvement in managerial incentives of the sort Jensen (1989) envisions.
235
STANGELAND, DANIELS & MORCK
Jensen essentially argues that an imminent threat of bankruptcy encourages
better management in highly levered firms. Thus, to the extent that the group
used the increased debt capacity created by its intercorporate risk sharing to
avoid taxes, its social benefits are more questionable.
PUBLIC POLICY IMPLICATIONS
HOW CAN PUBLIC POLICY ACCENTUATE the desirable features of corporate
groups like Hees-Edper and mitigate their undesirable features? As argued
elsewhere, vibrant and open capital and product markets are probably the
strongest antidote against the growth of seemingly perverse organizational
structures (Morck, 1995; Daniels & Halpern, 1995). With vigorous markets,
the ability of managers to devise and maintain inefficient organizational forms
is constrained. In this respect, we believe that further relaxation of the foreign
property rule, a continued liberalization of external trade barriers, and reduced
protectionism of domestic capital market suppliers are all necessary steps.
However, we believe that other policy instruments are also in order.
Nuanced reforms to securities regulations that give private investors
both the ability and the incentive to prosecute alleged breaches of disclosure
obligations would be useful, particularly at a time when resources for public
enforcement are so limited. Reforms to corporate and securities proxy rules
that impair institutional shareholders' voices (such as the shareholder commu-
nication rules that require shareholders to bear the costs of a dissident proxy
circular in the event of a disagreement with management) would also be useful
(Pound, 1991). Both of these reforms would bolster the capability of share-
holders to monitor and to intervene as required, with the result that less
reliance would be placed on the putative superiority of internal versus external
systems of capital allocation. Also, federal regulators should recognize that some
variation on section 9.1 of the Ontario Securities Commission Regulations
(which forces disclosure about intercorporate transactions in such groups) is
critically important in preventing abuse in groups like Hees-Edper. A regulation
of this nature should be a key part of any federal securities law if Ottawa asserts
its jurisdiction in that area.
Another set of reforms to securities regulation focuses on tax distortions
that implicitly favour debt over equity instruments by allowing the deduction
of interest. Removing the interest subsidy on debt would remove an incentive
to share risk across companies in a corporate group like Hees-Edper solely in
order to to reduce corporate taxes. The effect of the change would be to
reduce the desirability of strained capital structures such as those in levered
buyouts (LBOs). On the whole, such a change could also be expected to
reduce the general corporate tax rate. If it is desirable to subsidize debt because
high leverage encourages more careful management decisions (as Jensen,
1989, argues) it should be recognized that conglomerates can and do circum-
vent this. Intercorporate risk-sharing in such groups allows for higher leverage
236
IN HIGH GEAR: A CASE STUDY OF THE HEES-EDPER CORPORATE GROUP
without a perspective-enhancing increased chance of bankruptcy. Perhaps
intercorporate dividends should therefore be taxed more vigorously, and over-
all corporate tax rates be reduced.
Finally, although more amorphous in character, we believe that changes
to the political climate in which Canadian corporations operate are appropriate.
It is perplexing that the substantial concentration of economic power that was
amassed in the Hees-Edper group received so little attention from the regulators
and the financial community. It strikes us as odd that a single group was able
to assemble control over more than 15 percent of the market capitalization of
the country's premier stock exchange with scarcely a hint of criticism by
regulators, politicians or the press. In this respect, we suggest that increased
scrutiny, more analysis and, indeed, reform of the economic and political
institutions that could accept such potentially destabilizing economic power
are in order. However, concerns about vertical equity should not be used to
thwart the creation of optimal organizational arrangements. Specifically, we
fear that deep-seated concerns about vertical equity will limit the capacity of
shareholders to devise workable incentive-based compensation arrangements that
could, in turn, spawn incentives for managers to use firm-level diversification
rather than explicit pay differentials to guard their firm-specific human capital
investments. The Hees-Edper group provides solid evidence of the ability of
strong incentive schemes to increase risk taking. It would be a pity if such
incentives were disallowed simply because of concern that successful managers
might earn too much.
237
STANGELAND, DANIELS & MORCK
ENDNOTES
1 A third concern raised at the time was that conglomerates could engage in
predatory pricing in one market by diverting profits from another (see,
e.g., Bradburd, 1980; Greening, 1980). But this should increase, not
decrease, the relative financial performance of conglomerates.
2 Investment Company Act of 1940 §3(1)(3), 15 U.S.C. §80a-3(a)(3) 1988.
See also Roe (1994), p. 260.
3 R.S.C. 1985, c. 1-22.
4 These policies are discussed more fully in Daniels & Halpern, "The Role of
the Closely Held Public Corporation in the Canadian Economy and the
Implications for Public Policy," Canadian Business Law Journal, forthcoming.
5 In contrast to the United States, Canada does not have clearly articulated
civil liability standards for certain disclosure documents. Moreover, the
incentives for private enforcement of existing statutory and common law
disclosure standards is subverted by the prohibitions on contingency fees
and class actions.
6 The recent (and growing) receptivity of Canadian courts to imposing
fiduciary duties from majority to minority shareholders under the rubric of
oppression, combined with the Ontario Securities Commission's 1990
enactment of its policy on related party transactions (O.S.C. Policy 9.1),
has greatly restricted the scope for majority opportunism.
7 Financial Post, August 7, 1990.
8 Betas here are estimated using the market model r = a + fir
M
 + e. The TSE
300 index is used as the market return r
M
. A different regression is done
using daily data for each firm in each year.
9 A firm's unlevered beta, denoted fiu, is calculated as fiu = ft x (value of
equity/value of firm), where the value of equity over the value of the firm
is approximated by one minus debt over assets. This construction under-
states flu if the firm's debt is sufficiently risky that it carries a hefty risk
premium. Thus, it may understate the level of risk in some Bronfman
companies at some times.
238
IN HIGH GEAR: A CASE STUDY OF THE HEES-EDPER CORPORATE GROUP
BIBLIOGRAPHY
Amihud, Yakov and Baruch Lev."Risk Reduction as Managerial Motive for Conglomerate
Mergers." Bell journal of Economics, 12, 2 (1981):605-17.
Aron, Debra J. "Ability, Moral Hazard, Firm Size, and Diversification." Rand Journal of
Economics, 19,1 (1988):72-87.
Baker, George P. "Beatrice: A Study in the Creation and Destruction of Value." Journal of
Finance, 47, 3 (1992):1081-119.
Benidickson, Jamie. "The Combines Problem in Canadian Legal Thought, 1867-1920."
University of Toronto Law journal, 43 (1993):799.
Brenner, Menachen and David H. Downes. "A Critical Evaluation of the Measurement of
Conglomerate Performance Using the Capital Asset Pricing Model." Review of
Economics and Statistics, 61, 2 (1979):292-96.
Berger, Philip and Eli Ofek. "Diversification's Effect on Firm Value." Journal of Financial
Economics, 37 (1995):39.
Caves, Richard. Multinational Firms, Competition and the Productivity in Host Country
Markets. Cambridge: Cambridge University Press, 1982.
Comment, Robert and Gregg Jarrell. "Corporate Focus and Stock Returns." Journal of
Financial Economics, 37 (1995):67.
Daniels, Ronald and Paul Halpern. "Too Close For Comfort: The Role of the Closely Held
Public Corporation in the Canadian Economy and the Implications for Public Policy."
Canadian Business Law journal (Forthcoming, 1995).
Daniels, Ronald and Jeffrey Macintosh. "Toward a Distinctive Canadian Corporate Law
Regime." Osgoode Hall Law Journal. (1991):884.
Diamond, Douglas. "Monitoring and Reputation: The Choice Between Bank Loans and
Directly Placed Debt." Journal of Political Economy 99 (1991):689.
Donaldson, Gordon. "Voluntary Restructuring: The Case of General Mills." Journal of
Financial Economics, 27, 1 (1990):117-41.
Firth, Michael. "Takeovers, Shareholder Returns and the Theory of the Firm." Quarterly
Journal of Economics, 94 (1980):235.
Gahlon, James M. and Roger D. Stover. "Diversification, Financial Leverage and
Conglomerate Systematic Risk." Journal of Financial & Quantitative Analysis, 14, 5
(1979):999-1013.
Jensen, Michael. "Agency Costs of Free Cash Flow, Corporate Finance, and Takeovers."
American Economic Review, 5 (1986):323.
. "Eclipse of the Public Corporation." Harvard Business Review, 61, 5 (1989).
John, Kose and Eli Ofek. "Asset Sales and Increase in Focus." Journal of Financial Economics,
37(1995):105.
Keynes, John Maynard. The Means to Prosperity. London: Macmillan, 1933.
Matsusaki, John G. "Takeover Motives during the Conglomerate Merger Wave." Rand
Journal of Economics, 24, 3 (1993):357-79.
Montgomery, Cynthia A. and Birger Wernerfelt. "Diversification, Ricardian Rents, and
Tobin's q." Rand Journal of Economics, 19, 4 (1988):623-32.
Morck, Randall. "On the Economics of Concentrated Ownership." Canadian Business Law
Journal, (Forthcoming, 1995).
Morck, Randall, Andrei Shleifer and Robert Vishny. "Do Managerial Objectives Drive Bad
Acquisitions?" Journal of Finance, 45 (1990):31.
239
STANGELAND, DANIELS & MORCK
Myers, Stewart and Nicholas Majluf. "Corporate Financing and Investment Decisions
When Firms Have Information That Investors Do Not Have." Journal of Financial
Economics, 13 (1984):187.
Panzar, John C. and Robert D. Willig. "Economies of Scope." American Economic Review,
71,2(1981):268-72.
Pound, John. "Proxy Voting and the SEC: Investor Protection Versus Market Efficiency."
Journal of Financial Economics, 29 (1991):241.
Roe, Mark. Strong Managers, Weak Owners: The Political Roots of American Corporate
Finance. Princeton, N.J.: Princeton University Press, 1994.
Rose, Nancy and Andrea Shepard. "Firm Diversification and C.E.O. Compensation:
Managerial Ability on Executive Entrenchment?" Unpublished manuscript, 1994.
Rugman, Alan. "The Comparative Performance of U.S. and European Multinationa l
Enterprises, 1970-79." Management International Review, 23, 2 (1994):4.
Wernerfelt, Birger and Cynthia A. Montgomery. "Tobin's q and the Importance of Focus in
Firm Performance." American Economic Review, 78, 1 (1988):246-50.
240
Paul Halpem,
Faculty of Management
University of Toronto
Vijay Jog, n
School of Business (
Carleton University
INTRODUCTION
O
N APRIL 28, 1983 BELL CANADA ENTERPRISES INC. (hereafter BCE)
succeeded Bell Canada as the parent corporation of a group of companies
with interests in telecommunications services, telecommunications equipment
manufacturing, energy and printing. At the time of the reorganization the new
management holding company claimed to have the largest number of share-
holders of any Canadian corporation. In the decade between 1983 and 1993,
BCE's asset base grew from $15 billion to $37 billion.
This study analyzes BCE's growth, its strategic directions, and its value-
creation performance in a number of ways, all within the context of agency
costs and the role of corporate governance. To this end we describe the nature
and size of BCE, evaluate the effectiveness of its corporate governance structure,
and analyze the potential for the principal-agent problems that can lead to
agency costs of equity.
1
 We also investigate the agency cost associated with
free cash flow, given the use of BCE's conglomerate structure as an internal
capital market; and we document changes to its strategic directions as gleaned
from its annual reports. We then emphasize BCE's acquisition-divestiture
strategies and its financial performance over the ten years since its inception.
Finally, we evaluate the market's assessment of BCE over this period by
investigating the market value of BCE equity as compared to its value estimated
as the sum of the market value of its constituent parts.
Our analysis indicates that BCE has displayed marginal performance over
its initial ten years. During the first six years, BCE operated much like a large
conglomerate, both purchasing and establishing companies, many of which
were in areas unrelated to its core activities. The financing of those trans-
actions was based on the steady cash flows derived from its regulated telecom-
munications business. Its many acquisitions, however, do not appear to have
increased shareholder value. A change in strategic direction occurred in the
post-1989 period, which appears to have been recognized by shareholders.
However, even with this change in strategic direction, it is apparent that BCE
management still sees itself as a manager of assets, rather than of businesses.
241
Bell Canada Enterprises:
Wealth Creation or Destruction?
HALPERN & JOG
From our analysis we conclude that BCE is an under-valued company
exhibiting the classic characteristics of the "adverse selection" problem. Its
past performance appears to have led investors to place a high discount on its
underlying assets, reflecting the previous poor performance of its acquisition
activities. It also appears that the diversified nature of the firm and its agency
problems have had negative consequences for the valuation and performance
of BCE. One way to unlock this under-valuation would be to reshape BCE as
a more focused firm - possibly with an arm's-length and minority relationship
with its regulated subsidiary, Bell Canada.
WHY BCE?
B
CE PROVIDES A UNIQUE OPPORTUNITY to assess the effects of agency costs
and corporate governance. It is the largest company in Canada, and its
shares are held by many individual and institutional investors, both domestically
and internationally. The company is well-known outside Canada, and is
interlisted on the New York Stock Exchange (NYSE). In addition, the gover-
nance structure of the company is consistent with the agency cost model.
There are small managerial equity holdings; and although many institutions
own large numbers of BCE shares, none of them is represented on the board.
BCE is also insulated from the discipline of the takeover market for two reasons.
The first is the sheer size of the company; although higher valued transactions
(takeovers) have occurred in the United States, there have been none of this
(potential) magnitude in Canada. The second reason is that there are foreign
ownership restrictions on Canadian telecommunications companies and these
preclude a foreign firm from acquiring an economically significant position in
BCE. Moreover, because of BCE's large market capitalization, institutional
investors in Canada, almost by default, have to hold shares of BCE, since it
commands such a large weighting in the TSE 300 Index.
Finally, as a regulated utility Bell Canada, BCE's largest subsidiary, has
historically contributed a large and relatively stable cash flow to the BCE entity.
While providing a check on spending and investment by Bell Canada, the
regulatory process has no oversight of the operating, investment and financing
decisions of BCE Inc.
2
The diversified nature of BCE and its size in relation to the rest of the
Canadian corporate sector raises issues that have interesting implications for
how we view corporate governance. These issues relate to agency costs arising
from relatively small managerial stakes (simply because the size of the corporation
prevents any individual from having a large stake); the issues of free cash flow;
and, due to its large size, entrenchment behaviour.
3
 While we can identify
closely held Canadian companies to investigate the effect of entrenchment,
they may not reflect the effect of free cash flow and the agency costs of equity
through low managerial stakes.
242
BELL CANADA ENTERPRISES: WEALTH CREATION OR DESTRUCTION ?
SOME THEORETICAL BACKGROUND AND BCE-SPECIFIC DATA
T
HE CONFLICT BETWEEN MANAGEMENT and shareholders, often referred to as
the principal-agent problem, underlies the analysis in this study. The
severity of the difficulties associated with agency issues is related to ineffective
monitoring of management and to the existence of managerial compensation
contracts that do not align the interests of management and shareholders. In
addition, other markets, such as the takeover and product markets, which
normally assist in reducing the severity of agency issues, may not be effective
in the case of BCE.
This study considers a number of agency issues. First, there is the agency
cost associated with low levels of managerial share ownership. In this case,
managers consume excessive perquisites since the cost of this consumption to
them is much lower given their small ownership stake. This perquisite
consumption relates not only to lavish offices, for example, but also to low
levels of effort. In addition, managers can strive to increase the size of the firm
based on their desire for greater prestige and visibility, their wish to leave a
legacy, and their anticipation of the increased compensation, which is generally
related to the growth of both sales and assets of the firm.
4
 In addition to excessive
perquisite consumption, low levels of share ownership by management can
lead to an adverse selection problem which has a negative effect on share
prices. The intuition behind the adverse selection problem is the same as for
the classic Akerlof lemon's premium in the market for used cars. This refers to
a case where potential shareholders may find it difficult to reach a correct
value of the firm due to their inability to separate "good" firms from "bad"
ones; this inability, in turn, arises from the possibility that some firms will
misrepresent firm information. If this is so, outside investors may demand a
higher risk premium from all firms, thus producing a wedge between the cost of
internal and external financing.
The second agency cost issue relates to the dollar investment by managers
in the firm. Even though their proportionate ownership interest may be small,
their dollar investment in the firm may represent a substantial proportion of
their personal wealth. Since these managers may have poorly diversified
personal portfolios, they may forgo projects that are profitable but may be risky
since they are affected by the cost of failure to a greater extent than widely
diversified shareholders. In addition, these managers may have an incentive to
engage in risk-reducing investments through a conglomerate corporate structure.
The existence of free cash flow(s), defined as the amount of internally
generated cash flow available over and above the funds needed to take all
value-maximizing investment decisions, can also lead to an agency cost that
reduces the share price. In this case, management uses the free cash flows to
invest in unprofitable projects that increase the size of the firm, and hence
their compensation and influence. This problem is particularly acute in
companies with steady cash flows from operations. The conglomerate structure
243
HALPERN & JOG
can lead to an internal capital market where cash flows from various parts of
the conglomerate firm are uced by management to invest in new businesses. In
such companies, managers typically believe they can make better use of corporate
cash flows than can individual shareholders. In this organizational form, the
divisions of the firm share a common general office whose functions are to
allocate the internally generated funds, monitor the performance of individual
divisions and engage in strategic planning. However, the monitoring by
investors of the executives who are responsible for these central office functions
is blunted, since the firm's reliance on outside capital markets is relatively
minor. These considerations imply that, in companies with certain character-
istics — such as low managerial equity stakes, conglomerate structure, free cash
flows and entrenched management - the share price will be depressed to
reflect the agency costs and the inability to control these costs through effective
monitoring, external market oversight, or increasing managerial equity stakes.
To provide a perspective on BCE in the context of the agency issues
mentioned here, we present some statistics on BCE common equity securities.
As can be seen in Table 1, the total number of shareholders for BCE has
decreased over time, and the average share holding has increased from 740
shares per shareholder in 1985 to 1,278 shares in 1993. To some extent, this
may reflect the growth of institutional-investor holdings in BCE stock. As
suggested in a number of articles (including that by Holderness & Sheehan,
1988), the presence of institutional investors may overcome the free-rider
problem caused by the cost of monitoring by individuals, and provide an
effective control of agency costs in corporations.
5
 A similar pattern can be
seen when the number of shareholders and the average holding per shareholder
are grouped by foreign and Canadian ownership. A striking feature of this
representation is the larger average holding per shareholder for foreign owner-
ship compared to Canadian ownership; this is consistent with the larger role
played by institutions as investors in the United States, compared to Canada.
Since BCE is extremely large, it is also clear that a typical shareholder is
not in a position to impose any capital market discipline. If it is assumed that
control of at least a 5 percent block is required to have a credible influence;
this would represent a holding of $700 million in 1993, an amount large
enough to rule out even some of the country's largest institutional investors.
As an indication of the sheer size of BCE, the market value of the corpo-
ration's equity increased from approximately $10.4 billion in 1985, equal to 9
percent of the total market value of shares on the TSE 300, to $14-2 billion in
1993, or 5.42 percent of the TSE 300 value. Despite its loss of relative weight
on TSE 300 between 1985 and 1993, BCE had the largest market value of
equity of all companies in Canada for all the years shown in Table 1.
As noted earlier, the ownership of shares of a company by management,
and the monitoring activity of investors and the board of directors, are
important elements in agency costs. In Table 2 we present some information
directed at the potential for agency costs. We first note the size of the board of
244
TABLE 1
BCE SHARE STATISTICS
Total Shareholders
Average Shareholdings
No. of C'dn Shareholders
No. of Common
Shares Owned (OOOs)
Average Shareholdings
No. of Other Shareholders
No. of Common
Shares Owned (OOOs)
Average Shareholdings
C'dn-Owned Shares (OOOs)
C'dn-Owned Shares (%)
Foreign-Owned Shares (%)
1985
332,440
740
325,877
231,332
710
6,563
14,714
2,242
231,332
94.02
5.98
Relative Weight on TSE 300 (%) 9.01
Market Value of
Equity ($ millions) 10,365
1986
338,528
783
331,623
250,695
756
6,905
14,580
2,112
250,695
94.50
5.50
7.93
9,850
1987
318,675
859
311,847
251,986
808
6,828
21,881
3,205
251,986
92.01
7.99
7.21
10,169
1988
319,202
907
312,320
266,385
853
6,882
23,160
3,365
266,385
92.00
8.00
7.14
10,786
1989
288,619
1,046
282,008
269,214
955
6,611
32,837
4,967
269,214
89.13
10.87
7.66
13,783
1990
277,295
1,101
270,805
273,734
1,011
6,490
31,676
4,881
273,734
89.63
10.37
7.97
12,064
1991
260,747
1,190
255,532
282,366
1,105
5,215
27,926
5,355
282,366
91.00
9.00
8.51
14,779
1992
254,521
1,200
249,431
280,918
1,126
5,090
24,427
4,799
280,918
92.00
8.00
6.96
12,672
1993
241,078
1,278
233,846
261,937
1,120
7,232
46,224
6,392
261,937
85.00
15.00
5.42
14,252
TABLE 2
GOVERNANCE STRUCTURE
Percentage Ownership
Board
No. of Directors
No. of Corporate Insiders
No. of Directors -
Other Companies
No. of Directors - Other
Percentage Ownership
Corporate Insider
Other Company
Law Firm
Financial Institution
Investment House
Individual
1983
0.0014
21
7
9
5
67.80
30.59
0.81
0.68
0.00
0.12
1984
0.0014
22
10
9
3
76.12
22.62
0.63
0.48
0.00
0.14
1985
0.0007
19
7
9
3
51.31
46.21
1.18
0.94
0.00
0.36
1986
0.0008
21
7
10
4
61.20
36.60
0.94
0.95
0.00
0.31
1987
0.0008
20
6
9
5
72.63
23.27
0.83
0.84
2.16
0.28
1988
0.0010
20
6
10
4
86.25
6.75
0.82
0.72
5.26
0.19
1989
0.0011
19
5
10
4
87.03
6.73
0.85
0.00
5.20
0.19
1990
0.0012
20
6
10
4
88.13
6.66
0.85
0.00
4.14
0.21
1991 1992
0.0005 0.0005
20 19
7 6
9 9
4 4
70.86 71.99
13.69 13.99
1.04 1.18
0.00 0.00
13.77 12.15
0.63 0.69
1993
0.0003
15
4
7
4
56.01
12.69
0.58
0.00
29.58
1.14
BELL CANADA ENTERPRISES: WEALTH CREATION OR DESTRUCTION?
directors. As seen in Table 2, the number of board members declined from 21
in 1983 to 15 in 1993; the most striking decrease occurred in 1993 with the
change in CEO. While there is no theory that suggests the optimal size of a
board to promote effective monitoring, one senses intuitively that the larger
the board the less effective it is likely to be. This is particularly true if there are
many "inside" members on the board. The relative importance of inside directors
to outside directors is also shown in Table 2. Over the sample period, the ratio
of inside directors to all directors ranged from a high of .45 in 1984 to a low of
.26 in both 1989 and 1993. The outside directors were composed primarily of
individuals from other companies. There were also a few directors from
investment houses, financial institutions and law firms. Finally, there were no
directors from either pension or mutual funds. This board structure is not
dramatically different from that of many other large Canadian companies and
suggests that the monitoring role of the board is not very strong.
The existence and continuation of agency problems are affected by the
board of directors and its role in monitoring the management. If we assume
that membership on the board is suggested by the incumbent management,
that information is provided to the board by the same incumbent manage-
ment, and that the members of the board have no (or very little) equity in the
firm, then we can infer that the board of directors is unlikely to discipline
management to act in the best interests of its shareholders. Thus, the role of
the board is crucial in resolving agency problems.
Our analysis shows that BCE has followed the standard practice of
nominating a majority of outside members to the various required sub-
committees of the board. Although not shown here, outsiders do represent a
majority on the sub-committees on Management Resources, Pension Fund
Policy, Investments, and Audit. Unfortunately, we know very little about the
role(s) played by members of the BCE board, especially by the outside members.
This does not imply that boards of directors generally (or the BCE board
specifically) are necessarily captive to incumbent management; there is some
empirical evidence that boards can and do occasionally engage in disciplinary
behaviour. However, the evidence is not rich enough to indicate any pattern
of intensity and success with respect to a board's monitoring when the firm's
management is performing in a "satisfying" mode as opposed to a "wealth-
maximization" role.
6
The other element of agency costs derives from the percentage of shares
owned by management. The higher this proportion, the better the alignment
of the interests of managers and shareholders. In their survey paper, Jensen &
Warner (1988) state:
Share ownership can be an important source of incentives for management,
boards of directors, and outside block holders. The pattern and amount of
stock ownership influence managerial behaviour, corporate performance, and
stockholder voting patterns in election contests. In addition, a firm's
247
HALPERN & JOG
characteristics can influence its ownership structure. The data suggest that for
some firms the amount of ownership by inside or outside block holders is
economically significant. The precise effects of stock holdings by managers,
outside shareholders, and institutions are not well understood, however, and
the interrelations between ownership, firm characteristics, and corporate
performance require further investigation.
In the case of BCE, the percentage of BCE shares owned by insiders who
are members of the board ranges from .0003 percent to a high of .0014 percent
(see Table 2). Such small percentages are not at all surprising considering the
size of BCE. Although not shown, some insiders held large numbers of BCE
shares (during the sample period) and therefore had large dollar investments in
the corporation. The effect of substantial dollar holdings by insiders is ambigu-
ous. If the total investment in BCE is a large proportion of the insider's wealth,
an insider could, conceivably, try to reduce the company's risk (thereby reducing
any risk to his or her personal portfolio) by being overly conservative with
respect to board decisions. On the other hand, if the amount invested in the
company is a small proportion of a board member's overall wealth, then any risk-
reducing behaviour will not be as important - although the agency issue will still
be present. In the case of BCE, since it is so large, it is unlikely that any attempt
by management to reduce risk through investment policy would be effective
because so many assets would have to be shifted to have an effect on the company's
risk. However, significant changes in assets and in risk exposure can be achieved
by large diversification investments and/or divestitures. Such risk-reduction
behaviour can be a significant factor in the decisions undertaken by manage-
ment and may influence the value of the firm.
It is also interesting to note how the ownership percentages of the
board members changed between 1983 and 1993. In the early years of the
sample period, there was relatively little equity owned by members of the
board who could be considered as truly "outsider" — i.e., not even associated
with BCE-related subsidiaries. However, in the last three years of the period,
the relative ownership of BCE shares shows that outsiders held a proportion-
ately larger number of shares. Nonetheless, this is unlikely to lead to any
attempt to influence BCE's management. As will be seen later, this increase
in ownership by outsiders coincides with some significant strategic changes
announced by BCE.
Another interesting characteristic of BCE is its somewhat larger degree
of immunity from product-market discipline. BCE has as its base Bell Canada,
which is a regulated telecommunications company. Although there is now
competition in the long-distance market and competition in local markets is
on the horizon, Bell Canada has a strong brand name and remains large and
financially strong. These factors ensure that even with these changes in the
marketplace, Bell Canada will continue as a significant player, generating cash
flows for the parent. This does not imply that Bell Canada will not restructure
248
BELL CANADA ENTERPRISES: WEALTH CREATION OR DESTRUCTION?
its operations to improve internal cash flows. However, the important issue is
how these cash flows will be used.
Finally, the traditional disciplinary constraints on agency costs are not
applicable to the BCE case. The takeover market has little relevance since the
acquisition of a controlling interest would require such a substantial invest-
ment in a hostile takeover - an amount that is not likely to be available even in
today's setting of globalized capital markets.
7
This brief review of BCE suggests that the company has been substantially
immune from the elements of the typical market-discipline environment
with which other public corporations must cope. Its management holds a
relatively small stake in the company, with outsiders owning relatively
higher percentage(s) only in recent years. Thus, BCE has the elements
consistent with high agency costs of equity. In addition, the holding company
structure with large numbers of subsidiaries also introduces an agency cost of
free cash flows.
BCE: FORMATION AND GROWTH
GENERAL
IT IS IN THE CONTEXT OF FORMATION AND GROWTH that a review of BCE is most
relevant and necessary. Since the reorganization of BCE in 1983 was a crucial
development in its future growth (and had been long sought by Bell Canada's
management), our review of BCE starts in that year.
Prior to the 1983 reorganization, Bell Canada was a regulated company,
subject to regulatory review and the determinations by the regulatory authority
(the CRTC) with respect to rates that were "fair and reasonable". To allow for
reasonable rates, the CRTC determined a revenue requirement based on the
expected costs to be incurred by Bell plus an allowed return on capital projected
over a period called the "test year". The total amount of this revenue require-
ment was then compared to forecast revenues based on existing rates; and rate
adjustments were then approved by the CRTC. Crucial in determining the
allowed return on capital was an estimate of the cost of equity capital, which
was applied to the dollar book value of the equity of the consolidated company.
Clearly, expected net income to Bell depended upon the allowed return to
equity capital; the actual return on equity, and hence net income, depended on
actual revenues earned and costs incurred over the period. Since there was no
distinction made between regulated and unregulated operations under this
form of regulation, there was de facto regulation of all of Bell Canada's operations.
Thus, if a subsidiary such as Northern Telecom had a good year, its profits
would be included in the regulatory process and would contribute to a lowering of
telephone rates. If other subsidiaries' financial results were poor, telephone
rates could be adversely affected. Whether or not the regulators would
treat these outcomes symmetrically was always a matter of concern to Bell
249
HH
management. In any case, the process did not provide Bell with any incentive
to engage in profitable competitive services or non-telephone operations.
The regulatory changes associated with the reorganization in 1983 were
very important to BCE, particularly with respect to the growth of its non-
regulated operations. Apart from ventures that were defined as integral to the
provision of telecommunications services, such as the company that prints the
telephone directories, BCE could now start companies and make investments
free from the implicit regulation that was the hallmark of the previous regulatory
regime. This new structure therefore provided an incentive to engage in these
competitive operations. The change was also accomplished by a redefinition of
corporate assets, some of which were regulated while others were not. An
allowed rate of return was applied to the regulated assets which required the
determination of a capital structure for the company's regulated operations.
These operations had been a subset of the previous Bell Canada.
Prior to the reorganization, Bell Canada had both wholly-owned
subsidiaries and majority control of a number of companies. However, the
reorganization provided an incentive to expand both through acquisition and
internal growth funded by Bell Canada's substantial cash flows. A review of
the 1984 composition of BCE shows that it owned 15 major subsidiaries.
8
 In
1987, this number increased to 20, further increasing to 29 in 1990, 33 in
1992 and then declining to 26 in 19.94- If all of the subsidiaries of these major
subsidiaries are included, the numbers of companies that were directly and
indirectly controlled by BCE management were 64 in 1984, 105 in 1987, 198
in 1990, 178 in 1992, and 148 in 1994. The number(s) of subsidiaries and the
increase over time are a clear indication of the expansion strategy followed by
management.
BUSINESS SEGMENTS
BCE INC. HAS OPERATIONS IN MANY AREAS that are both related and unrelated
to telecommunications. Table 3 shows the sectoral breakdown of BCE
Revenue, Assets and Earnings by business segment.
9
 The two major segments
are telephone-related: Telecommunications and Equipment Manufacturing.
For Revenue and Earnings a third segment, denoted as "Other", includes
operations that are non-telecommunications related, such as financial services
and pipelines, among others. For the Assets category, there are two additional
segments, "Investments" and "Corporate". The former reflects investments in
financial assets and investments in non-telephone related companies; the latter
is the dollar amount, at book value, for BCE headquarters.
For Revenue, growth occurred in the two major segments, with
Equipment Manufacturing growing at a faster rate than Telecommunications.
The "Other" segment increased steadily through 1987, dipped in 1988,
increased again (to its maximum) in 1990, then decreased quickly thereafter.
The pattern for the two major segments in the Assets category again displays
250
HALPERN & JPG
BELL CANADA ENTERPRISES: WEALTH CREATION OR DESTRUCTION?
growth over the period, with Telecommunications growing more rapidly than
Equipment Manufacturing. The "Investments" and "Other" segments in this
category display growth up to 1990, at the end of which there was a significant
reduction, signifying the change in strategy we address below. There was also a
dramatic increase in the "Other" segment in 1989, followed by another large
increase in 1990. This reflects significant investments by BCE. Beginning in
1991, the dramatic reduction is a result of the changes in operating strategy
discussed below. The Corporate segment decreased in the period from 1987 to
1989, increased from 1990 to 1992, then fell again in 1993.
For Earnings, which we define as earnings before interest and taxes, we
provide data for each of the three major segments. The data for 1993 are not
presented since they are not available using a consistent definition, making
any comparison with previous years virtually impossible. Earnings for the
Telecommunications segment have continued to grow, reflecting the stable
nature of the regulated operations. The Equipment Manufacturing segment is
more volatile but, apart from 1988, showed positive growth over the period
ending in 1992.
10
 The "Other" segment was extremely volatile, reflecting not
only the uneven operations of the entities in this segment but also the divestiture
and closing of some of the companies.
Figures 1, 2 and 3 illustrate the breakdowns for Table 3 in percentage
terms. For Revenue (Figure 1), the "Other" segment accounts for a very small
percentage of the total, while the Telecommunications segment provides a
slightly diminishing proportion over time. Beginning in 1991, the percentage
of revenue from Equipment Manufacturing exceeded the percentage from
Telecommunications. In the analysis of the Assets category (Figure 2) the
three non-telecommunications segments are accumulated into "Other". In
Assets, the Telecommunications segment continues to account for the largest
component; and the "Other" segment increases until 1990 (with a small
retrenchment in 1988); from 1991 onward, the proportion is small. Figure 3
shows clearly that the Telecommunications segment has been a consistent
source of earnings for BCE.
In all three sectors, assets grew at a higher rate than revenue and earnings.
The Equipment Manufacturing segment now has 34 percent of assets and
generates 53 percent of revenue, but produces only 30 percent of earnings.
Despite the fact that revenues from the Telecommunications segment fell from
57 percent in 1983 to 43 percent in 1992, this sector still contributes over
two-thirds of earnings. It can also be seen that the Telecommunications segment
has been a steady performer, with operating margins averaging 28 percent and
return on assets in the 12 percent range. On the other hand, both the
Equipment Manufacturing and "Other" segments have had low and volatile
margins. Clearly, the corporation's foray into non-telecommunications businesses
was not accompanied by attractive and steady performance; BCE's overall
performance has clearly been supported by the regulated side of its business.
251
TABLE 3
SECTORAL BREAKDOWN OF BCE - REVENUE
Revenue
Telecommunications
Equipment Manufacture
Other
Total
Assets
Telecommunications
Equipment Manufacture
Other
Investments
Corporate
Total
Earnings
Telecommunications
Equipment Manufacture
Other
Total
1983
5,076
3,276
550
8,902
10,571
2,213
337
1,183
636
14,940
1,436
297
77
1,810
1984
5,541
4,359
715
10,615
11,213
3,356
423
1,423
1,072
17,486
1,586
470
112
2,167
ASSETS AND EARNINGS ($ MILLIONS)
1985
5,966
5,829
1,051
12,846
11,882
3,867
1,918
1,802
1,114
20,583
1,735
581
178
2,494
1986
6,390
6,114
1,063
13,567
12,704
3,926
3,589
2,190
1,305
23,714
1,843
583
197
2,623
1987
6,758
6,471
1,092
14,321
13,830
3,849
3,815
3,577
954
26,025
1,772
606
194
2,572
1988
7,092
6,598
755
14,445
15,981
5,428
236
3,291
645
25,581
1,838
250
40
2,128
1989
8,011
7,161
1,509
16,681
17,334
5,861
11,797
3,640
629
39,261
2,035
623
76
2,734
1990
8,468
7,851
2,054
18,373
18,326
6,229
12,980
3,286
1,166
41,987
2,259
820
(3)
3,076
1991
8,151
9,379
964
18,494
19,595
11,066
(120)
0
2,992
33,652
2,358
1,067
146
3,571
1992
8,415
10,222
935
19,572
21,106
11,895
599
0
,056
36,656
2,495
1,178
39
3,712
1993
8,614
10,550
663
19,827
22,079
12,608
851
0
1,170
36,708
n/a
n/a
n/a
n/a
TABLE 3 (CONT'D)
1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993
PERCENTAGES
Revenue
Telecommunications
Equipment Manufacture
Other
Total
Assets
Telecommunications
Equipment Manufacture
Other
Total
Earnings
Telecommunications
Equipment Manufacture
Other
Total
57.02
36.80
6.18
100.00
70.76
14.81
14.43
100.00
79.35
16.41
4.24
100.00
52.20
41.06
6.74
100.00
64.12
19.19
16.69
100.00
73.16
21.67
5.16
100.00
46.44
45.38
8.18
100.00
57.73
18.79
23.49
100.00
69.57
23.30
7.14
100.00
47.10
45.07
7.84
100.00
53.57
16.56
29.87
100.00
70.26
22.23
7.51
47.19
45.19
7.63
100.00
53.14
14.79
32.07
100.00
68.90
23.56
7.54
100.00 100.00
49.10
45.68
5.23
100.00
62.47
21.22
16.31
100.00
86.37
11.75
1.88
100.00
48.02
42.93
9.05
100.00
44.15
14.93
40.92
100.00
74.43
22.79
2.78
100.00
46.09
42.73
11.18
100.00
43.65
14.84
41.52
100.00
73.44
26.76
0.00
100.20
44.07
50.71
5.21
100.00
58.23
32.88
8.89
100.00
66.03
29.88
9.00
100.00
43.00
52.23
4.78
100.00
57.58
32.45
9.97
100.00
67.21
31.73
1.05
100.00
43.45
53.21
3.34
100.00
60.15
34.35
5.51
100.00
n/a
n/a
n/a
n/a
Note: n/a = not available.
HALPERN & JOG
FIGURE 1
BCE REVENUE: SECTOR BREAKDOWN
STRATEGIC DIRECTIONS
IN LIGHT OF THE UNEVEN AND GENERALLY UNATTRACTIVE performance in its
non-regulated business sectors, it is worthwhile to evaluate BCE's changes in
strategic direction over the ten-year period under review. There appear to
have been two distinct phases in the evolution of BCE strategic philosophy.
The first phase, from 1983 to approximately 1989, reflects diversification and
growth through acquisitions directly by BCE or through subsidiary companies.
This strategy is closely associated with the CEO, Jean de Grandpre. The second
phase, from 1990 to 1993, reflects the corporation's return to its core activities
- by divesting non-telecommunications assets and purchasing telecommunica-
tions-related assets both in Canada and abroad. Growth in the second phase
was accomplished both by acquiring majority positions and by making non-
majority investments. Clearly, the latter are portfolio investments and, in this
sense, the company did not change from the holding company philosophy that
it followed during the first phase. Throughout both phases (periods), the
254
BELL CANADA ENTERPRISES: WEALTH CREATION OR DESTRUCTION?
FIGURE 2
BCE ASSETS: SECTOR BREAKDOWN
corporation's regulated telecommunications activities in Canada have provided
significant cash flows.
Phase 1 - The Diversification Years
Beginning in 1983 with the acquisition of TransCanada Pipeline (TCPL), BCE
entered its diversification phase funded by its available pool of internal capital.
As noted in the 1983 annual report,
BCE will avoid large investments in businesses which have financial charac-
teristics fundamentally different from those of the existing businesses - for
example, companies which experience severe fluctuations in earnings through
an economic cycle....
TCPL was attractive because of the growth and stability of its income stream
and asset base.
255
HALPERN & JOG
FIGURE 3
BCE EARNINGS: SECTOR BREAKDOWN
In considering the supply of funds during this acquisition phase, the following
was noted in the annual report:
We have a large and growing pool of capital; and BCE will continue to study
opportunities for investment and diversification, either directly or through
subsidiary companies.
BCE's diversification strategy was continued in 1984 with the purchase of a
number of large real estate assets. The diversification approach was affirmed in
the following statement in the annual report.
An important aspect of BCE's management activities is to encourage entry
into new businesses which may not fall naturally within the business plans,
experience or resources of a particular business within the group.
The importance of the telecommunications segment in the BCE organization
was reinforced in 1986, when severely depressed energy prices combined with
weaknesses in parts of the real-estate markets had a negative impact on BCE
256
BELL CANADA ENTERPRISES: WEALTH CREATION OR DESTRUCTION?
earnings. As noted in the annual report, "The negative impacts of some of
these conditions on some of our investments were more than offset by strong
growth in other operations, especially telecommunications"(italics added). The
company also extended its diversification activities to additional markets. As
the annual report noted, "the security provided by the diversity of our positions
in various industries is now fully reinforced by the additional security of
diverse markets for each company's products and services". During 1986, the
company raised $300 million of equity of which $200 million was used to
maintain its percentage holding in BCE Development Corporation (BCED),
its real estate operation.
In 1988, while affirming their strategy of diversification, management
noted a change in its previous policy - of generally seeking a "strong voting
position" in companies in which BCE invested.
Recent transactions do, in effect, signal a more flexible attitude towards
control in situations where we have strong confidence both in management
and in the fundamental soundness of the business prospects.
Finally, in 1989, the company's corporate strategy began to be questioned
openly. That year, the company entered the financial services business with
the purchase of Montreal Trust Inc. from Power Financial Corporation. A
selection of quotes from the 1989 annual report illustrate these questions and
the beginning of the shift in emphasis.
We have reassessed what sort of businesses are best suited to our corporate
strategies, and how these investments should fit together.
We intend to continue our investments in telecommunications. We have also
entered into financial services and we expect the future will see growth in this
area. In addition, BCE will maintain certain other long term investments,
such as natural gas transportation, but will not hesitate to review and alter
other holdings in the light of changing circumstances.
In view of what happened in 1989, when disastrous results from one sub-
sidiary, BCED, had the effect of wiping out the contribution to BCE's con-
solidated net income of several well-performing companies, some shareholders
may question the whole notion of a diversified holding company, or retain
some nostalgia for earlier, more simple times.
BCE assets outside of telecommunications must meet the criteria of financial
compatibility. Financial services and natural gas pipelines, although different
from our core telecommunications business, fit this corporate strategy.
257
HALPERN & JOG
TABLE 4
BCE ACQUISITIONS 1983 - 1989
COMPANY
DATE OF
ANNOUNCEMENT COMMENT
TransCanada December 5, Bell acquires 5.3 million shares from Dome, valued at
Pipelines 1983 $168 million, tenders an offer for all shares outstanding
at $31.5 per share; approximate value of $670
million, or an 8.34% premium.
Offers 0.85 Bell shares for 2.6 million shares.
Approximately $71 million - a 27% premium.
First move into real estate. Approximate value $160
million. General tender offer for shares outstanding at
$3, representing a 13.33% premium.
TCPL announces bid for Encor to expand oil and gas base
and offers $8.75/share. Approximate value $980 million.
BCE gives Kinburn a loan of $190 million.
Esso Bank Note September 21,
1984
Daon Development January 21,
Corporation 1985
Encor
Kinburn
Northwest Tel
November 16,
1987
February 9,
1988
August 29,
1988
Montreal Trust Inc. March 8,
1989
Encor
Pays CNR $200 million for all shares of Northwest Tel,
giving Northwest Tel a monopoly over Northern BC,
Yukon, and NWT.
Package of $875 million in cash and shares. Purchased
from Power Financial Corporation, representing a
premium of $4-38/share, 2.5 times book value.
January 5, TCPL proposes spinning off Encor to BCE to raise cash
1989 and reduce debt. TCPL to get $570 million in cash to
pay down debt.
Table 4 sets out the list of major diversification acquisitions undertaken during
this period (Phase 1, as referred to earlier) when management saw its role as
managers of assets, not as managers of businesses. In 1987 the annual report stated:
Asset growth is a particularly relevant measure of BCE's performance, because
the role of BCE's management is primarily one of managing assets rather than
business.
Five essential ways asset management can enhance BCE's performance:
1. Acquisition
2. Divestiture
3. Re-grouping of existing units
4. Creation of new operating units
5. Public equity participation.
258
BELL CANADA ENTERPRISES: WEALTH CREATION OR DESTRUCTION?
The management of BCE is committed to using all the options available to it
to pursue the ultimate objective of increasing the company's value to its share-
holders.
This emphasis on asset growth is consistent with agency costs of equity
and free cash flow rather than shareholder value maximization through
efficient operation of companies.
Phase 2 - Returning to Core Activities
In 1990 a new CEO took over from de Grandpre. Raymond Cyr began a
process which attempted to refocus the operations of BCE. The company
concentrated on the telecommunications business both in Canada and inter-
nationally, where there were perceived to be many growth opportunities. The
result was a decided improvement in the financial results of the company. In
addition, the company reduced its investment in Encor and TCPL and
announced its intention to divest the remainder of its energy-related assets by
the end of 1992. BCE also wrote down the value of loans made to Kinburn and
wrote off BCED. However, as part of an assistance plan for BCED, the company
invested $250 million in a joint venture with Catena Developments, which acted
as the manager of the rescue attempt. Tables 5A and 5B list the major acqui-
sitions and divestitures during these years. This focus continued through 1991.
BCE's strategy in the immediate future is to continue refining our corporate
focus, divesting non-core interests and improving profitability. In a longer per-
spective, we will expand our role on the international market, with the goal of
becoming a strong global competitor.
The final year of Cyr's tenure as CEO was marked by a continuation of
the return to BCE's core businesses. As noted in the annual report, there was a
divestment of $1.1 billion of non-core assets.
BCE's major goal will be to continue to improve our earnings by focusing
closely on telecommunications, asserting our competitive strengths in Canada
and expanding our international business through alliances, joint ventures,
and investments.
Thus the company continued to make acquisitions during this period,
including the acquisition of 20 percent of Mercury Communications Limited,
a British provider of telecommunications services, and a majority interest in a
U.K.-based cable TV and telecommunications company. This international
focus was also reflected in the telephone directory business, with the acquisition
of the Caribbean Publishing Company and the formation of a joint venture to
publish directories in India.
In 1993 a new CEO took over and the refocusing on core activities, both
domestically and internationally, continued. The company completed the
259
HALPERN & JOG
TABLE 5A
BCE ACQUISITIONS 1990 -1993
COMPANY
DATE OF
ANNOUNCEMENT COMMENT
BF Realty January 24, Assets transferred into BF Realty with a $500-million loan
1990 from BCE and Carena.
Kinburn/SHL April 17, BCE seized 5.1 million shares of SHL as part of a Kinburn
Systemhouse 1990 loan covenant. Kinburn failed to pay $350 million of a
$400'million loan.
Telemax September 27, Joint bid for 20.4% of Mexican government Telephones
1990 with an approximate value of $400 million.
SHL Systemhouse February 3, SHL strategic partnership issues 32.5 million shares at
1992 $12.75, valued at $414 million.
SHL Systemhouse June 10, Alliance deal falls apart.
1992
Mercury November 11, British cable investment, valued at $960 million.
Communications 1992 This gives BCE a 20% stake in Mercury.
Talisman March 10, Exchange of Encor shares for Talisman shares. BCE to
1993 hold 19% of Talisman shares, worth $234 million. This
is considered a more liquid investment.
Talisman/Encor April 15, Share offer enhanced by $17.7 million.
1993
Jones December 2, 30% stake in U.S. cable company, valued at $330 million.
Intercable 1993 BCE Telecom International also paid $55 million for an
option to buy control, as well as invest $100 million in
expansion plans.
divestiture of its non-telecommunications assets, acquired non-controlling
positions in foreign companies and engaged in the formation of a number of
joint ventures. As noted in the 1993 annual report, "BCE will continue to
invest in rapidly growing telecom markets in Canada and abroad".
However, even with its redefined focus on core activities, senior BCE
management continued to view its role as that of a manager of assets. The
shareholders, while viewing the divestiture of the diversification activities
positively, need not be satisfied with the continuation of the asset manage-
ment philosophy and the holding company approach. BCE shareholders do
not need BCE to make partial investments in publicly traded corporations; if
they want to make such investments, they can make them on their own.
260
BELL CANADA ENTERPRISES: WEALTH CREATION O R DESTRUCTION?
TABLE 5B
BCE DIVESTITURES 1989 - 1993
COMPANY
DATE OF
ANNOUNCEMENT COMMENT
BCED
BCED
BCED
Encor
TransCanada
Pipelines
Kinburn
June 27,
1989
August 11,
1989
January 24,
1990
June 27,
1990
September 10,
1990
October 24,
1990
Montreal Trust Co. December 2,
1993
BF Reality
and Brookfield
December 2,
1993
O&Y offers to buy BCED for $557 million.
O&Y deal fails. Analysts believe they backed out of the
deal.
Closed out BCED. BCED takes $610 million write-down.
BCE took a $400 million write-down for its share of
BCED.
Unloading 48.9% of Encor in share issue. Valued at $100
million. $7 for 3 shares and 2 warrants. Shares trading
at $2.
Issues warrants to sell 48.9% stake in TCPL for
approximately 1.3 billion. $119 million after tax gain.
BCE takes $224 million write-down due to loan default.
Sold to Bank of Nova Scotia for $290 million with a $400
million write-down. Some analysts believe price $100
million less than expected.
Sold to Carena Development Corp., taking a
$350 million write-down.
EVALUATION OF THE OPERATING STRATEGIES
ACQUISITIONS, DIVESTITURES AND WHITE-DOWNS
To IDENTIFY THE MAJOR ACQUISITIONS and divestitures for which dollar values
are reported, we reviewed the Financial Post cards and BCE annual reports,
considering only acquisitions made by BCE and not those made by wholly
owned subsidiaries such as Bell Canada or partially owned companies such as
Northern Telecom. There are a number of acquisitions and divestitures in
every year, some of which may be relatively small individually but, depending
on whether BCE is in a net acquiring or divesting mode, may be large in the
aggregate. Also, there are some transactions for which dollar values are not
available. The total amount of investments in other companies and sub-
sidiaries can therefore be substantial. Further, BCE's capital expenditures are
extremely large. From the consolidated statements we observe that capital
expenditures in every year from 1987 to 1993 were never less than $3.2 billion.
261
HALPERN & JOG
TABLE 6
ACQUISITIONS & DIVESTITURES ($ MILLIONS)
ACQUISITIONS
TCPL
Esso Bank Note
Daon
Encor
Kinburn
Northwest Tel
Montreal Trust
Encor
BF Realty
Telemax
Mercury
Talisman
Jones Intercable
DATE
Dec. 5/83
Sept. 21/84
Jan. 21/85
Nov. 16/87
Feb. 2/88
Aug. 29/88
Mar. 8/89
Jan. 5/89
Jan. 24/90
Sept. 27/90
Mar. 11/93
April 10/93
Dec. 2/93
ORIGINAL $
838.00
71.00
160.00
980.00
263.00
200.00
875.00
570.00
500.00
400.00
960.00
251.70
330.00
1994$
1,218.14
100.38
223.39
1,213.76
299.30
240.01
1,024.60
675.19
561.30
436.94
977.88
254-04
329.26
CAR (-5, +5)
-0.0376
-0.0121
0.0661
-0.0075
0.0290
0.0032
-0.0044
-0.0055
-0.0493
-0.0078
0.0072
0.0092
0.0097
T-STAT
-2.27
-1.19
6.51
-0.82
1.79
0.32
-0.43
0.52
-6.86
-0.99
0.86
1.06
1.17
DIVESTITURES
BCED
Encor
TCPL
Kinburn
Montreal
Trust Co.
Montreal
Trust Co.
BF Realty
Jan. 24/90
June 27/90
Sept. 10/90
Oct. 24/90
Dec. 2/93
Dec. 2/93
Dec. 2/93
400.00
100.00
1,300.00
224.00
290.00
400.00
350.00
449.04
112.26
1,420.05
242.67
289.35
399.10
349.21
-0.0493
-0.0166
-0.0032
0.0670
0.0097
0.0097
0.0097
-6.86
-2.34
-0.43
7.84
1.17
1.17
1.17
Except for 1991 and 1992, the value of acquisitions was generally large in all
the years (between $1.0 billion and $1.9 billion). Finally, disposal of invest-
ments began in 1990 and ranged from $400 million to $1.3 billion. In our
analysis, we consider only those transactions that were in excess of $200 million.
262
BELL CANADA ENTERPRISES: WEALTH CREATION OR DESTRUCTION?
FIGURE 4
BCE ACQUISITIONS 1983 - 1993
1,500
1,000
500
Acquisitions Acquisitions ($ 1994)
&
/ c#
^ ^ ^ V
^ <&
^v ^ <//v <A/yyv
</
*f
w
Note: Acquisitions made before 08/89 were made with Grandpre as CEO. Between 08/89 and 04/93,
Cyr was CEO, Also note that Encor was purchased initially by TCPL and was spun off to BCE.
Table 6 and Figures 4 and 5 summarize these major acquisitions and
divestitures in original dollars as well as in 1994 dollars. Table 6 also shows the
reaction of capital markets to the corresponding announcement of these
events, as reflected in the BCE share price. The first series of transactions is
concerned with the acquisition and ultimately the sale of TCPL. December
1983 registered the first of a series of transactions in which BCE acquired a signif-
icant portion of TCPL shares; by 1987, BCE had invested approximately $1 billion
in TCPL. In 1987 TCPL purchased Encor and, in 1989 under a reorganization, all
of TCPL's oil and gas assets were consolidated into the newly formed Encor Inc.,
with BCE receiving shares in this newly created company. In addition, BCE
purchased $227 million of Encor preferred shares, which were convertible into
Encor equity. Beginning in 1990 BCE began to sell off the Encor and TCPL
assets, and by 1993 all TCPL shares had been sold for a total of approximately
263
HALPERN & JOG
FIGURE 5
BCE DIVESTITURES 1990-1993
1 50fl
1 000
c
g
6
JS
J
1 3
^ 500
0
BCEI
01/9C
3
llll
Enco
06/9(
1
r
)
TCP
09/9C
L
1 1 Divestitures
1 1 Divestitures ($ 1994)
lll l 1
Kinburn Montreal Montreal BF Realty
10/90 Trust Co. Trust Co. 12/93
12/93* 12/93*
Note: Montreal Trust was sold in two blocks.
$1.3 billion. In 1993, Encor shares were sold to Talisman for $275 million worth
of Talisman shares. BCE received approximately $100 million in 1990 for the sale
of some Encor shares. The cost of the combined TCPL and Encor transaction
was approximately $1.2 billion. BCE received approximately $1.6 billion over
the ten years that the corporation held equity in these companies. This total
transaction was nominally profitable.
The second series of transactions, which relate to BCE's foray into real
estate, was not as profitable. In 1985 BCE purchased approximately 69 percent
of the shares of Daon for approximately $160 million. The real estate market
faltered badly and, in mid-to-late 1989, Olympia and York offered to purchase
Daon, but the deal fell through. At the beginning of 1990, BCE wrote down
its investment in Daon, renaming it BCED. BCE then entered a joint venture
with Carena in which Carena would manage the BCED assets. BCE agreed to
advance up to $250 million to the joint venture. The real estate market failed
to improve, however, and in 1994 BCE sold its interests in BCED, renamed BF
Realty, to Carena in exchange for an agreement that Carena would provide
financial support to a subsidiary of BF Realty in order to enable the subsidiary
to complete its restructuring. These acquisitions and divestitures were
extremely unprofitable for BCE, and did little to generate confidence in the
264
BELL CANADA ENTERPRISES : WEALTH CREATION O R DESTRUCTION?
ability of management to engage in a profitable diversification strategy under a
holding company structure.
The third set of transactions centred on another diversification thrust,
this one into the financial services industry, with the acquisition of Montreal
Trust from Power Corporation for $875 million in shares and cash. The deal
was originally transacted in 1989. In December 1993 BCE agreed to sell
Montreal Trust to the Bank of Nova Scotia in exchange for (bank) common
shares valued at approximately $290 million. BCE took a $400 million charge
against its investment in Montreal Trust.
In 1988 the company sold its interests in certain subsidiary companies to
Kinburn for notes valued at approximately $190 million. In 1990 Kinburn
defaulted on its loan. BCE then wrote down the loans and seized 5.1 million
shares of SHL Systemhouse as a part of a Kinburn loan covenant. BCE had
previously held a small interest in SHL. Early in 1993 the company sold its
equity investment in SHL, but the terms were not disclosed. There had been
previous attempts to sell this interest.
During this period BCE continued to acquire companies that were in the
telecommunications area. In 1988 the company acquired Northwest Tel for
$200 million. Some of the international diversification transactions include
the acquisition of an interest in the equity of Telemax, the Mexican govern-
ment telephone system, for approximately $400 million; the acquisition of a
20 percent equity interest in Mercury Communications, a British cable company,
for approximately $1 billion; and the acquisition of a 30 percent interest in the
equity of Jones Intercable, a U.S. cable company, for approximately $330 million.
ABNORMAL RETURNS ON ANNOUNCEMENTS
To DETERMINE THE REACTIONS OF INVESTORS to announcements of acquisitions
and divestitures, we estimated the cumulative abnormal returns (CAR) and
the corresponding t-statistics for a period of 11 days, centring on the date of
each announcement. Although not shown here, most of the observed reactions
to BCE's acquisition-divestiture program, as indicated by the CARs, were
found to be small and statistically insignificant." These findings of no meaningful
reaction by BCE's shareholders to BCE's divestiture and acquisitions are not
surprising for two reasons. First, while some of the transactions were large in
terms of absolute dollar amounts, they were still small relative to the size of
BCE, and thus market reaction, if any, would be difficult to detect. Second, for
the events associated with write-downs there should be no abnormal return,
since this is not a cash flow item and these write-downs reflect the economic
realities that should already be reflected in the share price of BCE. A signifi-
cant share price reaction would be observed only if the write-down provided
new information to the market. For example, the BE Realty/BCED transac-
tions in January 1990 could reflect the disappearance of the Olympia and York
bid and the formation of a joint venture, which could lead to continuing losses.
However, no meaningful statistically significant reactions were observed.
265
HALPERN & JOG
FIGURE 6
CUMULATIVE WEALTH INDEX, BCE vs TSE 300
Even in the absence of statistically significant reactions, one of the note-
worthy features of these acquisition-divestiture phases at BCE is that, in most
instances, these acquisitions resulted in significant value destruction for BCE
shareholders. Except for a nominal gain recorded for the TCPL divestiture,
BCE's investments in the realty, energy, financial services and software integration
areas suffered major losses. Its current portfolio of acquisitions is relatively
new, and whether these acquisitions ultimately prove to create value awaits
the verdict of time. Moreover, from the reaction of the security market to the
announcements of these acquisitions, investors did not interpret them as
wealth-enhancing.
Mercury Communications and Jones Intercable reflect minority positions
and their impact on BCE's cash flows will depend upon activities and manage-
ment decisions unaffected by BCE management. As straight portfolio invest-
ments, the insignificant abnormal returns associated with these transactions
are to be expected.
12
266
BELL CANADA ENTERPRISES: WEALTH CREATION OR DESTRUCTION?
FIGURE 7
BCE CUMULATIVE EXCESS RETURNS
Notes: CRW is calculated with Ri - Rf.
CRW* is calculated with standard market model.
Both omit October 1987.
PERFORMANCE OF COMPANY RETURNS
AN IMPORTANT PERSPECTIVE OF BCE PERFORMANCE has to do with the impact
of operating, financing, and investment decisions on its shareholders. Figure 6
shows the cumulative wealth index based on monthly returns for BCE and the
TSE 300 Index for the period from 1983 to 1994. The cumulative excess
returns shown in Figure 7 are measured using the risk-free rate as well as the
standard market model over the same period.
13
 Figure 6 shows that BCE stock
outperformed the TSE 300 Index especially in the post-1990 period, a period
consistent with the change in corporate strategy. Similar conclusions can be
drawn from Figure 7, which shows positive cumulative excess returns in that
latter period. Over the period from 1985 to 1990 the cumulative excess wealth
falls. This is consistent with almost continuously negative excess returns,
267
HALPERN & JOG
implying that during this period BCE's shareholders experienced negative risk-
adjusted returns. The cumulative excess returns have yet to reach the relative
performance of BCE in its early years.
Viewing BCE as a portfolio of investments suggests another way to
observe the wealth effects of investments made by BCE management. As a
portfolio, the excess return over an investment horizon should also provide
another indication of the success of the investments made by BCE. The Jensen
performance measure is used to identify any wealth gains or losses. It is the
estimated value of the intercept from a regression of the excess monthly return
on BCE equity on the excess monthly return on the TSE 300 Index. The
excess return is defined as the rate of return on equity for month t, less the
corresponding monthly yield on a Government of Canada Treasury bill.
Table 7 sets out the values of the intercept and summary statistics for
the regressions over various time periods. We first consider the effect of
the diversification activities on the performance of the company. Looking
over the period from 1983 to 1989 when all of the diversification activities
occurred, the Jensen measure was very small (.076 percent) and not signifi-
cantly different from zero. What can we say about the period when the new
strategy emerged in 1990 to refocus and return to the core activities by
concentrating on the telecommunications business? The Jensen measure for
this period is .024 percent and, again, is not significant. This is probably due
to the fact that this period was still affected by the continued presence of
Montreal Trust, which was not sold until the end of 1993. The poor perfor-
mance of this subsidiary would have affected the performance of the overall
company. Therefore, it may be too early to evaluate the effect of the new strategy.
Using the four three-year non-overlapping periods to evaluate the perfor-
mance of BCE, beginning in 1983, we also see that none of the intercepts is
statistically significant, and the values of the Jensen measure are small.
Clearly, the overall performance of the corporation during this period does not
indicate that BCE created shareholder wealth from a risk-adjusted perspective.
VALUATION OF BCE
THE HOLDING COMPANY DISCOUNT
ANOTHER WAY TO IDENTIFY THE EFFECTS of the potential agency costs from
small managerial equity stakes and free cash flows on value creation in the
context of a holding company is to compare the market value of the equity
of BCE with the sum of the values of its constituent parts. If the market
value of the equity of the total entity were found to be less than the sum of
the market values of the component parts, this finding would be consistent
with the presence of agency costs and what is often referred to as the "holding
company discount". In general, the BCE share price will be affected by
three factors.
268
BELL CANADA ENTERPRISES: WEALT H CREATION OR DESTRUCTION?
TABLE?
JENSEN REGRESSIONS
PERIOD
1983 - 1985
1984 - 1986
1985 - 1987
1986 - 1988
1987 - 1989
1988 - 1990
1989 - 1991
1990 - 1992
1991 - 1993
1983 - 1989
1990 - 1993
ALPHA
0.01076
0.00044
-0.00245
-0.00494
0.00076
-0.00046
0.00292
0.00363
0.00165
0.00076
0.00024
TSTAT
0.24
0.01
-0.08
-0.18
0.03
-0.02
0.10
0.13
0.06
0.02
0.01
• Growth opportunities generated both by the management of the
individual entities in the BCE structure and by the ability of BCE
management to develop or acquire profitable subsidiaries or
portfolio investments.
• The negative impact of agency costs.
• A negative impact due to the "holding company discount".
The holding company discount is often referred to as discount
observed on closed-end investment funds, where such funds usually have a
market price lower than the net asset value of their holdings. The size of
the discount is variable and, at times, the closed-end fund sells at a premium
to its net asset value. As noted earlier, BCE management is not involved in
the management of the businesses, but rather in the management of the
assets of the overall firm and the investment of the cash flows generated by
the constituent parts. To the extent that shareholders view BCE's manage-
ment as a manager of the assets of the individual companies similar to a
portfolio investment, there is an analogy to the closed-end fund.
Can the effect of growth opportunities and the presence of agency costs
exist simultaneously? While the two components can co-exist, it is the net
result of the two factors that is important. For example, at the BCE (i.e.,
holding company) level, the agency cost argument suggests that management
consumes a large amount of perquisites and its investment decisions need not
be optimal. The free cash flow theory proposes that investment in diversification
activities or other areas will be excessive and beyond the optimal amount.
Therefore, even if some of the investments do follow the net present value
269
HALPERN & JOG
rule and generate above-normal profits, the possible use of free cash flow to
finance non-value-maximizing investments and the expectation by the market
of continued future behaviour of this type will have a negative effect on the
market value of the equity. The net result in such circumstances could be a
negative impact on share prices.
Is the holding company discount a factor separate from the influence of
agency costs? It is difficult to explain the perception that a closed-end invest-
ment fund - issuing equity claims and purchasing a set of securities - results in
a loss of wealth. The fact that many of these funds have not been "opened",
even though doing so would improve investors' wealth, suggests either that
the market value of the fund does not reflect its true value (if a sale of the
underlying securities were undertaken) or that there may be constraints to
removing the existing management. In the latter case, it could be that man-
agement is entrenched and making decisions which are either reducing wealth
through excessive perquisite consumption or not investing the cash flows from
the securities in an optimal way. Barclay et al. (1993) find that the larger the
managerial (or managerial affiliated) blocks in closed-end funds, the larger the
discounts to net asset value tend to be. They observe that the average discount
over their sample period for closed-end equity funds with management affiliated
blocks is 15 percent, and for those without blocks the discount is 8.2 percent.
They only include blocks which have not been accumulated for the purpose
of opening the fund.
Our search for equivalent Canadian closed-end investment funds has
been disappointing. We could find only one closed-end fund, that can be truly
defined as a fund, which holds primarily Canadian common stocks. For 1993
this fund had a discount of 17 percent.
14
Barclay et al. (1993) attribute this discount to an agency cost that is
based on benefits accruing to managers of these funds who, by having large
blocks, can entrench themselves and receive private benefits. These benefits
include the earning of management fees, fees for purchasing research, commis-
sions on securities transactions, acquisition of services from related companies
and the employment of relatives and friends.
15
 In the holding company context,
the discount, if it exists, can be associated with the agency costs that we have
identified. As noted previously, the holding company provides an internal
capital market due to the generation of cash flows; to the extent that these
cash flows are stable, the company is able to issue debt to finance its activities.
Freed from direct equity market monitoring, management can engage in
diversionary behaviour leading to a reduction in value. There is, however, no
literature that documents the size of holding company discounts and whether
they change over time. Although we believe that the holding company discount
is really a manifestation of the agency cost issue, we include it in our analysis
as a separate variable.
270
BELL CANADA ENTERPRISES: WEALTH CREATION OR DESTRUCTION?
THE VALUATION OF COMPONENTS AND THE DISCOUNT
THE TASK OF VALUING INDIVIDUAL SUBSIDIARIES of BCE and the comparison of
their aggregate value to the overall BCE value is not easy for a variety of reasons.
To provide a perspective on this difficult task, it should be noted that there are
a large number of companies in the BCE structure, including wholly owned
subsidiaries, majority owned subsidiaries, and portfolio investments both in
Canada and abroad. For example, in 1993 there were 142 subsidiaries, of
which 94 had BCE ownership greater than 90 percent, 20 were majority
owned, and 28 were portfolio investments. This compares with the 1983
total of 73 subsidiaries, of which 46 were wholly owned, 7 majority owned,
and 20 portfolio investments.
16
In addition, both the assets and revenues of these subsidiaries and portfolio
investments have changed dramatically, since they are affected by BCE's
acquisitions and divestitures. The assets increased dramatically in 1989 and
fell in 1993, reflecting the Montreal Trust purchase and divestiture, respectively.
Prior to 1988, wholly owned assets, as a proportion of total telecommunications
and other assets, ranged from approximately 3 percent to 11 percent of total
BCE assets. From 1989 to 1992, the proportion increased to approximately 42
percent, reaching a maximum of 47 percent in 1992. The value in 1993 fell to
18 percent following the major diversifications of that year. Although the
wholly owned non-financial subsidiaries have generated approximately $1 billion
in revenue and now have over $4 billion in assets, they have generated only
small or negative earnings.
BCE also has some significant holdings in wholly owned subsidiaries.
Some of the subsidiaries are actually larger than many medium-size Canadian
companies. For example, Tele-Direct had a revenue of $497 million in 1993
and employed 2,700 people worldwide. There are also a variety of minority
investment holdings of BCE such as Jones Intercable, Mercury Communications,
Encom and Videotran, among others.
To estimate the value of the components of BCE, we have grouped these
subsidiaries and portfolio investments into three major classes — Bell Canada,
subsidiaries with publicly traded equity in Canada, and all other wholly owned
subsidiaries. Since Bell Canada is wholly owned, an estimate of value must be
inferred by examining companies that are comparable in terms of risk. Since the
companies in the second class are publicly traded, the market value of this class
equals the price per share of each of the publicly traded subsidiaries multiplied
by the number of shares of each company owned by BCE. These subsidiaries,
along with BCE's interest in them and their estimated market values, are shown
in Table 8. Apart from Bell Canada, we have not valued any of the wholly
owned subsidiaries belonging in the third class, but assume initially that their
values are zero. However, given the poor performance of the diversification
activities described thus far, it is possible that the wholly owned subsidiaries
have had a negative influence on the share value of BCE, or that the market,
271
HALPERN & JOG
TABLE 8
BCE's PUBLICLY TRADED SUBSIDIARIES
PERCENTAGE OWNED
1983 1984 1985 1986 1987 1988 1989 1990
BCE Development Co. - - 67.9 65.8 67.0 67.0 67.0 67.0
BCE Mobile
Communications ____ _ 80.6 73.0 69.7
B F Realty Holdings _______ _
Bruncor Telephone Co. - - 30.7 31.2 31.0 31.2 31.2 31.4
Encor _____ _ 48.9 67.0
Maritime Telegraph 32.5 31.4 31.4 32.0 32.5 33.1 33.7 33.8
Memotech Data - - - - - 31.6 30.8 31.5
New Brunswick
Telephone Co . 32.5 31.3 _____ _
Newfoundland
Telephone Co . 54-4 53.3 _____ _
NewTel Enterprises - - 53.4 54.2 54.9 55.5 56.0 55.7
Northern Telecom 51.9 51.9 52.0 52.3 52.5 52.8 53.1 53.1
Quebecorlnc. - - - - - 21.2 21.2 21.6
TransCanada
Pipelines 42.3 47.2 47.8 48.5 49.3 49.1 48.9 24.4
with limited information on these subsidiaries, has placed a discount on the
implied value of the non-traded subsidiaries.
17
 The approach, then, is to compare
the overall value of the BCE as reflected in its market price to the estimated
value of the first two classes of subsidiaries, assigning a zero value to all other
wholly-owned subsidiaries except Bell Canada. Since both the market value of
the publicly-traded subsidiaries and BCE's ownership of them is known, the
main challenge is to estimate the value of Bell Canada.
In order to estimate the implied value of Bell Canada, we identified a set
of comparable companies that are in a similar business to Bell Canada,
although they are certainly smaller. Recently, the introduction of long
distance competition has affected all of the companies, but not equally, since
they do not have the same ratios of long-distance to local exposure. The
former influence may result in a lower discount rate if size is inversely related
to risk; while the latter influence results in higher risk and hence a higher
272
BELL CANADA ENTERPRISES: WEALTH CREATION OR DESTRUCTION?
TABLE 8 (CONT'D)
BCE'S SHARE OF SUBSIDIARY MARKET VALUE ($ MILLIONS)
1983
BCE Development Co. -
BCE Mobile
Communications -
BF Realty Holdings -
Bruncor Telephone Co. -
Encor -
Maritime Telegraph 73
Memotech Data —
New Brunswick
Telephone Co. 74
Newfoundland
Telephone Co. 73
NewTel Enterprises -
Northern Telecom 2,864
Quebecor Inc. -
TransCanada
Pipelines 595
Total 3,680
BCE Market
Value ($millions) 7,204
% of BCE
Market Value 51.0
1984 1985 1986 1987 1988
236 271 252 345
- - - - 1,002
---- -
102 100 102 104
- - - - 0
90 101 108 114 123
112
80
76
93 94 100 108
2,702 2,950 2,679 2,747 2,502
66
968 1,008 980 999 1,107
3,916 4,489 4,231 4,313 5,467
8,289 10,365 9,850 10,169 10,786
3 47.24 43.31 42.95 42.41 50.69
1989
135
1,493
-
105
234
163
115
_
_
136
3,444
49
1,269
7,143
13,783
51
1990
14
832
-
108
187
164
93
_
_
127
4,197
51
640
6,414
12,084
82 53.08
discount rate. How these two influences affect the value of Bell Canada is
difficult to determine; and so we assume that they net out.
Table 9 shows valuation ratios for five telecommunications companies -
BC Telecom, Bruncor, Maritime Tel&Tel, Newtel Enterprises, and Telus. For
each of these companies the following three ratios are calculated for the years
1985 to 1993: market value of equity to earnings, market value of equity to
revenue, market value of equity to assets. By applying the ratio for the average
of these companies (and excluding some obvious outliers due to write-offs and
highly unusual items) to the appropriate Bell Canada financial variable for
273
HALPERN & JOG
TABLE 9
ESTIMATION OF IMPLIED BELL CANADA MARKET VALUES 1985 - 1993
1985 1986 1987 1988 1989 1990 1991 1992 1993
MARKET-VALUE-TO-EARNINGS RATIO
BC Telecom 10.2 10.1 8.66 8.88 10.5 10.5 12.4 10.7 13.6
Bruncor 12.5 10.9 10.4 10.5 61.8 10.9 11.0 11.2 55.8
Maritime
Tel & Tel 9.59 10.5 9.08 8.98 10.0 8.65 9.9 9.25 11.5
Newtel
Enterprises 9.86 9.79 9.51 9.5 9.51 11.4 11.7 10.1 12.5
Telus
Corporation 000 0 0 25.4 11.9 10.2 12.3
Average 10.5 10.3 9.42 9.46 10 10.4 11.4 10.3 12.5
MARKET-VALUE-TO-SALES RATIO
BC Telecom 0.95 0.98 0.88 0.87 1.11 1.1 1.29 1.08 1.32
Bruncor Inc. 1.38 1.27 1.21 1.14 1.07 1.02 1.07 1.2 1.42
Maritime
Tel & Tel 1.05 1.02 0.99 0.98 1.08 1.0 1.11 1.09 1.26
Newtel
Enterprises 1.27 1.16 1.1 1.04 0.95 0.88 1.1 1.1 1.34
Telus
Corporation 0 0 0 0 0 1.52 1.76 1.52 1.76
Average 1.16 1.11 1.04 1.01 1.05 1.11 1.27 1.2 1.42
MARKET-VALUE-TO-ASSETS RATIO
BC Telecom 0.39 0.42 0.42 0.44 0.54 0.57 0.67 0.56 0.71
Bruncor Inc. 0.6 0.55 0.54 0.49 0.46 0.42 0.43 0.48 0.58
Maritime
Tel & Tel 0.47 0.45 0.43 0.41 0.43 0.4 0.45 0.43 0.48
Newtel
Enterprises 0.55 0.48 0.47 0.33 0.39 0.35 0.44 0.43 0.55
Telus
Corporation 000 0 0 0.58 0.72 0.57 0.69
Average 0.5 0.47 0.46 0.42 0.45 0.46 0.54 0.49 0.6
274
BELL CANADA ENTERPRISES: WEALTH CREATION OR DESTRUCTION?
TABLE 9 (CONT'D)
VALUES FOR BELL CANADA ($ MILLIONS)
1985 1986 1987 1988 1989 1990 1991
Operating Revenue
5,769 6,255 6,378 6,624 7,273 7,655 7,729
Net Income
(after extra-ordinary items)
652 711 731 793 875 966 986
Total Assets
11,941 12,693 13,508 14,512 15,699 16,605 17,163
1992 1993
7,863 7,957
1,006 871
18,414 18,945
IMPLIED BELL CANADA VALUE BASED ON :
Market Value-to-Earnings Ratio
6,868 7,346 6,886 7,502 8,758 10,007 11,213
Market Value-to-Sales Ratio
6,712 6,930 6,651 6,669 7,650 8,465 9,787
Market Value-to- Assets Ratio
6,024 6,024 6,274 6,065 7,125 7,695 9,267
Average 6,535 6,776 6,603 6,745 7,844 8,722 10,089
10,352 10,869
9,414 11,305
9,090 11,364
9,619 11,179
each year from 1985 to 1993, we obtain three estimates of the implied value of
Bell Canada. These values are shown at the bottom of Table 9. The implied
value for Bell Canada for any year is equal to the average of the three estimated
values for Bell Canada, using the different valuation ratios.
In Table 10, the implied equity values for Bell, publicly traded subsidiaries,
and the sum of these two values are presented for the years 1985 to 1993. The
sum of these two values is the implied value of BCE, assuming that the non-
traded, wholly owned subsidiaries of BCE have no value. The corresponding
actual market values of BCE equity are shown in column 5 of the table. To
determine if the actual market value is above or below the implied value we
calculate the discount, defined as the actual market value minus the implied
value, both in dollar terms and as a percent of the implied value. As can be
seen in the table, the dollar discount is always negative and ranges from a low
in 1985 of $675 million to a high in 1992 of $6.09 billion. The percent discount
had its lowest value of 6 percent in 1985 and its highest value of 32 percent in
1992. The discount value for 1993 is 25 percent.
275
HALPERN & JOG
TABLE 10
ESTIMATE OF THE UNDERLYING BCE MARKET VALUES AND DISCOUNTS ($ MILLIONS)
YEAR
1985
1986
1987
1988
1989
1990
1991
1992
1993
ESTIMATED
VALUE OF
BELL
6,535
6,767
6,604
6,746
7,845
8,723
10,090
9,620
11,180
VALUE OF
PUBLICLY
TRADED
SUBSIDIARIES
4,505
4,263
4,368
5,480
7,154
6,425
8,596
9,141
7,895
BCE
IMPLIED
VALUE
11,040
11,030
10,972
12,226
14,998
15,147
18,686
18,761
19,075
BCE
ACTUAL
MARKET
VALUE
10,365
9,850
10,169
10,786
13,783
12,064
14,779
12,672
14,252
DISCOUNT
(ACTUAL
IMPLIED)
-675
-1,180
-803
-1,440
-1,216
-3,084
-3,907
-6,089
-4,822
DISCOUNT
(%)
-6.11
-10.70
-7.32
-11.78
-8.11
-20.36
-20.91
-32.46
-25.28
REDUCTION
IN BELL
IMPLIED
VALUE FOR
BREAKEVEN (%)
10.33
17.44
12.16
21.35
15.50
35.35
38.72
63.30
43.14
Obviously, one possible explanation of the discount is that the value of
Bell Canada has been incorrectly calculated and is too high. If this were true,
we consider the size of the estimation error that would be necessary in order to
make the implied and actual market values of BCE the same. In the last column
of Table 10, the overestimate of the implied Bell value is calculated as a
proportion of the implied value. In order to break even, the value of the
implied Bell market value would have to be reduced by approximately 43 per-
cent in 1993 and 64 percent in 1992: the lowest value is 10 percent in 1985.
While it is possible that we may have overestimated the value of Bell, it is
unlikely that the differences would be as large as are required to achieve the
breakeven value. In addition, we have assumed a zero value for the wholly
owned subsidiaries — whic h is clearly incorrect. In 1993, these subsidiaries
accounted for almost $800 million in revenue and $4 billion in assets. This
assumption therefore reduces the implied value of BCE; the result should be
a premium, not a discount, when the actual and implied market values are
compared.
In contrast to some evidence consistent with improved value-creation
performance noted in the previous section, the size of the implied discount
indicates that the discount has actually increased over time. Some of it may be
the result of (our) undervaluation of Bell Canada. However, it should be
emphasized that in Table 10 we have assigned absolutely no value to the wholly
owned subsidiaries of BCE. If there is any value in these subsidiaries, then our
analysis would indicate an even higher discount associated with BCE shares -
276
BELL CANADA ENTERPRISES: WEALTH CREATION OR DESTRUCTION?
an observation consistent with the agency cost arguments of free cash flow,
small managerial stakes, entrenchment, and internal capital markets through
the holding company structure.
SUMMARY AND CONCLUSIONS
T
HE UNDERLYING PURPOSE OF THIS STUDY is to provide an analysis of the
effects of agency costs and corporate governance on corporate value. BCE
was used as an example in our analysis because it represents the stereotype of a
widely held large firm where monitoring by the board of directors and share-
holders is unlikely to be effective. In addition, entrenchment of management
exists because of the large size of the company and constraints on foreign
ownership of Bell Canada. Another interesting aspect of BCE is the importance
of free cash flow and the holding company structure of the company with
respect to agency costs.
Our analysis of BCE is consistent with the intuitive implications of the
agency costs argument: if managers have a very small equity stake in the
company, there is a greater chance that dispersed external shareholders may
place a lower value on the firm. Although the board of directors is expected to
monitor management performance and reduce any misalignment of interests,
it is very likely that the board's efforts in this connection may not be effective.
Our analysis also indicates that the performance of BCE is not as stellar
as it could be in the light of its changing corporate strategy, its historical
propensity to make diversification investments in unrelated areas, and its
continued holding-company mentality of managing assets and not maximizing
shareholder value. Although a return to its core business has been evident
since 1990, our analysis also indicates that the market continues to place a
substantial discount on BCE shares. This discount can be explained by the
market's expectation of management investment of free cash flows generated
primarily from the regulated telecommunications company into wholly owned
and portfolio investments in related and quasi-related businesses across the world.
What changes are necessary to increase the value of BCE and thereby
reduce the discount we have observed? In order for changes to have any effect
on the discount, the market must anticipate that those changes will reduce the
agency costs and improve the efficiency of BCE. One approach is to improve
the monitoring of management by having a more active board on which there
are outside representatives. This is the crux of the TSE guidelines on the
composition of the board, although we would require that "independent"
directors not be so defined if they are associated with any subsidiary and/or
affiliated BCE company. However, while this change in board structure is
necessary, it is not sufficient to generate the changes that may be required. Even
with the best structured board, its effect on operations will still depend upon the
strength it shows in its interactions with the CEO and inside directors. A strong
CEO can control even the strongest and best intentioned group of outside board
277
HALPERN & JOG
members. Also, without a direct linkage between the outside directors' compen-
sation to shareholder wealth, the influence of the board may well be muted.
A second approach to unlocking this discount would be to reduce BCE's
ownership in wholly owned subsidiaries. This action may generate a more
transparent method for valuing individual components within the BCE
structure. However, this is unlikely to reduce the discount dramatically for two
reasons. First, we have measured the discount by assuming that the subsidiary
values are zero. If the subsidiary values were positive, however, the discount
would actually be larger, not smaller. Second, the conversion of wholly owned
subsidiaries to partially owned subsidiaries with publicly traded equity will not
remove the holding company discount, since BCE will still be valued at a
discount similar to a closed-end investment fund.
A third approach would be to sit back and wait for the discipline of the
product market to take effect in the markets in which BCE subsidiaries and
investments sell their goods and services. Some of the subsidiaries and partially
owned companies are in competitive industries and the product market will
have an effect here. However, as long as cash flows are generated from other
parts of the holding company, BCE can subsidize money-losing operations in
competitive markets.
At present, Bell Canada is BCE's major subsidiary and under the current
regulatory structure it is a regulated monopoly which has only modest incentives
to become more efficient and large incentives to acquire assets and increase
the size of its rate base. With the arrival of competition in supplying long distance
telecommunications services, there has been a push to improve the efficiency
of some operations. The most recent action is the elimination of a large number
of jobs. Further, if price-cap regulation is introduced as an alternative to the
current rate-base, rate-of-return form of regulation, there will be an added
incentive for Bell Canada to improve its efficiency and hence its cash flows.
However, those larger cash flows will ultimately be channelled to BCE and we
are left with the same problem we started with - dissipation of wealth through
uneconomic investments.
We are caught in a dilemma. On the one hand we argue that BCE share-
holders will have their share value maximized if BCE spins off or sells off its
non-telecommunications assets and returns to a pure telecommunications
company, which may or may not include the long-distance aspects of its business.
This could be equivalent to the U.S. decision under which AT&T was broken
into local and long-distance companies. However, in this case the impetus is
shareholder maximization and not purported efficiencies due to competition.
On the other hand, there may be economies of scope which arise from the
operation of a number of related companies. This factor suggests a spin off/sell
off of some companies, but retention of the telecommunications companies.
Shareholder value would be maximized if the retained companies were managed
more effectively and the emphasis on asset management were eliminated.
However, the unanswered question is whether BCE management can manage
278
BELL CANADA ENTERPRISES: WEALTH CREATION OR DESTRUCTION?
this resulting company to create value, perhaps assisted by a more active board
of directors. From our analysis, BCE management has historically been
unsuccessful in coming even close to reaching this objective.
ENDNOTES
1 The popularity of the agency theory approach to analyzing firm value is
generally attributed to the Jensen-Meckling paper in 1976, although similar
arguments can be traced to Veblen (1924) and Manne (1965),
2 Technically, regulation is on the regulated operations of Bell Canada,
thereby removing some of the competitive operations from regulatory
control. The introduction of competition in long distance and proposed
changes in regulation could have an effect on cash flows from Bell
Canada.
3 Since there is voluminous literature in this area, we refer to it only briefly.
The seminal papers are: Manne (1965), Jensen & Meckling (1976), Myers
& Majluf (1984), Demsetz & Lehn (1985), Jensen (1986), and Hoshi et al
(1989).
4 See Jensen &. Meckling (1976), Jensen (1986), and Elitzur & Halpern
(1995).
5 However, Holderness 6k Sheehan (1988) also state on page 319,
We note at the outset that in light of the current lack of knowledge about
large-block shareholders and the paucity of precise hypotheses about them in
the literature, this paper is as much a descriptive analysis as it is a test of
specific hypotheses on the role of majority shareholders.
6 See Morck, Shleifer & Vishny (1989), Weisbach (1988), and Warner,
Watts &Wruck( 1988).
7 In light of the recent and very large leveraged buyouts (LBOs) and takeovers,
it may be argued that no company is now immune from being taken over.
However, BCE is quite different from many other large companies which
have been taken over or have successfully completed an LBO transaction.
Unlike these companies, in the case of BCE, there is a large degree of
uncertainty about the potential response of the regulators of its telephone
subsidiary, Bell Canada, and about its holdings of other regulated tele-
phone companies. In addition, there is a legislated cap on the proportion
of equity ownership of a telecommunications company by a foreign company.
This would eliminate the incentive for a foreign firm, which could have
access to capital markets to finance a takeover of BCE, to engage in such a
takeover.
279
HALPERN & JOG
8 This number includes only those which are classified as a level 2 subsidiary
by Statistics Canada in its Intercorporate Ownership Statistics. A level 2
subsidiary is defined as one on which the parent exerts (can exert) direct
control. This number, therefore, excludes those companies which are in
turn owned by these fifteen subsidiaries. We have also excluded "numbered"
subsidiaries of BCE in this count.
9 Earnings are defined as operating earnings before interest and taxes.
10 For 1993 the after-tax net income contribution from the equipment manu-
facturing segment is -$588 million (loss), reflecting the serious problems
faced by Northern Telecom.
11 These abnormal returns and their t-statistics are based on standard event
methodology using mean returns estimated over 120 days, ending 30 days
prior to the actual event date as the expected return for calculating the
abnormal returns on the event date. Daily data is used in all cases and CARs
are reported for the (—5 +5) window with event date termed as day zero.
12 Note that BCE paid $400 million for a 20 percent stake in Telemax in
1990. The recent devaluation of the Mexican peso may imply a possibility
of a write-down of this investment similar to the one recently announced
by Labatt's.
13 The cumulative wealth index is defined as the net cumulative value of $1
invested net of the risk-free rate. Cumulative excess returns is calculated
using the market model with the TSE 300 representing the index. The
regression period used in these calculations is a 60-month moving period,
updated monthly.
14 The discount is a sufficiently important problem that at least one manage-
ment has tried to reduce it through innovative means. Canadian General
Investment Ltd. decided to list its shares on the London Stock Exchange,
where the average discounts are 7 percent and 12 percent for single country
funds. The discount at the announcement date of February 13, 1995 was
28 percent. Management anticipated a reduction of the discount to 15 per-
cent. As of February 25, 1995, the discount was 26 percent; and the shares
will be listed in London at the end of February, 1995. See Report on
Business, February 14, 1995, pp. B13.
15 For a discussion of the literature on agency costs associated with high owner-
ship and its effect on equity value see Daniels & Halpern (forthcoming,
1995).
16 See Statistics Canada, Catalogue number 61-517, various years.
17 In its 1993 annual report, BCE lists the value of its investment in these
wholly owned subsidiaries and its minority interests. These amount to
approximately $2 billion.
280
BELL CANADA ENTERPRISES: WEALTH CREATION OR DESTRUCTION?
ACKNOWLEDGEMENT
W
E WOULD LIKE TO THANK the Financial Research Foundation of Canada
and the Social Sciences and Humanities Research Council for providing
partial funding for this study. We also wish to thank James Blake for his
extremely valuable research assistance.
BIBLIOGRAPHY
Akerlof, G. A. "The Market for 'Lemons': Qualitative Uncertainty and the Market
Mechanism." Quarterly journal of Economics, 84 (August 1970):488-500.
Barclay, M., C. Holderness and J. Pontiff. "Private Benefits from Block Ownership and
Discounts on Closed-end Funds." Journal of Financial Economics. (June, 1993):263-91.
Daniels, R. and P. Halpern. "The Role Of The Closely Held Public Corporation In the
Canadian Economy and the Implications for Public Policy." Canadian Business Law
journal. (Forthcoming, 1995).
Demsetz, H. and K. Lehn. "The Structure of Corporate Ownership: Causes and
Consequences." Journal of Political Economy, 93 (1985):1155-77.
Elitzur, R. and P. Halpern. "Executive Compensation and Firm Value." Paper presented at
the Conference on Corporate Decision-making in Canada, March 20-21, 1995,
Toronto, Canada.
Holderness, C. G. and D. P. Sheehan. "The Role of Majority Shareholders in Publicly Held
Corporations." Journal of Financial Economics, 20 (1988):317-46.
Hoshi, T., A. Kashyap, and D. Scharfstein. "Corporate Structure, Liquidity, and
Investment: Evidence from Japanese Industrial Groups." Quarterly Journal of Economics,
CVI( 1989): 33-60.
Jensen, M. C. "Agency Costs of Free Cash Flow, Corporate Finance, and Takeovers."
American Economic Review, 76 (1986):323-29.
Jensen, M. C. and W. Meckling ."Theory of Firm: Managerial Behavior, Agency Costs and
Ownership Structure." Journal o/Financial Economics, 7 (1976):305-60.
Jensen, M. C. and J. B. Warner. "The Distribution of Power Among Corporate Managers,
Shareholders, and Directors." Journal of Financial Economics, 20 (1988):3-24.
Manne, H. "Mergers and Market for Corporate Control." Journal of Political Economy, 73
(1965):! 10-20.
Morck, R., A. Shleifer, and R. W. Vishny. "Alternative Mechanisms for Corporate
Control." American Economic Review, 79 (1989):842-52.
Myers, S. C. and N. S. Majluf. "Corporate Financing Decisions When Firms Have
Investment Information that Investors Do Not." Journal of Financial Economics, 13
(1984):187-221.
Shleifer A., and R. W. Vishny. "Large Shareholders and Corporate Control." Journal of
Political Economy, 94 (1986):461-88.
Veblen, T. The Engineers and the Price System, New York: Viking, 1924.
Warner, J. B., R. L. Watts, and K. H. Wruck. "Stock Prices and Top Management
Changes." Journal of Financial Economics, 20 (1988):461-92.
Weisbach, M. S. "Outside Directors and CEO Turnover." Journal of Financial Economics, 20
(1988):431-60.
28 J
HOLDERNESS/MEHROTRA/GILL
Clifford G. Holderness
Wallace E. Carroll School of Management
Boston College
Commentary on Part III
Large-Block Shareholders and Conglomerates
Sense or Nonsense?
INTRODUCTION
T
HE PARADIGM THAT HAS DOMINATED RESEARCH in financial economics for
over 60 years is that of the widely held corporation. Yet, we now know
that most public corporations, even many large ones, have shareholders who
own high-percentage blocks of common stock. For example, Mikkelson &
Partch (1989) find that in a random sample of 240 NYSE and Amex industrial
corporations, officers and directors on average control 20 percent of the votes
(median 14 percent). Similarly, Holderness & Sheehan (1988) report that 11
percent of all NYSE and Amex corporations have majority shareholders who
control at least 50 percent of the common shares.
Ownership of Canadian corporations is even more concentrated than
that of American corporations. In their study elsewhere in this volume, Rao
and Lee-Sing classify only 23 percent of Canadian corporations as being widely
held, compared with 40 percent of American corporations. I believe the
Canadian situation to be more "natural" than that in the United States
because a series of populist laws passed by Congress in the 1930s disenfran-
chised natural large-block shareholders, specifically financial institutions
(Roe, 1990). Despite (or perhaps because of) these laws, evidence shows that
since the 1930s the ownership of American corporations has become signifi-
cantly more concentrated.
1
The relationship between blockholders and conglomerates is one that
has been overlooked in the rapidly growing literature on ownership structure.
This potentially important relationship is, however, addressed in two studies
presented at this conference on Corporate Decision-Making in Canada: "In
High Gear: A Case Study of the Hees-Edper Corporate Group" by Ron
Daniels, Randall Morck and David Stangeland; and "Bell Canada Enterprises:
Wealth Creation or Destruction?" by Paul Halpern and Vijay Jog. These studies
address the role of blockholders in the context of two very different companies,
albeit with some important similarities. The Hees-Edper Group is controlled
by the Bronfman family. Hees-Edper, in turn, owns high-percentage blocks of
stock in other corporations. BCE, the largest Canadian corporation, holds
dozens (and at times hundreds) of high-percentage stakes in other companies.
282
DISCUSSANTS' COMMENTS ON PART III
Both Hees-Edper and BCE hold blocks in companies in diverse businesses. In
short, both are conglomerates with blockholder features.
My comments here will focus on the relationship between blockholders
and conglomerates generally, and on these two studies specifically.
CHANGING VIEWS OF ECONOMISTS ON
CONGLOMERATES
E
CONOMISTS' VIEWS OF CONGLOMERATES have changed significantly over the
last 25 years.
2
 In the 1960s and 1970s economists largely applauded the
efforts of corporations to expand into unrelated fields. A number of arguments
were advanced in support of such expansion. Conglomerates could allocate
capital more efficiently than markets; conglomerates could do a better job of
identifying and developing promising young managers; and conglomerates
could take cash generated by mature industries and funnel it to promising
growth areas, solving what has become known as the free-cash flow problem.
By the mid-1980s this attitude had changed, and conglomerates were not
held in such high regard. In the United States several highly publicized hostile
takeovers were aimed at companies that had diversified. Upon gaining control,
the acquirer would often sell divisions unrelated to the company's core operations.
At the same time, academic studies were appearing to confirm that a focused
firm typically creates more value for its shareholders than does an unfocused
conglomerate.
3
Although the current opinion of academics and the financial press is
anti-conglomerate, there are nonetheless some glaring exceptions, notably
active large-block shareholders, such as Warren Buffett and Laurence Tisch.
Although such individuals are almost universally classified as active investors
(see Jensen,1993), they could also be classified as conglomerate investors.
Buffett, for example, counts among his investments large blocks of stock in
transportation (US Air), newspapers (Washington Post), financial services
(Salomon Brothers), and food (See's Candy). Tisch controls an insurance
company (CNA), a movie theater chain (Loews), and a television network (CBS).
Are these active investors to be praised for their conglomerates, while
others are criticized? Are conglomerates somehow more effective when con-
trolled by an active investor? To answer such questions - questions which are
important both in Canada and the United States - we need a theory of
conglomerates. The studies on Hees-Edper and BCE go beyond praise or criticism
of conglomerates. Both studies seek to understand conglomerates in a deeper
sense. My remaining comments offer suggestions for extending these studies and
thus furthering our understanding of the relationship between conglomerates
and blockholders.
283
HOLDERNESS/MEHROTRA/GILL
TOWARD A THEORY OF CONGLOMERATES
AND BLOCKHOLDERS
INCENTIVES OF TOP MANAGEMENT
INDIVIDUALS, NOT CORPORATIONS MAKE DECISIONS. Consequently, to under-
stand any organizational form, we must examine the punishment and reward
structure confronting top decision-makers. Both studies pursue this issue.
Halpern and Jog identify a major change in policy at BCE in the early
1990s to refocus operations. During the 1980s BCE had diversified into a large
number of industries, many of which had little, if anything, to do with BCE's
primary operations in communications. This policy changed in the early
1990s, coincidental with the arrival of a new CEO, Raymond Cyr. I would like
to see this issue explored further. Why was there a change in CEO? Were both
CEOs compensated in the same manner? How was their compensation struc-
tured? In particular, was the new CEO's compensation package more oriented
toward incentive-based compensation (bonuses and stock options)? Was the
change to a more focused operation at BCE actually the result of Cyr's efforts?
Daniels, Morck and Stangeland also explore the incentives of top
executives. They claim that the Hees-Edper Group uses "extreme incentive-
based compensation schemes ... to pay managers, and in some cases employees".
4
They further state that Hees-Edper's top managers faced asymmetric incentives:
if their gambles paid off, the managers profited personally; if their gambles
failed, on the other hand, the managers did not suffer the consequences. This
sounds like the incentives confronting savings and loan executives in the
United States during the 1980s: win with a risky investment and you get the
upside; lose with a risky investment and the government (through its guarantee
of deposits) is left with the loss.
The compensation scheme at Hees-Edper is worthy of additional investi-
gation. The authors cite only one newspaper article as proof of the asymmetric
compensation system. Stronger and more definite information about the
precise nature of the compensation would be helpful. For both studies I would
also like to see investigations of top management turnover. Is it related to firm
performance? Is it related to the composition of the board of directors? Recent
papers by Rose & Shepard (1994), Mehran (1995), and Dial & Murphy
(1995) offer examples of how the authors of the BCE and Hees-Edper studies
could expand upon their already-illuminating discussions of executive
compensation.
THE ROLE AND INCENTIVES OF THE BOARD OF DIRECTORS
THE BOARD OF DIRECTORS IS CLOSELY RELATED to the top management team.
The board has the legal power to manage the corporation, a right which it
delegates, but does not transfer outright, to top managers. Halpern and Jog
284
DISCUSSANTS' COMMENTS ON PART III
report that board membership at BCE fell from 21 in 1983 to 15 in 1993. This
is interesting given that it has been argued that small boards are typically more
effective than large boards. Some commentators maintain that a board with
fewer than 10 members is optimal, even for the largest corporations. Halpern
and Jog, however, do not explore why the reduction in board size at BCE took
place, nor do they discuss the implications of the change on firm performance.
Neither do Daniels et al. discuss boards within the Hees-Edper Group or at the
Hees-Edper portfolio companies.
I would like to see both studies investigate the role of the board of directors.
Boards need not be ineffective in the presence of a large-block, even majority,
shareholder. For example, there is research that analyzes the serious constraints
placed by a board of directors on a domineering majority shareholder — in this
case, Ted Turner of Turner Broadcasting (Holderness &. Sheehan, 1991). Do
directors in the Hees-Edper Group have such powers? Similarly, do directors of
companies in which BCE owns large blocks have such powers? What is the
role of the Bronfman family on the Hees-Edper boards?
Similarly, it would be illuminating to determine the relationship
between top executive turnover and firm performance. Here the work of
Weisbach (1988) is potentially valuable. Investigations in this area should
recognize the importance of a three-way classification of directors into insiders,
grays (those who have business relations with the company, such as lawyers
and investment bankers), and true outsiders. The traditional classification of
insiders versus outsiders can miss important effects.
SHARED BENEFITS OF BLOCKHOLDER CONTROL OF CONGLOMERATES
THERE ARE TWO (NOT MUTUALLY EXCLUSIVE) REASONS why someone would
hold a high-percentage block of stock. First, a blockholder can use voting
power to improve management. The benefits related to this are often called
"shared benefits" of control because they accrue to all shareholders. Second, a
blockholder can use voting power to secure corporate benefits that do not
accrue to other shareholders. The related benefits in this case are often called
the "private benefits" of control.
There are several interesting research topics on the shared benefits of
blockholder control of a conglomerate. For example, one could hypothesize
that a blockholder would have a greater effect on the management of a con-
glomerate than on the management of a focused corporation. The explanation
might take this form: blockholders are primarily monitors; monitors have
generalized skills that are not limited to specific firms or industries; thus,
blockholders are most effective in conglomerates. Certainly, Warren Buffett
and Laurence Tisch seem to fit this profile.
Although this is a plausible explanation, I know of no supporting or,
for that matter, contradicting empirical evidence. The two studies under
discussion, however, touch upon potential shared benefits of blockholder
285
HOLDERNESS/MEHROTRA/G1LL
control of a conglomerate. BCE, for example, often buys complete control of
non-telecommunications companies but only buys minority control of companies
in its primary business of telecommunications. This pattern could reflect
regulatory constraints, or it could reflect the shared benefits of control.
The Daniels, Morck and Stangeland study offers little discussion of what
types of block investments are made by Hees-Edper, much less why they are
made. Development of this point could yield valuable insights.
Finally, it would be interesting to see how management in the target
firms changes after either BCE or Hees-Edper makes a block investment. The
authors could investigate the stock-price reaction when BCE or Hees-Edper
initially announces a large block investment. Similarly, the authors could
investigate operational changes and top management turnover following block
investments.
5
CONSTRAINING PRIVATE BENEFITS OF BLOCKHOLDER
CONTROL OF CONGLOMERATES
MANAGEMENT IMPROVEMENTS ARE THE POTENTIAL shared benefits for minority
shareholders of block investments. Consumption of certain types of private
benefits by the blockholder represents the downside for minority shareholders
of block investments.
6
 Simply put, what prevents a controlling shareholder
from maximizing expected cash flows and then expropriating those cash flows
through excessive compensation, perquisites, and transactions with the
corporation at non-market prices? There is evidence from several sources that
blockholders typically consume private benefits. This appears to explain why,
for example, trades of high-percentage blocks of stock are typically priced at
substantial premiums to the post-announcement exchange price (Barclay &
Holderness, 1989, 1992). Private benefits also appear to explain, at least in
part, why closed-end funds in the United States often trade at a discount to
net asset value (Barclay, Holderness & Pontiff, 1993).
There is also considerable empirical evidence that blockholders (at least
in the United States) are constrained in their power to consume private benefits
at the expense of minority shareholders. If this were not the case, blockholder-
controlled firms would eventually cease to exist as minority investors realized
the folly of investing in such firms. The evidence in the United States is that
the law, as opposed to organizational factors, constitutes the primary constraint
on dominant shareholders.
7
Finally, it is important to know how, if at all, Canadian law constrains
controlling shareholders such as BCE, the Hees-Edper Group, and the
Bronfman family. A related point would be to determine how Canadian
corporations adapt to controlling shareholders, perhaps by giving the board
additional powers.
286
DISCUSSANTS' COMMENTS ON PART III
CONCLUSION
T
HE ESSENCE OF PRIVATE PROPERTY is the collocation of wealth effects and
decision rights. The natural state of a market economy is to have owners
manage. For this reason, and over time, more corporations will have large-
block shareholders.
The relationship between blockholders and conglomerates is potentially
important and has received little attention to date. Investigations such as
those conducted by Halpern and Jog into BCE and by Daniels, Morck, and
Stangeland into the Hees-Edper Group will help us better understand this
intriguing relationship.
ENDNOTES
1 Preliminary research indicates that ownership of stock by directors and
officers of a randomly chosen American corporation has approximately
doubled in fractional terms since 1935. Holderness, Kroszner & Sheehan
(1995).
2 This issue is reviewed at length by Daniels et al.
3 See the papers in the journal of Financial Economics, Symposium on
Corporate Focus (January 1995).
4 Daniels et al, this volume.
5 For examples of specific types of investigations, see Barclay & Holderness
(1991).
6 Only consumption of pecuniary private benefits will reduce the wealth of
minority shareholders (Barclay &. Holderness, 1992).
7 See Holderness & Sheehan (forthcoming).
REFERENCES
Barclay, Michael J. and Clifford G. Holderness. "Private Benefits from Control of Public
Corporations." Journal of Financial Economics, 25, (1989):371-97.
. "Negotiated Block Trades and Corporate Control." Journal of Finance, 46,
(1991):861-78.
. "The Law and Large-Block Trades." Journal of Law & Economics, 35 (1992):265-94.
Barclay, Michael J., Clifford G. Holderness and Jeffrey Pontiff. "Private Benefits from Block
Ownership and Discounts on Closed-end Funds." Journal of Financial Economics, 33,
(1993):263-91.
Dial, Jay and Kevin J. Murphy. "Incentives, Downsizing, and Value Creation at General
Dynamics." Journal of Financial Economics, 37 (1995):261-314.
Holderness, Clifford G. and Dennis P. Sheehan. "Constraints on Active Investors," in
Relational Investing. Edited by John C. Coffee, Jr. and Ronald Gilson. Oxford: Oxford
University Press (Forthcoming).
. "Monitoring an Owner: The Case of Turner Broadcasting." Journal of Financial
Economics, 30 (1991 ):325-46.
287
HOLDERNESS/MEHROTRA/GILL
. "The Role of Majority Shareholders in Public Corporations: An Exploratory
Analysis." Journal of Financial Economics, 20 (1988):317-46.
Holderness, Clifford G., Randall Kroszner, and Dennis P. Sheehan. "Were the Good Old
Days Good? Changes in Managerial Stock Ownership since the Great Depression,"
unpublished working paper, Boston College, Chestnut Hill, MA, 1995.
Jensen, Michael C. "The Modern Industrial Revolution, Exit, and the Failure of Internal
Control Systems." Journal of Finance, 48 (1993):831-80.
Symposium on Corporate Performance. Journal of Financial Economics, 37 (1995): 1-126.
Mehran, Hamid. "Executive Compensation Structure, Ownership, and Firm Performance."
Journal of Financial Economics, 38 (1995):163-84.
Mikkelson, Wayne H. and M. Megan Partch. "Managers' Voting Rights and Corporate
Control." Journal of Financial Economics, 25 (1979):263-90.
Roe, Mark J. "Political and Legal Restraints on Ownership and Control of Public
Companies." Journal of Financial Economics, 27 (1990):7-41.
Rose, Nancy L. and Andrea Shepard. "Firm Diversification and CEO Compensation:
Managerial Ability or Executive Entrenchment?" Unpublished Working Paper,
Massachusetts Institute of Technology, Cambridge, MA, 1994.
Weisbach, Michael. "Outside Directors and CEO Turnover." Journal of Financial Economics,
20(1988):431-60.
288
DISCUSSANTS' COMMENTS ON PART III
Vilcos C. Mehrotra
Faculty of Business
University of Alberta
Commentary on Part III
B
OARDS OF DIRECTORS ARE MORE THAN INTERFACES between managers of
modern public corporations and their owners. In principle, they are the
appointed monitors of corporate management put in place by shareholders to
act in the latter's best interests. The reality of boards is quite different. Several
corporate declines could have been arrested if boards had been more vigilant
and assertive. Recent derivative-blamed losses at Metallgesellschaft and
Procter and Gamble, and even at Barings, could have been prevented had the
board(s) of those firms been more vigilant and in tune with the corporations'
strategic objectives and follow-on activities.
To be sure, boards are goaded into action when matters get really out of
hand. Witness the push at General Motors in 1992 that resulted in a major
overhaul of the board itself. But such examples of board assertiveness are rare
and tend to occur far too late. In GM's case, it was prolonged and mounting
inaction in the face of lacklustre market and operating performance that led
the board, studded with star directors, finally to replace the CEO. An ounce of
prevention, as the old saying has it, would have avoided pounds of cure - in
the form of real and opportunity losses at GM. Effective corporate governance
is about prescribing the right dose of "prevention". Too much interference
with line managers detracts from creative risk-taking; too little oversight leads
to loss of accountability with value-damaging consequences.
If we are to see any material change in board behaviour, the incentives and
controls that underlie its role will have to be overhauled. Modest steps in this
direction have already been taken. The Dey report (1994) sponsored by the
Toronto Stock Exchange (TSE), similar in spirit to the Cadbury report (1992) in
the United Kingdom, has several useful recommendations aimed at improving
board oversight and governing effectiveness. Appointing more "outside and
independent" directors to the board and splitting the roles of the Chairperson
(of the board) and the Chief Executive Officer (CEO) are two recurring themes
in the Dey and Cadbury reports. However, including more outsiders on the
board will not achieve the desired changes unless some additional measures
(such as the use of federal versus provincial legislation to effect these changes)
are considered. Whatever one's views of the efficacy of legislative changes, most
would agree on the need for consistency and uniformity of legislation across
different jurisdictions and over time. In this regard, replacing provincial
securities legislation with a national jurisdiction deserves close scrutiny.
289
HOLDERNESS/MEHROTRA/G1LL
The Dey report falls short of recommending legislative changes to
achieve most of its recommendations, relying instead on the coercive powers
of the TSE to incorporate the changes. Again, this follows the spirit of the
Cadbury report. Legislative changes, if adopted, must take into account
corporate governance issues unique to Canadian corporations, in at least one
important respect. In both of the other two major English-speaking
economies, the United Kingdom and the United States, the ownership
structure of companies is diffused throughout a large base of shareholders.
Canadian companies, on the other hand, tend to be closely held or, at least
closely controlled through pyramidal ownership arrangements. This has
pointed up a different sort of conflict among a corporation's stakeholders,
namely, the intra-shareholder conflict. Whereas in the United States a
potential conflict exists between debt and equity claims on corporate cash
flows, in Canada the issue is cast more in terms of dominant versus minority
shareholders, presumably because dominant shareholders can extract private
benefits from the corporation that are not available to minority shareholders
with otherwise similar dividend rights.
1
 Reallocation of private benefit consump-
tion can be more than merely redistributive — to the extent that it reduces ex
ante contracting costs with outside equity providers and thereby the cost of
equity capital, it can lead to increased real investment and therefore benefit
society.
What is needed is a mechanism to check the ability of controlling share-
holders to transfer wealth to themselves. To be sure, safeguards to protect
minority shareholders do exist — such as fairness-opinion legislation in the case
of minority buy-outs and, in extreme cases, judicial injunctions — which can
provide additional relief to minority shareholders. Furthermore, any change
that restricts the ability of controlling shareholders to extract private benefits
is likely to reduce the incentive of those shareholders to monitor manage-
ment, precisely because the benefits of such monitoring can no longer be
privately appropriated. Ultimately, the net benefit of imposing restrictions on
the activities of controlling shareholders remains an open empirical issue.
Having noted this caveat, one change that deserves scrutiny is board reform
designed to induce additional external monitoring of the firm's management.
The obvious question is who should take on this role?
Take the case of institutional shareholders, the one group that almost
everyone agrees is in a position to provide, and benefit from, corporate over-
sight and monitoring. This is especially true in the Canadian context where
corporate ownership structures, market size limitations, and foreign equity
holding restrictions render "voting with one's feet" less effective. Legislation
aimed at easing foreign-equity-holding restrictions may not directly increase
institutional monitoring (access to U.S. equities may even reduce the incentive
of Canadian institutions to engage in domestic monitoring), but it is likely to
put pressure on Canadian boards to deliver shareholder returns more in line
with those available internationally.
290
DISCUSSANTS' COMMENTS ON PART 111
In addition, under current legislation, institutional investors face something
of a "Catch-22" in terms of board representation. Getting truly independent
institutions to sit on corporate boards and to wield real influence will not
become a reality unless those institutions hold significant equity in the firms.
But large equity holdings put the independent institutions in the dangerous
position of being defined as controlling shareholders and hence as insiders on
the board. Such branding puts restrictions on the trading behaviour and
communicating abilities of institutions and detracts from the very purpose
such holdings were intended to achieve. Legislative changes aimed at raising
the ceiling used to define inside or controlling shareholders is a possible solution
to the above problem. The definition of outside directors may, alternatively, be
modified to accommodate the same result. Clearly, a 20 percent equity stake
held by the family or heirs of the founding member is very different from a
similar stake held by a pension fund. The pension fund should therefore be
treated as an outsider subject to a stated purpose of investment.
The rest of this commentary deals with four studies that were part of the
discussion group on The Board and Beyond. I have attempted to put the
evidence presented in these studies in a policy perspective. Where appropriate, I
have also taken the liberty of pointing out shortcomings and offering sugges-
tions for strengthening the evidence. The studies discussed in this session were
"Alternative Mechanisms for Corporate Governance and Board Composition"
by Jean-Marie Gagnon and Josee St-Pierre; "Executive Compensation and
Firm Value", by Ramy Elitzur and Paul Halpern; "In High Gear: A Case Study
of the Hees-Edper Corporate Group", by Ron Daniels, Randall Morck and
David Stangeland; and "Bell Canada Enterprises: Wealth Creation or
Destruction", by Paul Halpern and Vijay Jog.
Halpern and Jog provide a case history of Bell Canada Enterprises, high-
lighting a failed diversification strategy, its ultimate abandonment, and the
subsequent re-focusing. The market evidence appears to be consistent with
value creation from increased corporate focus. A diversified company can
trade at a discount to the sum of its divisional values due to cross-divisional
subsidies, sheer valuation complexities arising from difficulties in separating
aggregate cash flows, lack of proper incentives for divisional managers, dimin-
ishing returns to scope in management (this goes back to early Coasian views
on the boundaries of a firm), or sub-optimal corporate capital structure for the
several divisions of the firm. BCE was no exception to this holding company
discount. Halpern and Jog show that the failed diversification strategy at BCE
was accompanied by a reduction in the proportion of outside directors on the
board since 1984, a period during which the implied discount on BCE assets
increased. Such evidence is consistent with management entrenchment, aided
further by foreign-equity-holding restrictions that all but force Canadian
institutions to hold large capitalization stocks such as BCE in their portfolios.
Two alternative policy recommendations emerge, given the peculiarities
of the BCE case. First, relaxing foreign-equity-holding restrictions would subject
291
HOLDERNESS/MEHROTRA/G1LL
BCE management to stricter capital market discipline by raising its cost of
funding non-value maximizing assets. Second, regulatory pressure could be
applied to make BCE shed Bell Canada, which might ultimately organize itself
into smaller regional utilities a la the AT&T breakup in the United States in
1984- Shedding the crown jewel — Bell Canada — would, more importantly,
force managers to rethink their diversification strategy on a stand-alone basis,
free from complicated cross-subsidies that appear to be reflected in much of
the discount in BCE shares today. Recent work by Daley, Mehrotra &
Sivakumar (1995) at the University of Alberta shows that gains to spinning -
off assets are primarily concentrated in firms that shed unrelated assets. Prima
facie, BCE's spin-off of Bell Canada would likely create value by creating an
independently viable spun-off unit. This would also force the remaining parent
assets to strive for efficiency gains.
The role of boards is particularly important in large conglomerate-style
companies where shareholders' interests and management actions often collide
head on. To wit, managers desire diversification across industries for risk-
reduction reasons since their human capital is inherently undiversifiable (forcing
them to bear firm-specific risk). At the risk of over-simplification, to the
extent there are diminishing returns to scope in management and/or complex
cross-subsidies in the organization of conglomerates, shareholders prefer strictly
"pure plays", assuming responsibility for diversification on their own by
putting together portfolios of shares in firms across many industries. The
board's task here is to harmonize the two objectives and to select an optimal
level of diversification.
Daniels, Morck and Stangeland discuss the rationale of the conglom-
erate organizational form in their case study of the Hees-Edper group and show
that the performance of this group is indistinguishable from other Canadian
companies of a similar size, even though the Hees-Edper group carries a higher
debt leverage and asset risk. They conclude that the alleged benefits of a
conglomerate organizational form were absent at Hees-Edper. An excellent
discussion on the merits and demerits of the conglomerate form provides the
introduction to this study, and the subsequent empirical evidence sets a back-
drop for several policy implications. The authors argue that more stringent
disclosure requirements for conglomerates may reduce the monitoring cost of
outside shareholders. Furthermore, they suggest that proxy solicitation rules
should be modified so that dissident shareholders do not have to bear a dispro-
portionate cost of proxy campaigns. In the end, they point out that capital-
and product-market discipline is the best device to improve corporate governance
in Canada. To this end, NAFTA and the increasing inter-provincial freeing of
trade is a step in the right direction and should be encouraged.
Elitzur and Halpern examine the evidence of the relationship between
salary, bonus, and total compensation to Canadian executives and various
measures of corporate performance since the Ontario Securities Act was amended
to require executive compensation disclosure for firms registered in Ontario.
292
DISCUSSANTS' COMMENTS ON PART 111
As with the broader U.S. evidence, they find firm size to be positively relate d
to all measures of compensation.
The more interesting findings have to do with changes in individual
components of executive compensation and firm performance. The authors
find no significant relation between compensation and market returns. Thus,
from the shareholders' perspective, the evidence presented offers little hope
that bonus-based incentive schemes create value for shareholders too. The
preceding concern is somewhat mitigated by the fact that over suitably long
intervals, stock returns and accounting performance measures do tend to
converge. In the debate on corporate governance, it is important not to be
distracted by the publicity surrounding high levels of managerial compen-
sation that has been featured in the media recently and to focus instead on the
marginal product of top managers. Thus, any effort to regulate managerial
compensation is likely to miss the point with respect to effective corporate
governance and may well lead to the best managers leaving for less regulated
regimes (such as the United States).
Another interesting finding is that the existence of shareholders'-rights
plans (poison pills) is positively related to changes in bonus and overall
compensation, which raises the question of whether such bonuses are awarded
by self-serving compensation committees. The recommendation of both the
Cadbury report and the Dey report — tha t compensation committees be
composed mainly of outside directors — i s particularly germane here.
Another conflict of interest arises when executives from companies that
have a consulting relationship with the firm are placed on the board. Such
directors should excuse themselves from auditing and compensation
committees.
Gagnon and St-Pierre provide evidence on the determinants of outside
versus inside directors on Canadian boards under various ownership structures.
They view the board as one form of corporate governance and, as such, partially
substitutable with other forms (of governance) such as executive compen-
sation, debt policy, dividend policy, and anti-takeover devices adopted with or
without shareholder approval. They point out that the actual frequency of use
of these various forms depends on their effectiveness as well as on their cost of
adoption. They assert that board changes are a low cost mechanism for
improving corporate governance effectiveness. Such an assertion depends
crucially on the ownership regime of the firm. Under diffuse ownership, board
changes are often costly affairs — the out-of-pocket expense of a proxy battle is
itself a substantial sum and may deter outsiders with insignificant ownership
from attempting it. The free-rider problem associated with bringing about such
changes is particularly acute under diffuse ownership regimes. Furthermore,
political considerations may preclude institutional owners from overtly
confronting management in some instances. In this regard, legislation to safe-
guard the confidentiality of proxy ballot-counting may be needed to elicit
timely institutional responses free from political considerations.
293
HOLDERNESS/MEHROTRA/G1LL
An important contribution of the Gagnon and St-Pierre study is the
view that different governance structures are substitutable at the margin, and
thus in equilibrium we would expect to see varying degrees of usage for each of
these mechanisms, depending on the circumstances and characteristics of the
firm. For instance, they discuss the role of debt in corporate governance and
point out that higher debt ratios relieve the need to supply additional monitoring.
I believe that using debt as a tool of corporate governance requires a richer theory
of debt versus investment opportunity set variables facing the firm than is
currently available in the literature. What is needed is panel data relating debt
use with corporate performance, adjusting for variables in the investment
opportunity set facing the firm. Empirically, the evidence in Gagnon and St-
Pierre is not surprising, given that they find a positive correlation between the
voting rights of outsiders and their board presence, and a negative correlation
between the voting rights of insiders and the presence of outsiders on the
board. This is expected. The interesting result is with the return on assets
variable, which is positively related to outsider presence on the board for
diffusely held firms. Interpreting this result involves solving the age-old chicken
and egg problem - and Gagnon and St-Pierre's methodology is not appropriate
for such a task. We need panel data, with lead and lag variables, to understand
such a relationship.
CONCLUSION
I
N ORGANIZATIONAL FORMS CHARACTERIZED BY separation of ownership and
control, the board of directors plays a central role in effective corporate
governance. This commentary is based on the studies presented in the session
on The Board and Beyond at the conference on corporate governance
organized by Industry Canada. The key policy recommendations contained
here are summarized below.
First, the establishment of a national securities law should provide
consistency and uniformity of legislation across different jurisdictions and should
be welcomed.
Second, any changes that are sought in an effort to improve Canadian
corporate governance must address the peculiarities of Canadian corporate
ownership. The presence of controlling shareholders is a mixed blessing in
Canada. While controlling shareholders provide valuable monitoring of man-
agement, they also contribute to a conflict between controlling and minority
shareholders. Laws that limit the ability of controlling shareholders to garner
corporate resources for private use have the potential to reduce ex ante contracting
costs between outside equity providers and controlling shareholders, thereby
reducing the cost of equity capital and leading to a higher level of social
investment. At the same time, regulators must take into account the above
trade-off.
294
DISCUSSANTS' COMMENTS ON PART III
Third, the role of institutions requires special scrutiny. We are probably
entering an era of increased institutional activism and, in general, regulators
should encourage this trend. To this end, rules governing proxy ballots must be
re-examined. In particular, the confidentiality of proxy ballots must be provided
for, and the playing field with respect to proxy campaigns between manage-
ment and outside shareholders has to be leveled.
Fourth, regulators must redefine "insiders" in the context of institution-
al owners, especially those who are clearly non-management oriented.
Current rules discourage activist institutions from accumulating large blocks
of a firm's shares.
Finally, while regulators should in general welcome the trend toward
increased disclosure of corporate policies, such as those dealing with manage-
ment compensation, such disclosure must not be used as a pretext for interfering
with market-determined compensation packages.
ENDNOTE
1 Seagram's latest foray into Hollywood via its purchase of MCA is a case in
point. It is not clear if the interests of Seagram's minority shareholders
were best served in this acquisition. Whether the acquisition creates value
for Seagram's shareholders in the long run remains to be seen; in the short
span of the announcement week, however, financial markets lopped off
approximately $2 billion from Seagram's equity value.
295
HOLDERNESS/MEHROTRA/GILL
Lee Gill
Corporate Law Policy Directorate
Industry Canada
Commentary on Part III
Alternative Mechanisms for Corporate Governance
and Board Composition
T
HE STUDY BY JEAN-MARI E GAGNON AND JOSEE ST-PlERRE on alternative
mechanisms for corporate governance and board composition presents a
good summary of the mechanisms of corporate governance and an econometric
analysis of the determinants of one of those mechanisms — the composition of
the board of directors. The study also provides some useful insights into the
possible ranking of these corporate governance mechanisms according to their
cost, effectiveness (for both shareholders and managers), and probability of use.
The authors point out clearly that these rankings were made on a purely
intuitive basis. However, they do not provide a full explanation of the intuition
behind these rankings. The explanations underlying the "cost of mechanisms"
are comprehensible, but more explanation and discussion of the rankings is
required especially with respect to effectiveness. For example, they offer little
to support the statement that executive compensation warrants being ranked
as the third most effective mechanism for corporate governance behind board
composition and charter defences, and ahead of debt policy.
Although the study goes on to provide some fairly convincing evidence
as to the usefulness of debt policy, it is not clear why this policy is not as good
as executive compensation. Is this result particular to a widely held corporation,
or does it apply to all types of corporations? The study also points, in several
places, to the value of poison pills to managers, and argues that they are not
particularly valuable for shareholders. However, the authors offer little
evidence to support their claim, nor do they explain why so many shareholders
are inclined to vote for shareholder-rights plans if they are so bad for them.
In the course of the Phase II consultations to the proposed changes to
the Canada Business Corporations Act (CBCA) some corporate executives told
us that one of the main reasons poison pills have been implemented is that
statutes such as the CBCA do not allow sufficient time for corporations to
appraise proposals properly, solicit views from shareholders, and respond to
bidders. We were informed by companies that have implemented poison pills
that they did so primarily to ensure that sufficient time was available to
respond adequately to a takeover bid. They also informed us that if we
increased our allowed response period to between 45 and 60 days, there would
be significantly less need for poison pills and fewer shareholders would continue
296
DISCUSSANTS' COMMENTS ON PART III
to approve their use. If that is true, it suggests that these pills are now being
implemented to the benefit of shareholders. Is that true?
More explanation could also be provided in the statistical part of the
study. There is little explanation, for example, of Table 7, which sets out the
results of one of the key groups of equations found in the study.
The variables used to explain the outside/inside director ratio in firms
with diffuse ownership and concentrated inside ownership are significant.
However, in the case of diffuse ownership the only significant variables are
regulated enterprises, blockholdings which do not have members represented
directly on the boards, return on assets, and voting rights of inside directors.
The significance of these variables requires discussion.
It is not surprising to find that there are more outside directors in
regulated enterprises. It is also not surprising that the ratio of outside to inside
directors falls when the voting rights of inside directors are extended. It is
interesting, however, that the ratio of outside to inside directors rises as a
result of increased holdings by institutional and other blockholding investors
that have direct representatives on the board. It would, perhaps, be even more
interesting to know whether that result has evolved over time because the
number of institutional investors and other blockholding investors has grown,
or because the type of firm that is prone to hiring (more) outside directors
attracts these types of investors.
Although it is not mentioned by the authors, Table 7 suggests that the out-
side/inside ratio is also positively related to the return on assets. What does this
suggest - that the more outsiders there are, relatively speaking, on the boards of
diffusely owned firms, the greater the return of the corporation? Or, might this
also be related to the possibility that institutional investors are involved in
surveying management and in ensuring an acceptable return on assets?
One thing the authors do not discuss in any detail is the non-significant
results. For example, the authors find that the complexity of the organization
(as represented by the knowledge and log-assets variables) has not induced a
higher level of monitoring by way of outside directors. Why is this the case?
Under Concentrated Inside Ownership the authors indicate that the
existence of blockholders continues to be a significant determinant of the
outside/inside director ratio on the board. Inside directors, although important
shareholders, accept (or may have to accept) monitoring by other large share-
holders. The authors detect a significant size effect and indicate that this result
is consistent with the hypothesis of a nonlinear relationship between corporate
size and the ratio of outside/inside directors. It is not clear, however, why this
result is not found in the diffuse ownership case. Was this result specific to
certain industries which are not large factors in the diffuse ownership case?
The authors have certainly put together a good summary and analysis of
the mechanisms of corporate governance. More work, however, has to be done
to elaborate on the details of these preliminary results and to quantify more
clearly the importance of corporate governance mechanisms.
297
HOLDERNESS/MEHROTRA/GILL
Commentary on Part III
Executive Compensation and Firm Value
T
HIS STUDY BY RAMY ElITZUR AND PAUL HALPERN is a very good examination
of the observed importance of firm-specific performance and size variables
in the determination of executive compensation in Canada. I hope there will be
more follow-up work undertaken in support of that which has already been done.
The authors indicate they found that the bonus received by an executive
is positively related to net income, a form of short-term performance. They
also found that the bonus received had a positive relationship with the return
on assets. However, there was no relationship to other performance variables, such
as return on equity or the ratio of cash flow from operations to assets.
In all cases, however, the adjusted R
2
 seems very low. Actually, I had
thought that none of these equations would be significant as the adjusted R
2
were all below one and in some cases they were negative. At least one case,
however, appears to have a significant explanatory value. I carried out a rough
calculation on the log bonus equation with respect to sales and other variables
found in Table 5. By my calculation, the F-test indicated that the equation
had a significant explanatory value at the 5 percent level, a surprising result
that might deserve some explanation. On the whole, it would have been useful
if the F statistics had been presented and discussed in the study.
Still, as the authors clearly indicate, in most cases the explanatory
variables are not significant. Only in a few cases do the equations used satisfac-
torily describe the levels of executive compensation found in the sample.
The authors suggest some of the possible reasons for this lack of signifi-
cance of the explanatory variables. First, they point out that they use cash
compensation only; they do not use option-related compensation. The authors
also acknowledge that all forms of compensation should be used. They note
that an observation that cash performance is unrelated to firm performance
does not lead to the conclusion that total compensation, including equity
compensation, is unrelated to firm performance. That is indeed true and I feel
that it is one of the major problems of the study. Having said this, the authors
clearly note their continued work on collecting information on deferred
compensation and that they intend to include this variable in future reports.
Starting with this year much of this information should be available from
Form 40 filed with the Ontario Securities Commission (OSC).
With respect to salary, as the authors point out, the equations had a
better fit. Salary was found to be positively related to size: measured either as
assets or sales, type of industry the firm is in, and whether the firm is interlisted
on an American stock exchange. Of particular interest is the fact that salary
was found to relate negatively to a five-year rate of return variable. This
should be explained further in the study.
298
DISCUSSANTS' COMMENTS ON PART III
Another piece of information now available from the OSC is the break-
down of compensation among the top five executives in a firm. Some of the
results might be more robust if the authors focused on the CEO or on individual
rather than aggregate statistics.
As a final point, Elitzur and Halpern mentioned that one assumption in
the model is that managers are risk-neutral. Substantiation for this assumption
would be useful. Many feel that managers are, on average, risk-averse and that
some of the agency problems would make them more risk-averse than the
shareholders might like. How would the model change if risk aversion were
assumed?
299
This page intentionally left blank 
H
pART iv   iNSTITUTIONAAL iNVESTORS
This page intentionally left blank 
Jeffrey G. Macintosh & Lawrence P. Schwartz O
Faculty of Law Consulting Economist O
University of Toronto Toronto
Do Institutional and Controlling Shareholders
Increase Corporate Value?
SHAREHOLDERS' INCENTIVES AND FIRM VALUE
D
OES THE CONCENTRATION OF SHAREHOLDINGS make any difference to
corporate value or behaviour? Ever since Berle & Means,
1
 this has been
perhaps the most important question confronting those with an interest in
corporate governance. In a capitalist economy the corporation is the primary
instrument for the generation of wealth. If, as Berle & Means suggested,
corporations are run by managers in the interests of managers first, and share-
holders second, then we might be inclined to regulate corporate and managerial
behaviour rather more closely than if managers were entirely faithful to the
goal of maximizing shareholder wealth. If the combination of market oversight
(i.e., product, capital, corporate control, and managerial labour markets) and
direct shareholder monitoring of management were sufficient to align manager
and shareholder interests, the role of the state in policing corporate governance
would likely be much reduced.
Since Jensen &. Meckling (1976),
2
 most economists have agreed that
there is indeed a relationship between the concentration of share ownership
and corporate performance.
3
 Jensen &. Meckling suggested that there is a linear
and monotonically increasing relationship between managerial ownership and
corporate performance. The reason is that as managers acquire an increasing
share of the equity, their share of the cost of perquisite consumption and bad
management increases correspondingly. Jensen & Meckling therefore predicted
that managerial consumption of perquisites and slack should diminish as
managerial ownership increases.
While Jensen &. Meckling started a revolution in modern thinking about
concentration of share ownership, their view of the relationship between
ownership concentration and corporate performance is insufficiently nuanced
in a number of important respects. First, the evidence suggests that hostile
takeovers serve an important disciplinary function, by removing inefficient (or
high perquisite-consuming) managers.
4
 However, increasing managerial owner-
ship tends to make it easier for managers to fend off a hostile takeover bid.
5
303
MACINTOSH & SCHWARTZ
Thus, the increased degree of managerial entrenchment that accompanies
increasing managerial share ownership may actually reduce corporate value in
the mid-range between the ownership level at which managers have no power
to defeat a hostile takeover, and the level at which they can absolutely defeat
a hostile takeover.
6
There is empirical evidence to support the proposition that there is a
non-linear relationship between managerial ownership and corporate
performance.
7
 Morck, Shleifer & Vishny
8
 found, for a sample of very large
(Fortune 500) firms, that as managerial ownership increases from zero to 5 per-
cent, the value of the firm also rises. However, between 5 percent and 25 per-
cent ownership, firm value falls. After 25 percent, it rises once again.
However, McConnell & Servaes found that there was an increase in value up
to between 40 percent and 50 percent ownership, and a decrease thereafter.
9
Interestingly, the size of the average firm in the McConnell & Servaes sample
is significantly smaller than that used in the Morck et al sample.
Wruck found a relationship similar to that identified by Morck et al., but
in relation to non-managerial blockholdings in a sample of NYSE and AMEX
firms.
10
 This suggests that some caution should be exercised in interpreting the
Morck et al. results as the product of managerial entrenchment, given that the
blockholdings in Wruck's sample were non-managerial holdings. Unless the
blockholders in Wruck's sample were associated (or were otherwise acting in
concert) with the managers, it is not clear why increases in blockholdings would
serve to further entrench managers. One would expect the opposite to be true.
11
Indeed, Morck et al. find that the non-linear relationship between
management ownership and firm value applies (with the same turning points)
to outside directors, with nearly the same empirical robustness as it does to the
two top officers. Given that outside directors have less capacity than inside
directors to consume the perquisites of control, this too suggests that some
care should be taken in interpreting the Morck et al. results.
12
Moreover, if managerial entrenchment is the explanation of the Morck
et al. results, the advent of wealth-reducing poison pills and (in the United
States) state anti-takeover legislation may have substantially altered the
relationship between managerial share ownership and firm value.
13
 Both poison
pills and state anti-takeover statutes give even those managers who hold very
little (or none) of the firm's equity the power to defend against a hostile
takeover. In the United States there is no requirement that shareholders must
approve the adoption of a poison pill. Since both pills and anti-takeover
statutes can be used as powerful entrenchment tools by managers,
14
 it would be
interesting to see if more recent data sets, spanning the era of poison pills and
state anti-takeover legislation (i.e., encompassing the mid to late 1980s)
would yield similar results.
In Canada, there are five potentially significant differences that distinguish
Canadian practices from those in the United States. First, poison pills must be
approved by shareholders.
15
 Managers are therefore less likely to entrench
304
INSTITUTIONAL AND CONTROLLING SHAREHOLDERS
themselves by getting such measures passed. Second, there is no legislation
in Canada comparable to state anti-takeover statutes. Third, Canadian courts
and regulators have been tough on poison pills.
16
 It is now generally thought
that a poison pill may buy target management more time to find an alternate
transaction, but it will not allow management to entrench itself. Thus, if there
is the same non-linear relationship in Canada between managerial share
ownership and firm value as that observed in the United States, the poison
pill should have made fewer inroads into this relationship.
The fourth significant difference is perhaps the most important. While
most public corporations in the United States do not have a controlling share-
holder, most public corporations in Canada do. This is significant, because the
Morck et al. study is based on U.S. data, and so was conducted in an environment
in which there were few controlling shareholders to play a counterweight role
to the managerial pursuit of self-interest. As discussed below, increasing
managerial shareholdings is likely to have very different effects in corporations
both with and without controlling shareholders.
Fifth, the largest firms in Canada tend to be much smaller than those in
the Morck et al. sample.
17
 This smaller size is undoubtedly an important reason
why many of the largest Canadian corporations have controlling shareholders
and the largest U.S. corporations do not. Because of wealth constraints in
both countries, it is difficult for individuals or families to have large stakes in
the largest corporations. The largest firms in Canada may correspond more
closely with the sample used by McConnell &. Servaes, and therefore we
might have a greater expectation that the curvilinear relationship between
ownership and control found in that study will apply in Canada.
In summary, there is good reason to be cautious in assuming that the
Morck et al. results hold in Canada — less because of the effect of poison pills
(for all the reasons noted above) than because controlling shareholders tend
to prevent management entrenchment. Indeed, there is also good reason to be
cautious in assuming that the findings of Morck et al. still hold in the United
States, in this era of poison pills and state anti-takeover legislation.
Another important limitation on the Jensen &. Meckling view of the
relationship between managerial share ownership and firm value is that it fails
to take into account the distribution of non-managerial shareholdings.
Managers are much more likely to remain in control where there is no other
significant blockholder to oppose them, given that a significant blockholder
can supply a counterweight to management's power. An aggregate managerial
holding of 20 percent, for example, means one thing where no other share-
holder holds more than 1 percent. It means something else entirely where
there is a single (non-managerial) 25 percent blockholder, or where there is a
controlling (non-managerial) shareholder with a block of 51 percent.
Moreover, non-managerial shareholders (NMS) come in many varieties,
and the incentives confronting different types of NMS, and hence their desire
and ability to supply a counterweight to management, vary considerably. At
305
MACINTOSH & SCHWARTZ
the very least, there are four different types of non-managerial shareholder:
retail shareholders; institutional shareholders; controlling (but non-management)
shareholders; and individual, corporate, and other large, but non-controlling
(and non-institutional) blockholders.
18
 Retail shareholders are most subject to
the collective-action problems first noted by Berle & Means. Typically, they
hold small stakes, do not have the deep pockets needed to finance litigation
aimed at holding managers accountable, have high coordination costs, and
face daunting free-rider problems. Although the motivations of institutional
shareholders vary, institutions generally have much larger stakes. They also
tend to have the financial resources needed to sue management if necessary,
lower coordination costs, and less daunting free-rider problems. While many
factors tend to blunt institutional incentives to monitor,
19
 large blockholders
(particularly individuals, who are free from the internal agency problems that
burden most institutions) often have potent incentives to monitor and
discipline management. Controlling shareholders have not only the incentive
but also the ability to ensure that managers do their bidding.
The ability to "entrench" presumes an ability to control. Morck et d. state that:
... we have recognized that pure voting power is probably not the main mech-
anism by which managers retain control. It is more likely that ownership is
also positively correlated with status as a founder, tenure with the firm, pre-
ponderance of inside directors, ability to persuade shareholders, and other
conditions that facilitate management control.
2 0
The point can be broadened by noting that control not only derives from a
variety of sources, but also (importantly) is not an indivisibility. The power to
elect directors is an important manifestation of control, but even the power to
elect aR of the directors does not result in unencumbered power to direct financing
and investment decisions. For example, in Canada, fundamental changes in the
life of a corporation (such as amalgamation, sale of all or substantially all of the
assets, reincorporation in another jurisdiction, changes to the articles, etc.) must
typically be approved by a special resolution (requiring the approval of two-
thirds of those shareholders voting). This passes a power of negative control (the
power to obstruct a special resolution) to a blockholder or a coalition of block-
holders holding one-third of the voted shares. Class and majority of the minority
voting entitlements resulting from statutory, administrative, or judicial approval
requirements cede further veto powers to particular classes or even to minority
shareholders within a minority class. These veto powers substantially strengthen
the ability of NMS to prevent managers from adopting wealth-transferring or
wealth-reducing measures. This point is often overlooked by economists, who
tend to assume that once the 50.1 percent ownership threshold is reached,
further share acquisitions do not increase control.
21
While the view that control is not an indivisibility has force in both the
United States and Canada, it is of particular importance in the Canadian
306
INSTITUTIONAL AND CONTROLLING SHAREHOLDERS
setting. While in the United States many corporate law statutes (including
those of Delaware) require only simple majority approval for fundamental
changes, the standard in Canada is approval by two-thirds of all shareholders
voting. Moreover, there are more class and minority voting entitlements in
Canada than in the United States. For example, under most Canadian corporate
law statutes, restricted and non-voting shareholders must be given the right to
vote on an equal footing with voting shareholders in respect of a number of
different types of fundamental changes.
22
 Perhaps more importantly, the
corporate law requires that restricted and non-voting classes of shareholders
separately approve a significant number of different types of transactions by a
two-thirds majority.
23
 These mandatory class voting entitlements give
significant power to restricted and non-voting shareholders in the context of
fundamental corporate changes.
In addition to these corporate law requirements, there are a number of
statutory, stock exchange, and administrative requirements that a variety of
transactions be approved by a majority of disinterested shareholders. For
example, Policy 9.1 of the Ontario Securities Commission requires that all
"going private" transactions, and transactions with related parties
24
 be
approved by a majority of the minority (i.e., non-conflicted) shareholders.
Policy 9.1 further requires that the company secure a valuation of the subject
matter of the transaction (in the case of a related-party transaction) or of the
firm itself (in the case of a "going private" transaction), and that a summary be
sent to shareholders prior to the shareholder vote. In addition, approval by a
committee of independent directors is recommended, but not required
(although because of the policy, such approval is quickly becoming standard
practice). Policy 9.1 also has valuation requirements (and recommends
independent director approval) in the case of takeover bids effected by insiders
("insider bids") and the issuer itself ("issuer bids").
25
 All of these requirements
greatly add to the power of institutions holding restricted and non-voting
blocks of shares.
In Canada, minority interests also have the power to sue the corporation
and/or its directors, officers and controlling shareholders, on the grounds that
their rights or interests have been disregarded. The corporate oppression
remedy
26
 (for which, in the case of public corporations, there is no direct analogue
in the United States) greatly adds to the power of minority shareholders. In
addition, under the so-called "public interest" powers,
27
 Canadian securities
regulators have much broader powers than their American counterparts to
intervene in transactions that they find objectionable. On balance, therefore,
there appears to be more opportunity in Canada than there is in the United
States for minority interests to block redistributive transactions.
The distribution, size and identity of NMS will interact with the extent
of managerial shareholdings to influence the degree of corporate monitoring
and hence corporate value. Indeed, there will be many permutations and com-
binations, given that the interaction of various different types of NMS (and
307
MACINTOSH & SCHWARTZ
managerial shareholders) will affect the incentives of other NMS. For example,
the presence of a large non-managerial blockholder may have a negative effect
on the incentives of institutional shareholders to engage in monitoring. Each
institution might reason that the blockholder will be an effective monitor,
reducing the payoff from engaging in additional monitoring. Conversely, the
absence of institutional or block shareholders will accentuate the incentives of
retail investors to monitor management. Incentives to monitor are not
independent of the size or distribution of other shareholdings, nor of the identity
of the other shareholders.
It is also important to note that entrenchment results from a short-
circuiting of the market for corporate control by management. There are,
however, other market controls that tend to limit the ability of managers to
engage in perquisite consumption or other forms of inefficient behaviour, even
where they have effectively entrenched themselves. Managerial labour markets,
capital markets, and product markets all supply some measure of discipline.
28
Indeed, if these markets operate with a high degree of rigour, the absence of an
effective control market would, in theory, have an inconsequential effect on
managerial incentives.
Take, for example, the case of a corporation with two classes of shares -
one fully voting, the other non-voting. The use of a dual class structure may
serve to entrench the managers, but it comes at a cost. As Morck argues,
29
 and
as supported by Morck & Stangeland,
30
 the use of a dual class structure is likely
to increase the corporation's cost of capital (in both current and subsequent
rounds of financing). At least part of this cost is likely to be passed on to the
managers, limiting to some extent their incentive to adopt such structures.
However, as argued by Daniels & Halpern,
31
 the efficacy of at least one
market — the managerial labour market — may also be partly short-circuited by
management entrenchment.
Do CONTROLLING SHAREHOLDERS INCREASE OR
DECREASE FIRM VALUE?
A
S THE ABOVE DISCUSSION MAKES CLEAR, both theory and evidence suggest
that it would be naive simply to assume that the presence of a controlling
shareholder will have an unambiguous effect on firm profitability. Controlling
shareholders can supply effective monitoring of managers; but they may also
engage in value-reducing redistributive transactions. For example, a controlling
shareholder might set up a transfer pricing arrangement on non-market terms
with another controlled company for the purpose and effect of redistributing
shareholder wealth to that other company.
Holderness & Sheehan suggest that a controlling shareholder might
appoint herself manager, then pay herself an excessive salary, hire relatives at
inflated salaries, negotiate favourable deals with other controlled companies,
308
INSTITUTIONAL AND CONTROLLING SHAREHOLDERS
loot the corporation, or invest in negative net present value projects.
32
 While
this is not necessarily an exhaustive list of the ways in which controlling
shareholders can cheat minority shareholders, it suggests that there are several
ways to achieve this result.
Examining a sample of large public firms trading on the NYSE and
AMEX, however, Holderness & Sheehan found little evidence that majority
shareholders (holding 50 percent or more of the equity) exploit minority
shareholders. When majority blocks were traded, the value of minority shares
increased in price appreciably, and such trades were accompanied by significant
changes in the composition of management and the board.
33
 Majority share-
holders who were CEOs earned only marginally more compensation on average
than other CEOs. There were only a small number of cases involving the
employment of relatives, and no suits at all alleging abuse of majority power.
Finally, Holderness & Sheehan found that investment policies, the frequency
of corporate control transactions, accounting rates of return, and Tobin's Q were
not significantly different for either widely held or majority controlled firms.
These findings are of particular interest in the Canadian setting, (subject
to differences in the legal regime as already described) in which the danger of
the types of redistributive behaviour referred to above is heightened. The key
factors that heighten the danger of predation are: the high concentration of
share ownership within individual corporations, and the fact that common
shareholders control many corporations - in extended corporate empires like
that controlled by the Bronfman interests. In such settings, there are many
more possibilities for redistributive behaviour.
It is our hypothesis that Canadian controlling shareholders engage in
both more effective monitoring of managers than non-control shareholders,
and redistributive transactions that shift wealth from non-controllers to
controllers. There are four measures of profitability used to test this hypothesis:
return on assets, return on sales, sales growth, and the ratio of share price to
book value per share. Since the redistribution of wealth will have a minimal
impact on return on assets, return on equity, and sales growth,
34
 we hypothesize
that the presence of a controlling shareholder (and the managerial monitoring
that this brings) will have a positive effect on these three measures of profitability.
The effect on the price-to-book ratio is likely to be more ambiguous. It is
probable that the price-to-book ratio will be positively affected by the more
proficient monitoring of managers brought to bear by the presence of a
controlling shareholder. However, it may also be negatively affected by the
redistribution of wealth from minority interests to the controlling shareholder.
We have no a priori way of knowing which effect will dominate. Thus, we simply
hypothesize that neither effect will dominate, and that the price-to-book ratio
will be unaffected by the presence of a controlling shareholder.
Note that we do not distinguish between controlling shareholders with
different attributes. For the purposes of this study, we define a controlling
shareholder as any non-institutional shareholder holding in excess of 10 percent
309
MACINTOSH & SCHWARTZ
of the total voting power in the corporation. We run a second set of regressions
with the control threshold set at 20 percent. The test of whether the presence
of a controlling shareholder affects firm value is therefore a somewhat crude
one. Specifically, there is evidence that not all types of controlling shareholders
have the same effect on firm value. For example, Morck & Stangeland present
Canadian evidence suggesting that firms controlled by descendants of founding
stockholders diminish firm value.
35
 Morck & Stangeland also find that there
are differences in corporate performance based on whether the controller is a
foreign company, a Canadian-owned subsidiary, or an independent private-
sector firm.
36
 Looking at U.S. data, Holderness & Sheehan find evidence that
individual controllers tend more toward entrenchment (and inefficient
behaviour) than corporate controllers.
37
Moreover, although controlling shareholders will ultimately appoint
management
38
 (indeed, that is what it means to be a controlling shareholder),
the controlling shareholder (CS) may or may not be an officer or director of
the corporation. A CS who is an officer or director (or both) may derive either
pecuniary or non-pecuniary benefits from occupying the position of
director/officer, and may wish to retain control for the purpose of maintaining
the flow of these benefits, which are not shared by other shareholders.
In contrast, a CS who is not a manager will not participate in any special
benefits flowing to management. The non-manager CS may still derive benefits
from control that do not flow to other shareholders (such as the psychic benefit
that goes with "pulling the strings"), or that are earned at the expense of the
other shareholders (e.g., the opportunity to indulge personal and non-optimal
investment preferences). In such cases, however, the element of side benefits
consumed as manager has been removed. Therefore, a non-manager CS can be
expected to monitor managers more effectively than a manager CS.
It should be noted, however, that there is evidence that almost all majority
(as distinct from merely controlling) shareholders are represented on the
board of directors or in management.
39
 Consequently, failing to distinguish
between management and non-management controlling shareholders may not
constitute a serious empirical failing.
Do INSTITUTIONAL SHAREHOLDERS INCREASE OR
DECREASE CORPORATE VALUE?
W
E HYPOTHESIZE THAT FIRM VALUE IS POSITIVELY CORRELATED with the
degree of institutional ownership, for the following reasons.
Many have argued that institutional shareholders are better able to
resolve problems of collective action than retail shareholders.
40
 Institutional
shareholders typically hold much larger stakes than retail shareholders. Thus,
they have more at stake in corporate decisions. They also have lower coordination
costs, are better able to evaluate management's performance, and possess the
3/0
INSTITUTIONAL AND CONTROLLING SHAREHOLDERS
resources to oppose management (whether in court or otherwise) in the event
that management attempts to proceed with wealth-redistributing or wealth-
reducing measures.
There is U.S. evidence that institutional shareholders are more likely to
vote than retail investors, and are more likely to oppose as least some forms of
wealth-reducing measures.
41
 There is also evidence that the market reaction to
the adoption of shark-repellent measures is negative for firms with low insti-
tutional ownership, but zero for firms with high institutional ownership.
42
Indeed, the incidence of such amendments appears to have decreased over
time and also appears to be a result of the fact that institutional shareholders
not only play an active role in preventing management from introducing such
measures, but also tend to vote against such resolutions when they are presented
for approval.
43
 McConnell & Servaes also find evidence that increasing
institutional ownership correlates with increasing firm value and profitability.
44
In Canada, the growth of institutional holdings has tracked that in the
United States.
45
 As indicated by Macintosh,
46
 over the past two or three decades,
the percentage of equity held in the portfolios of life insurance companies and
public pension funds has greatly increased. However, the percentage of equity
held in mutual fund portfolios appears to have shrunk over the past 25 years.
47
Several commentators have argued that Canadian institutional investors
have become increasingly important in matters of corporate governance.
48
 As
summarized by Macintosh,
49
 the influence of institutional investors has been
felt in a number of ways, including (but not necessarily in order of importance):
• Voting against management (often in conjunction with other
institutional investors).
• Threatening to exercise dissent rights.
• Suing to enjoin a transaction.
• Enlisting the support of securities regulators to stop a transaction.
• Publicly expressing dissatisfaction with management, or a partic-
ular course of action recommended by management.
• Mounting or participating in a proxy battle to unseat manage-
ment (in rare cases).
• Supporting institutional organizations (such as the Pension
Investment Association of Canada) and soft dollar brokers (like
Fairvest Securities) which actively support institutional causes.
• Creating proxy voting guidelines, either individually or through
representative organizations, dealing with matters like poison
pills, executive compensation, blank cheque preferreds, etc.
• Meeting with management, either individually or collectively, to
discuss matters of concern.
3/1
MACINTOSH & SCHWARTZ
Anecdotal evidence suggests that the number of Canadian institutions
willing to engage in public activism vis-a-vis corporate managers is small.
Indeed, even some of the very largest institutional investors, such as the
Ontario Municipal Retirement System (OMERS) have preferred to avoid
activities that attract public attention.
50
 However, anecdotal evidence also
suggests that a good many institutions engage in "quiet diplomacy", chatting
with management behind closed doors about matters of concern.
The increasing willingness of Canadian institutions to engage in such
activities, and their commensurately decreasing propensity to sell their invest-
ments when dissatisfied with management,
51
 suggests that institutions do,
indeed, engage in useful monitoring that might be expected to enhance
corporate value.
On the other side of the ledger, there are many problems that tend to
blunt institutional incentives.
52
• Institutional investors are not exempt from free-rider problems
(increasing corporate value will benefit other, non-contributing
shareholders, and in particular may benefit the institution's rivals).
• Institutions may be co-opted by management (e.g., a bank might
vote with management in the fear that if it does not, the bank
will lose the corporation's deposit and/or loan business).
• Pension fund managers may follow a "golden rule" — a mutual
back-scratching arrangement under which fund managers
appointed by management from one corporation will refrain from
engaging in activism in return for similar behaviour from other
fund managers.
• Political pressures brought to bear on public pension funds.
• Limited monitoring capabilities, given large portfolios, limited staff,
and limited ability to engage in active management activities.
• The need or desire to maintain liquid portfolios, which results in
the acquisition of small blocks without significant voting power.
• Legal restraints on institutional monitoring activities.
53
• Agency conflicts within institutional investors
• An institutional culture of "passivity".
• Fear of political reprisals for too direct involvement in corporate
activities.
• Fear that approaching other shareholders with concerns about
management, will trigger a "race to the exit" which will cause
the share price to fall.
• Potential fiduciary conflicts between maximizing fund value and
corporate value when fund managers become corporate directors.
3/2
INSTITUTIONAL AND CONTROLLING SHAREHOLDERS
• The proliferation of non-voting shares in Canada.
• Difficulties in identifying other shareholders.
• Poison pills.
Some researchers have suggested that institutional shareholders actually
diminish corporate value. The co-option referred to above may not merely render
institutions ineffective monitors, but may convert them into unwilling allies of
management against other (non-co-opted) shareholders. There is evidence, for
example, that institutions tend to vote with management in proxy contests,
54
and that institutions that are more subject to co-option are more likely than
other institutions to vote with management.
55
It has also been argued by some commentators that because the
performance of investment managers is evaluated quarterly, such managers
tend to have extremely short investment horizons and therefore undervalue
long-term investment projects.
56
 According to this view, long-term projects
are systematically undervalued by the market, resulting in an inefficiently
low level of expenditures in long-term projects such as research and devel-
opment. However, there appears to be little evidence to support the view
that the market undervalues long-term activities. In fact, the evidence
supports the view that the market does not inappropriately discount long-
term investment projects.
57
On balance, both theory and U.S. data support the view that institutional
shareholders increase corporate value. We thus feel comfortable with the
hypothesis that the presence of institutional investors will have a positive effect
on all four of our accounting measures of profitability. The measure of institu-
tional ownership we use is described later in this study, in the Data section.
It is important to note, however, that institutional shareholders frequently
play a somewhat different monitoring role in Canada than in the United
States because the great majority of Canadian public corporations have a
controlling shareholder. As elaborated above, we hypothesize that this results
in better monitoring of management, thus diminishing the marginal value of
institutional monitoring of managers. However, we also hypothesize that the
presence of a controlling shareholder will result in a higher probability of
redistributive transactions. Canadian institutions are therefore more likely to
monitor controlling shareholders than management.
Initially, this might seem like an arid distinction, given that controlling
shareholders will act through their appointed managers. Indeed, there is signifi-
cant similarity in the characters of redistributive events for controlled and
non-controlled corporations. For example, in a case where the controlling
shareholder is also a manager, the firm may be just as prone (perhaps even
more prone)
58
 to overpay managers as in a case where the firm is widely held.
However, it is virtually certain that a non-trivial class of redistributive events
will occur more frequently in controlled corporations than in non-controlled
corporations. For example, controlling shareholders can be expected to use
313
MACINTOSH & SCHWARTZ
their powers of control to engineer non-arm's-length transactions more often
than the managers of a non-controlled corporation.
An interesting question is whether the marginal value of increased
institutional ownership is greater or less when there is a controlling shareholder.
When there is a controlling shareholder, the increasing presence of institu-
tional shareholders should mitigate the incentive and/or ability of controlling
shareholders to engage in redistributive transactions. It should not interfere
with the controlling shareholder's superior capability to monitor management,
but (assuming that controlling shareholders are good managerial monitors)
neither will increasing institutional presence add much to managerial monitoring.
When there is no controlling shareholder, increasing institutional presence
should have a greater (positive) marginal effect on managerial monitoring,
since dispersed shareholders are likely to supply much less effective monitoring
of management than a controlling shareholder. However, because there is
probably less danger of redistributive transactions, an increasing institutional
presence will result in a smaller (positive) marginal influence in discouraging
such transactions. It is therefore difficult to tell, a priori, whether an increasing
institutional presence will have a greater effect when there is a controlling
shareholder or when there is not.
To test which effect is dominant, we constructed an interaction term,
which is the percentage of institutional ownership multiplied by a dummy
variable, which is zero if there is no controlling shareholder, and 1 if there is a
controlling shareholder. Because we cannot determine on an a priori basis
which effect will dominate, we hypothesize that (randomly) the interaction
variable will be positively related to our four measures of profitability. Note,
however, that we have already suggested that the danger of redistribution by a
controlling shareholder is most likely to show up in the form of a lower price-
to-book ratio. If this is so, and if controlling shareholders are less likely to
engage in redistributive transactions as institutional ownership rises, the inter-
action term is most likely to be positive when the dependent variable is the
price-to-book ratio.
DATA
T
HE UNIVERSE OF FIRMS FROM WHICH the data set is drawn is the TSE 300
Composite Index, which consists of 292 firms, given that some firms have
more than one class of common equity included in the TSE 300. Inclusion of
an issuer in the data set required that data be available regarding institutional
holdings, control holdings, assets, return on equity (ROE), return on assets
(ROA), the ratio of price to book value, and sales growth.
The greatest difficulty we encountered was in assembling data relating to
institutional holdings. To date, no data set has been assembled dealing with
Canadian firms. Canadian data is simply not as readily available as U.S. data,
given that Canadian shareholders are subject to public reporting requirements
314
INSTITUTIONAL AND CONTROLLING SHAREHOLDERS
only at a 10 percent ownership level, compared to a 5 percent level in the
United States. Many institutions have holdings of less than 10 percent. This is
partly the result of a need to achieve prudent diversification, but it is also a
consequence of the desire to keep the institution's holdings confidential.
Some institutional investors are expressly forbidden to cross the 10 per-
cent ownership threshold. For example, banks are forbidden by the Bank Act
(subject to limited exceptions) from holding more than 10 percent of a single
issuer.
59
 Thus, few bank holdings will be publicly reported. Mutual funds are
similarly forbidden from holding more than 10 percent, although in this
case by a National Policy statement
60
 issued by the Canadian Securities
Administrators rather than by a legislative requirement. The mutual fund policy,
however, allows a mutual fund to hold a block that is larger than 10 percent
with the permission of the regulators.
An additional difficulty in determining institutional holdings is the
uneven reporting of 10 percent blocks. In many cases, one entity (a bank, for
example) will market a number of mutual funds under common management.
Many funds take the view that they need not aggregate such holdings for
reporting purposes. In other cases, external managers will manage a variety of
funds (such as pension and mutual funds) with diverse beneficial ownership.
Many of these fund managers contend that they need not aggregate these
diversely owned funds for reporting purposes.
Our data relating to institutional holdings is drawn from two sources.
The first is Vickers Institutional Holdings Guide, a commercial service assembled
by Vickers Stock Research Corporation (based in New York). While Vickers
digests holdings of less than 10 percent, the service depends on voluntary
disclosure by institutional investors and hence is unavoidably incomplete, due
to the fact that some funds choose not to disclose their holdings. Our second
source is a database graciously furnished by Fairvest Securities Corporation, a
Toronto-based institutional broker that has been a champion of institutional
causes (and whose activities are funded by soft dollar commissions). Fairvest
maintains a list of all 10 percent or greater holdings in all TSE 300 corpora-
tions, and in some cases large institutional holdings of less than 10 percent.
The information concerning these holdings is obtained from public filings,
information circulars, prospectuses and other sources. We combined the
Vickers and Fairvest databases to produce our own institutional ownership
database.
The accuracy of the database is diminished by a number of factors. First, as
noted above, we are certain that we do not have all holdings of less than 10 per-
cent. Second, the Fairvest database is continually updated as new information is
received, but no historical record is kept showing institutional ownership at
specific points in time. Given that the database was provided to us in early
November 1994, it speaks as of that date. The Vickers database, however, was
compiled as of March 31, 1994. Combining the two databases thus results in a
timing mis-match. While we would prefer that all our data speak as of the
3/5
MACINTOSH & SCHWARTZ
same date, since Fairvest does not keep historical records of share ownership,
this was simply not possible.
Third, there is an additional timing mis-match between the institutional
shareholder ownership data and the accounting information for our sample of
firms. Given that our accounting data relates to the 1993 fiscal year, we are
thus relying on data that describe institutional holdings subsequent to the
reporting period as a proxy for institutional holdings throughout the reporting
period. Ideally, when attempting to relate institutional ownership to flow
variables such as return on assets or sales growth, the statistic on institutional
ownership would be a weighted composite reflecting degrees of institutional
ownership throughout the reporting period, given that institutional ownership
can change over the course of a year. Once again, because of the unavailability
of data on institutional holdings during the reporting period, we had little
choice in this regard.
Subsequent to the conference for which this study was prepared (in
March 1995), an attempt was made to obtain accounting data for the 1994
fiscal year. This was not possible, because as of June 1995, the Financial Post
database from which our accounting data are drawn contained fiscal 1994
accounting data for only a fraction of the firms in our portfolio.
While we would have preferred that our database distinguish between
voting and restricted or non-voting shareholdings, the Vickers database does
not explicitly distinguish between voting and non-voting shares.
61
 The
Fairvest database, on the other hand, records the largest voting block of shares
(as a percentage of total voting power) and the identity of the shareholder. Most
such blocks are in excess of 10 percent. Our combined database therefore reflects
institutional ownership of voting, restricted-voting, and non-voting equities.
Because we include non-voting and restricted-voting shares as well as
voting shares, we implicitly make two important assumptions. First, institutions
possess a lever over management even when they hold restricted- or non-voting
shares.
62
 Second, institutions also possess a lever over management when they
hold a large share of a single class of equity, rather than a high percentage of
the total equity.
While we would rather have worked with a more finely graded sample,
we believe that both of our assumptions are defensible on theoretical grounds.
While restricted voting shares confer limited voting power, and non-voting
shares confer none at all, holdings of both classes of share may nonetheless
give institutions both the incentive to monitor management and a lever over
management. Arguably, holding a non-voting block gives institutional
investors a greater, rather than a lesser, incentive to monitor management and
controlling shareholders. Without a significant voice, institutional shareholders
will perceive exit to be a more important source of self-protection. The value
of the exit option will, in turn, depend on a reliable and accurate flow of
information about the company - giving institutions an incentive to monitor
on a regular basis.
3/6
INSTITUTIONAL AND CONTROLLING SHAREHOLDERS . . .
Moreover, while there is a tendency in the literature to draw a clear line
between exit and voice, the two are, in fact, related. The very fact that
institutions can stampede for the exit and cause the firm's share price to fall,
by itself provides them with a lever over management. In many cases, manage-
ment compensation depends partly on share price. Managers will thus want to
avoid any conduct that will cause the share price to fall. Indeed, many senior
officers have a significant portion of their wealth invested in the firm's stock -
an arrangement that further increases their incentive to avoid a substantial
drop in share price.
63
Further, predatory behaviour towards restricted or non-voting shareholders
that causes a drop in the corporation's stock price will adversely affect the
corporation's cost of capital in subsequent rounds of financing. This in turn
will raise the probability of bankruptcy (which will almost certainly result in
the managers losing their jobs), in addition to depreciating the value of the
managers' human capital in the external labour market. Also, institutions
holding restricted- or non-voting equity retain the power to embarrass manage-
ment by going public with their concerns about management's intended
course of action. Since many managers prefer to avoid negative publicity, the
power to go public allows institutions to exert influence over management
even when they do not hold voting stock. Thus, while we would have preferred
to segregate the institutional ownership sample by voting power, we suggest
that there is good theoretical justification for mixing voting, restricted-voting,
and non-voting holdings in our database. Where a corporation had more than
one common share class in the TSE 300 Composite Index, we determined
institutional ownership only for the traded class with the largest public float.
The institutional ownership variable (% INST) represents the percentage
of institutional ownership in the class selected for inclusion in the study,
rather than the percentage which institutional holdings constituted of all
common share classes. As indicated above, we assume that institutions possess
a lever over management by virtue of holding large positions in a single class
of common shares, rather than a large share of the total equity of the company.
We contend that this assumption is theoretically defensible for the same reasons
that inclusion of restricted- and non-voting shares is defensible. In addition,
class and "majority of the minority" voting requirements (alluded to earlier)
give those who hold large positions in a single class of shares an effective veto
over many types of corporate fundamental changes (even though the consti-
tution of the corporation may designate these shares as restricted-voting or
non-voting). Given that many types of redistributive activities must be
consummated as fundamental changes subject to these statutory, administrative,
and judicial voting entitlements, this gives institutional blockholders a
solid bargaining position vis-a-vis management.
Controlling shareholdings (CONTROL) were determined from the Fairvest
database. We adopted mechanical cutoffs to determine who is a controlling
shareholder, defined as any non-institutional shareholder (or where it
317
MACINTOSH & SCHWARTZ
appeared likely that shareholders were acting in concert, coalitions of share-
holders) holding either 10 percent or more (in the first set of regressions) or
20 percent or more (in the second set) of total voting power.
64
 These were
identified from the Fairvest data base. In our final sample of 99 companies,
there were 63 firms satisfying the first criterion, and 55 satisfying the second.
Accounting data for each issuer was compiled from the Financial Post
Surveys for the fiscal year ending closest to March 31, 1994. In most cases, the
fiscal year end was December 31, 1993. From this source, we obtained data on
total assets, the ratio of share price to book value per share, sales growth for
the 1993 reporting period, return on assets, and return on equity.
From the total of 292 TSE 300 firms, we collected complete data for 99
firms. These 99 firms are broadly representative of the industry structure of the
TSE 300 Index.
METHODOLOGY
W
E USE STANDARD OLS REGRESSION METHODOLOGY to determine if rising
levels of institutional ownership and/or the existence of a controlling
shareholder affect firm value. To this end, we define four independent variables.
The first, % INST, as described above, is the percentage of institutional owner-
ship of the firm's outstanding common shares that are listed in the TSE 300
index. The second, CONTROL, also described above, is a dummy variable that is
1 if there is a control shareholder, and 0 if there is no control shareholder. The
third, ASSETS, is designed to control for possible size effects.
65
 The fourth
independent variable, INTERACTION, is an interaction term equal to institutional
holdings multiplied by the CONTROL dummy. The theoretical basis for the
interaction term arises from the earlier discussion about the manner in which
interactions between different types of shareholders may affect corporate value.
Four dependent variables are regressed on the independent variables
described above. These are: the ratio of share price to book value per share;
return on assets; return on equity; and sales growth.
All of the accounting data were taken from the Financial Post Surveys, in
which the price-to-book ratio is defined as the ratio of average share price to
book value per share. In turn, the average price is equal to the sum of the high
and low stock prices for the year divided by two, and book value is defined as
the year-end book value per share (excluding preferred shares). Although the
latter is defined to include any intangibles reflected on the balance sheet, the
typical balance sheet does not place a value on intangibles. Thus, the price-to-
book ratio is largely based on the book value of tangible assets.
The price-to-book ratio is a proxy for Tobin's Q, which is the ratio of the
firm's market value of debt and equity to the replacement cost of its tangible
assets.
66
 As indicated by Ross & Westerfield, "... firms with high Q ratios tend
to be those firms with attractive investment opportunities or a significant
competitive advantage".
67
 Tobin's Q has been extensively used in similar types
3/8
INSTITUTIONAL AND CONTROLLING SHAREHOLDERS . .
of investigations in the United States. The other dependent variables are
commonly accepted indicia of firm performance.
Note, however, that measuring profitability is as much art as science. No
single measure can give an unambiguous signal as to firm performance. For
example, sales growth may be associated with future profitability, but may also
be a sign of managerial empire building undertaken at the expense of future
profits. This is why we use four different measures of profitability.
In the first set of regressions, the four measures of profitability are
regressed on ASSETS, % INST, and CONTROL, with the CONTROL threshold set at
10 percent. In a second set of regressions, the same four measures of profitability
are regressed on the same independent variables, plus the INTERACTION variable.
The third and fourth sets of regressions repeat the above with CONTROL set at
20 percent.
RESULTS
W
E DESCRIBE OUR REGRESSION RESULTS BELOW. Note that in each table we
first describe the results with the CONTROL threshold set at 10 percent.
These results are described in each table as regression "A". We then describe
the results when the CONTROL threshold is raised to 20 percent. This is
described in each table as regression "B". In each case, "Model 1" refers to the
regression without the interaction term as an independent variable, and
"Model 2" refers to the regression with the interaction term.
PmcE-TO'BooK RATIO
WHEN PRICE-TO-BOOK RATIO is REGRESSED against ASSETS, % INST, and CONTROL,
none of the independent variables is statistically significant (see Table 1(A),
Model 1). When the interaction term is added as an independent variable, the
results change somewhat (see Table 1(A), Model 2). In this case, the percentage
of institutional holdings is statistically significant at the 10 percent level,
although the sign of the coefficient is negative, counter to our hypothesis that
increasing institutional holdings increases firm value. The CONTROL variable is
statistically significant, but again the coefficient is negative. The INTERACTION
coefficient is positive and statistically significant.
When the control threshold is raised to 20 percent and the regressions
rerun without the INTERACTION term, the CONTROL variable is statistically
significant at the 5 percent level, again with a negative coefficient (see Table
1(B), Model 1). When the INTERACTION term is introduced, again only the
CONTROL variable is statistically significant, again with a negative coefficient
(see Table 1(B), Model 2). The INTERACTION term is positive, but not quite
significant at conventional levels. In none of the regressions is the ASSET
variable either statistically or economically significant.
319
MACINTOSH & SCHWARTZ
TABLE 1(A)
DETERMINANTS OF PRICE-BOOK RATIO - CONTROL 10%
INDEPENDENT
VARIABLE COEFFICIENT
STANDARD
ERROR T-STAT1STIC
MODEL 1
ASSETS
% INST
CONTROL (10%)
Constant
Standard Error of Y est
R2
No. of Observations
Degrees of Freedom
-0.0015980
-0.0035428
-0.2230006
2.32191285
1.32214303
0.03019484
99
95
0.00100208
0.00882972
0.27994731
-1.5947184
-0.4012464
-0.7965807
MODEL 2 (INCLUDES INTERACTION TERM)
ASSETS
% INST
CONTROL (10%)
INTERACTION
Constant
Standard Error of Y est
R2
No. of Observations
Degrees of Freedom
-0.0015103
-0.0265151
-0.8412449
0.037277
2.72774422
1.29929986
0.07327550
99
94
0.00098566
0.01400213
0.40392588
0.0178324
-1.5322851
-1.8936485***
-2.0826714**
2.0904018**
Notes: Statistically significant at 5% level, 2-tailed test.
Statistically significant at 10% level, 2-tailed test.
These regressions supply some (weak) evidence that increases in insti-
tutional holdings result in a lower price-to-book ratio. They supply more
robust evidence that the presence of a control shareholder results in a lower
price-to-book ratio. There is also some evidence that when there is a control-
ling shareholder, increased institutional holdings result in an increase in the
price-to-book ratio.
320
INSTITUTIONAL AND CONTROLLING SHAREHOLDERS
TABLE 1(B)
DETERMINANTS OF PRICE-BOOK RATIO - CONTROL 20%
INDEPENDENT
VARIABLE COEFFICIENT
STANDARD
ERROR T-STAT1ST1C
MODEL 1
ASSETS
% INST
CONTROL (20%)
Constant
Standard Error of Y est
R2
No. of Observations
Degrees of Freedom
-0.0016572
-0.0046251
-0.5322349
2.49518792
1.29926333
0.06346930
99
95
0.00098005
0.00867397
0.26504631
-1.6910123
-0.5332189
-2.0080828*
MODEL 2 (INCLUDES INTERACTION TERM)
ASSETS
% INST
CONTROL (20%)
INTERACTION
Constant
Standard Error of Y est
R2
No. of Observations
Degrees of Freedom
-0.0015594
-0.0201169
-1.0222615
0.0306753
2.16252486
1.28447317
0.09430507
99
94
0.00097043
0.01218708
0.37906480
0.017147
-1.6069738
-1.6506784
-2.6967989*
1.7889563*
Notes: Statistically significant at 1% level, 2-tailed test.
Statistically significant at 5% level, 2-tailed test.
Statistically significant at 10% level, 2-tailed test.
321
MACINTOSH & SCHWART Z
TABLE 2(A )
DETERMINANTS OF RETURN ON ASSETS - CONTROL 10%
INDEPENDENT
VARIABLE COEFFICIENT
STANDARD
ERROR T-STAT1ST1C
MODEL 1
ASSETS
% 1NST
CONTROL (10%)
Constant
Standard Error of Y est
R2
No. of Observations
Degrees of Freedom
-0.0007253
0.18195008
4.62878708
-1.7786128
7.96604675
0.15773592
99
95
0.00603769
0.05319997
1.68671135
-0.1201334
3.42011590'
2.74426781*
MODEL 2 (INCLUDES INTERACTION TERM)
ASSET
% INST
CONTROL (10%)
Interaction
Constant
Standard Error of Y est
R2
No. of Observations
Degrees of Freedom
-0.0010586
0.26924555
6.97814216
-0.141654
-3.3207895
7.93723692
0.17261906
99
94
0.00602131
0.08553704
2.46752544
0.108936
-0.1758223
3.14770709"
2.82799197"
-1.300344
Note Statistically significant at 1 % level, 2-tailed test.
RETURN ON ASSETS
THE RESULTS FOR RETURN ON ASSETS are quite different. Although once again
the ASSET control variable is not statistically or economically significant in
any of the regressions, the coefficients on % INST and CONTROL are both
positive and statistically significant both with (Table 2(A), Model 2) and
without (Table 2(A), Model 1) the INTERACTION variable present. The
INTERACTION term is negative but not statistically significant.
322
INSTITUTIONAL AND CONTROLLING SHAREHOLDERS .
TABLE 2(B)
DETERMINANTS OF RETURN ON ASSETS - CONTROL 20%
INDEPENDENT
VARIABLE COEFFICIENT
STANDARD
ERROR T-STATIST1C
MODEL 1
ASSETS
% INST
CONTROL (20%)
Constant
Standard Error of Y est
R2
No. of Observations
Degrees of Freedom
-0.0019615
0.17597162
2.87380689
-0.2858958
8.14858437
0.11869363
99
95
0.00614657
0.05440051
1.66228982
-0.3191224
3.23474212*
1.72882420***
MODEL 2 (INCLUDES INTERACTION TERM)
ASSETS
% INST
CONTROL (20%)
INTERACTION
Constant
Standard Error of Y est
R2
No. of Observations
Degrees of Freedom
-0.0025013
0.26147897
5.57851356
-0.169312
-1.7614647
8.08668768
0.14116810
99
94
0.00610959
0.07672653
2.38648709
0.107953
-0.4094200
3.40793405"
2.33754189'
-1.568392
Notes: Statistically significant at 1% level, 2-tailed test.
Statistically significant at 5% level, 2-tailed test.
Statistically significant at 10% level, 2-tailed test.
When the control threshold is raised to 20 percent, the results are broadly
similar, except that the Control variable is not significant (Table 2(B),
Model 1) unless the Interaction term is added to the regression (Table 2(B),
Model 2). Otherwise, as with the 10 percent control definition, both the pres-
ence of a controlling shareholder and increasing institutional holdings appear
to result in a higher return on assets.
323
MACINTOSH & SCHWARTZ
TABLE 3(A)
DETERMINANTS OF RETURN ON EQUITY - CONTROL 10%
INDEPENDENT
VARIABLE COEFFICIENT
STANDARD
ERROR T-STATISTIC
MODEL 1
ASSETS
%INST
CONTROL (10%)
Constant
Standard Error of Y est
R
2
No. of Observations
Degrees of Freedom
0.01610556
0.39340964
12.0896625
-11.778928
23.5075896
0.10554815
99
95
0.01781706
0.15699170
4-97743918
0.90394035
2.50592635'
2.42889206'
MODEL 2 (INCLUDES INTERACTION TERM)
ASSETS
%INST
CONTROL (10%)
INTERACTION
Constant
Standard Error of Y est
R
2
No. of Observations
Degrees of Freedom
0.01555183
0.53841534
15.9921547
-0.2353
-14.340624
23.5660489
0.11055609
99
94
0.01787758
0.25396371
7.32620507
0.3234364
0.86990699
2.12004835'
2.18287019"
-0.727502
Note: Statistically significant at 5% level, 2-tailed test.
RETURN ON EQUITY
THE RESULTS FOR RETURN ON EQUITY are similar to those for return on assets.
The ASSETS size control is not significant in any of the regressions. The coef-
ficients on the % INST and CONTROL variables are positive and statistically sig-
nificant with (Table 3(A), Model 2) or without (Table 3(A), Model 1) the
INTERACTION TERM. The INTERACTION term itself is negative but not significant.
When the control threshold is raised to 20 percent, the CONTROL variable
ceases to be statistically significant. The % INST variable, however, continues
to be significant both with (Table 3(B), Model 2) and without (Table 3(B),
Model 1) the INTERACTION term.
324
INSTITUTIONAL AND CONTROLLING SHAREHOLDERS
TABLE 3(B)
DETERMINANTS OF RETURN ON EQUITY - CONTROL 20%
INDEPENDENT
VARIABLE COEFFICIENT
STANDARD
ERROR T-STATISTIC
MODEL 1
ASSETS
% INST
CONTROL (20%)
Constant
Standard Error of Y est
R2
No. of Observations
Degrees of Freedom
0.01312820
0.38069448
8.33923236
-8.3990426
23.8607040
0.07847468
99
95
0.01799842
0.15929570
4.86752102
0.72940895
2.38986034'
1.71324013"
MODEL 2 (INCLUDES INTERACTION TERM)
ASSETS
% INST
CONTROL (20%)
Interaction
Constant
Standard Error of Y est
R2
No. of Observations
Degrees of Freedom
0.12609538
0.46284288
10.9376908
-0.162661
-9.8166472
23.9543410
0.08100432
99
94
0.01809779
0.22727892
7.06923874
0.319778
0.69674427
2.03645318"
1.54722329
-0.50867
Note Statistically significant at 5% level, 2-tailed test.
Statistically significant at 10% level, 2-tailed test.
To summarize, this suggests that the increasing presence of institutional
shareholders raises return on equity. It also suggests, although somewhat more
weakly, that the presence of a controlling shareholder also raises return on
equity.
325
MACINTOSH & SCHWARTZ
TABLE 4(A)
DETERMINANTS OF SALES GROWTH - CONTROL 10%
INDEPENDENT
VARIABLE
STANDARD
COEFFICIENT ERROR T-STATISTIC
MODEL 1
ASSETS
%INST
CONTROL (10%)
Constant
Standard Error of Y est
R2
No. of Observations
Degrees of Freedom
-0.0276291 0.05109070
-0.0224936 0.45017612
13.6814696 14.2728834
17.1980469
67.4083751
0.01431223
99
95
-0.5407869
-0.0499693
0.95856381
MODEL 2 (INCLUDES INTERACTION TERM)
ASSETS
% INST
CONTROL (10%)
INTERACTION
Constant
Standard Error of Y est
R2
No. of Observations
Degrees of Freedom
-0.0286734 0.05134607
0.25096130 0.72940737
21.0408758 21.0415419
-0.443735 0.9289396
12.3671443
67.6838828
0.01669912
99
94
-0.5584345
0.34406191
0.99996834
-0.477679
SALES GROWTH
THE % INST VARIABLE is NEGATIVE WITHOUT the INTERACTION term (Table
4(A), Model 1), and positive with the INTERACTION term (Table 4(A),
Model 2), but is not statistically significant in either regression. The CON-
TROL variable is positive without the INTERACTION term, and positive
with the INTERACTION term, but again is not statistically significant in either
regression. The INTERACTION term has a negative sign, but is not signifi-
cant either.
When the CONTROL threshold is raised to 20 percent, the results are
similar (Table 4(B), Model 1 and Model 2). In none of the regressions is the
ASSETS control significant.
In our sample, sales growth does not appear to be affected by either the
presence of a controlling shareholder, or by institutional holdings.
326
INSTITUTIONAL AND CONTROLLING SHAREHOLDERS
TABLE 4(B)
DETERMINANTS OF SALES GROWTH * CONTROL 20%
INDEPENDENT
VARIABLE
STANDARD
COEFFICIENT ERROR T-STATISTIC
MODEL 1
ASSETS
% 1NST
CONTROL (20%)
Constant
Standard Error of Y est
R2
No. of Observations
Degrees of Freedom
-0.0307057 0.05093946
-0.0335046 0.45084163
10.4081062 13.7761475
20.4184369
67.5310039
0.01072266
99
95
-0.6027881
-0.0743157
0.75551646
MODEL 2 (INCLUDES INTERACTION TERM)
ASSETS
% INST
CONTROL (20%)
INTERACTION
Constant
Standard Error of Y est
R2
No. of Observations
Degrees of Freedom
-0.0331060 0.05110259
0.34667099 0.64176543
22.4335455 19.9613456
-0.752783 0.9029539
13.8578867
67.6396566
0.01798372
99
94
-0.6478353
0.54018333
1.12384936
-0.83369
WINDOW DRESSING?
T
HERE IS SOME EVIDENCE THAT institutions engage in "window dressing" -
that is, selling poor performers at the end of the year and buying stocks
that are likely to be more superficially appealing to their sponsors.
68
 Because of
the timing mis-matches in our data noted earlier, we cannot be sure that some
institutions are not purchasing stocks with a high return on equity and on
assets at year's end. In other words, instead of institutional monitoring
resulting in a higher return on assets and return on equity, it may be that a
high return on assets and a high return on equity results in higher institutional
ownership.
It should be noted, however, that some of our institutional sample
(obtained from the Fairvest database) speaks as of early November 1994. This
precedes most of the year ends in the sample. If window dressing does indeed
327
MACINTOSH & SCHWARTZ
occur in the fourth quarter (as the evidence suggests),
69
 then the institutional
holdings shown in Fairvest, which tend to be the largest holdings (i.e., those
in excess of 10 percent), will tend to be relatively unaffected by possible
window dressing. It is precisely these institutional shareholdings that are most
likely to supply useful monitoring.
Second, many of the reporting dates for institutions in the Vickers
database also precede the year end. For example, of 243 institutional holders
of Canadian Pacific, 101 (just under 42 percent) reported as of a date prior to
the end of 1993. This reporting pattern appears to be broadly representative of
the database as a whole. Thus, even for the Vickers database, many institu-
tional holdings are reported prior to the end of the fourth quarter.
Perhaps most importantly, the literature offers only equivocal support for
the view that all we have detected is window dressing. Lakonishok, Shleifer,
Thaler & Vishny
70
 investigated a large sample of (mostly private) U.S. pension
funds to determine whether these institutions engaged in window dressing to
curry favour with plan sponsors. The study found some evidence of window
dressing by comparing fourth quarter selling and purchasing behaviour of the
sample funds to their behaviour in the first three quarters. In the fourth
quarter, there was evidence that funds were more likely to sell losers (stocks
realizing returns in the previous year in the two lowest quintiles of the entire
CRSP data sample, which consists of firms traded on the New York Stock
Exchange, American Stock Exchange, and over-the-counter).
71
However, the evidence that institutional investors tend to accelerate
their selling of losers in the last quarter was strongly offset by the finding that,
in general, institutional traders are contrarian in their trading strategies. After
reviewing trading behaviour during the first three quarters, the authors
concluded that:
72
... when it comes to purchases, funds are clearly contrarian: relative to avail-
ability they overbuy losers... and underbuy winners. ... Second, when it comes
to sales, funds oversell winners relative to their holdings ... but they also over-
sell losers. ...
73
The Lakonishok et al. study found, however, that smaller funds had a
stronger propensity to dump losers in the fourth quarter than larger funds.
74
Although there was weak statistical evidence, as noted above, that the
funds accelerated their soles of losers, there was robust evidence that the funds
accelerated their purchases of losers. While initially this appears counter-intuitive,
the authors suggest a plausible explanation: the continued holding of past
losers may be difficult to defend to plan sponsors. However, it may not be difficult
to defend the buying of stocks that were losers before they were held by the fund.
This evidence strengthens, rather than detracts from our results. In general,
the funds in the sample were shown to be contrarian investors: they sold
winners and bought losers. Assuming that ROE and ROA are correlated with
328
INSTITUTIONAL AND CONTROLLING SHAREHOLDERS .
stock returns (the basis for the study by Lakonishok et ai.), then institutions
engaging in contrarian trading will sell, rather than buy stock with high ROE
and ROA. The Lakonishok et al. study suggests that the existence of window
dressing should not affect this result, since the accelerated sales of losers in the
fourth quarter were offset by the accelerated purchases of losers. The net effect
on the balance of winners and losers in the overall portfolio should thus be
minimal.
A second study by Lakonishok, Shleifer & Vishny, however, is also
worthy of note. The second study used the same data sample, disaggregated by
size, to determine if institutions engage in positive feedback trading. Positive
feedback (PF) trading is the opposite of contrarian trading. It occurs when an
institution buys last quarter's winners, and sells last quarter's losers. The study
found robust evidence of PF trading in the two smallest quintiles of traded
firms by size, but no evidence of such trading in the three largest quintiles.
These studies using American data are merely suggestive of what might
occur in the Canadian market. On average, Canadian funds are smaller than
their U.S. counterparts. Further, TSE stocks tend to correspond in size to those
in the lowest two quintiles in the Lakonishok et al sample. Obviously, we can-
not rule out the possibility that window dressing or, more generally, PF trading
strategies drive our results.
We note, however, that because institutional blocks tend to be large
compared to retail holdings, and because it is time-consuming and expensive
for brokers (and hence their institutional clients) to break up large blocks,
much institutional trading is consummated with other institutions. Hence, it
is not entirely clear that window dressing (or more generally PF trading
strategies) will result in a large increase in the percentage of a particular firm
that is institutionally held. For every institutional buyer/seller, there must be a
seller/buyer on the other side of the transaction. If that trader is another
institution engaging in a different trading strategy, then such trading will effect
no change at all in the percentage of a given firm that is institutionally held.
It is also likely that some retail traders engage in PF trading strategies. In
theory, retail PF trading might exceed institutional PF trading. If this is the
case, then firm profitability will tend to lead to decreased, rather than
increased, institutional ownership. We have no evidence that this is the case,
but present it as a theoretical possibility. At the very least, it seems unrealistic
to believe that all institutional purchases made for the purpose of window
dressing (or PF trading) will be made by retail sellers. It is also unrealistic to
believe that all institutions engage in window dressing (or PF trading).
Finally, we note that our regressions yield some support for the view that
as institutional holdings rise, the price-to-book ratio falls. Window dressing,
however, should create an artificial demand for certain stocks based on factors
other than risk/return characteristics, since the only end in view is to make
the portfolio holdings more superficially appealing to plan sponsors. It should
thus create a price pressure effect, driving the price upward and increasing the
329
MACINTOSH  & SCHWARTZMACINTOS H & SCHWARTZ
price-to-book ratio. We do not observe this. In short, we think it is unlikely
that window dressing or PF trading drive our results.
THE PmcE-ToBooK RATIO AND UNTRADED EQUITY
M
ANY CANADIAN CORPORATIONS have more than one class of common
shares. In some cases, one class of common shares is publicly traded,
while the other is not. In calculating the price-to-book ratio as defined earlier,
however, the Financial Post uses the book value for all common shares, but the
price for only the traded class of equity. Where this is the case, the price-to-
book ratio will be biased downward.
In fact, it is not uncommon for controlling shareholders to hold a class of
superior voting equity, whether traded or non-traded. Where such equity is
traded, it tends to trade at a premium to the restricted- or non-voting equity.
76
This downward bias on the price-to-book ratio offers an alternative
explanation for some of the regression results in which price-to-book ratio is
the dependent variable. In particular, the regressions appear to show that the
presence of a controlling shareholder has a negative effect on the price-to-
book ratio. This may be an artifact of the manner in which the price-to-book
ratio is computed.
SUMMARY AND CONCLUSIONS
THE ROLE OF INSTITUTIONAL SHAREHOLDERS
TAKEN AS A WHOLE, THESE RESULTS OFFER the first empirical support in Canada
for the hypothesis that institutional investors increase firm value. There is a
positive and statistically significant relationship between both return on assets
and return on equity (two commonly used measures of profitability) and
institutional holdings. Although in all regressions of price-to-book against
institutional holdings, the coefficient on institutional holdings was wrongly
signed to support the hypothesis that an increased institutional presence
increases value, in none of these regressions was the coefficient statistically
significant.
There is also some support for the hypothesis that institutional monitoring
acts to reduce the danger of redistributive transactions engineered by control-
ling interests. The price-to-book ratio is the measure of profitability most likely
to be affected by redistributive transactions effected by controlling shareholders.
Thus, if institutional shareholders play a useful role in monitoring for
redistributive events, the interaction term should have its strongest effect
where the dependent variable is the price-to-book ratio. This is, in fact, the
case. The interaction term was negatively signed and was not statistically
significant in the regressions of ROA, ROE, and sales growth on the three
independent variables - institutional holdings, control, and the interaction
330
INSTITUTIONAL AND CONTROLLING SHAREHOLDERS
term. However, it was positive and statistically significant in one of two regres-
sions in which price-to-book was the dependent variable, and nearly significant
at conventional levels in the second. Thus, there is some support for the
hypothesis that institutional shareholders monitor for redistributive events.
The evidence as a whole tends to confirm the hypothesis that increased
institutional holdings result in a more profitable firm. This suggests that
institutional investors do indeed perform useful monitoring of corporate
managers and/or controlling shareholders, and that this monitoring is of benefit
to other non-controlling shareholders.
That the percentage of institutional ownership achieved statistical
significance in many of the regressions is particularly interesting given the
"noisiness" of the institutional ownership database. Database errors are likely
to reduce the chances of finding a statistically significant relationship. As
indicated earlier, there are a number of database errors in our institutional
ownership sample, stemming from the difficulties we encountered in identifying
all institutional owners, the timing mis-match between the two sources of
institutional ownership data, and the timing mis-match between the owner-
ship data and the accounting data.
Given that the institutional ownership sample combined voting, restricted-
voting, and non-voting shares, the results suggest that institutions derive a
lever over management from sources other than voting power. Further investi-
gation (through stratification of the sample into voting, restricted-voting, and
non-voting shares) is in order.
Because the institutional ownership sample focused on the percentage of
institutional ownership in single traded classes of equity (where in some cases
the firm had two classes of equity), the results suggest that institutions derive
power from holding large positions in single equity classes. This too deserves
further investigation through further stratification of the sample.
One important caveat is in order. There is the possibility that our results
are driven by "window dressing" or, more generally, positive feedback trading
by institutional managers. For the reasons given earlier, we do not think that
this is the case, but we cannot rule it out.
THE ROLE OF CONTROLLING SHAREHOLDERS
THE PRESENCE OF A CONTROLLING SHAREHOLDER had a somewhat more
ambiguous effect on firm profitability. There is fairly strong support for the
hypothesis that the presence of a controlling shareholder resulted in a lower
price-to-book ratio. However, there is even stronger support for the hypothesis
that the presence of a controlling shareholder resulted in higher returns on both
assets and equity, although there was no discernible effect on sales growth.
Except for the absence of an effect on sales growth, these results are
consistent with our hypotheses about the effect of a controlling shareholder
on firm value. Earlier we hypothesized that the presence of a controlling
331
MACINTOSH & SCHWARTZ MACINTOS H & SCHWART Z
shareholder should result in better monitoring of managers. This in turn
should result in higher ROA and ROE (which we assume are not as likely to
be affected by redistributive transactions as price-to-book). However, in an
efficient market, where controlling shareholders regularly engage in some
redistribution of profits at the expense of non-controlling interests, the price
of firms with controlling shareholders will be discounted to reflect this risk.
The fact that the price-to-book ratio is less when there is a controlling
shareholder suggests that, even though such corporations generate higher
profits, these profits are siphoned off by controlling shareholders.
Again, there is a caveat. Because of the way in which the price-to-book
ratio is computed, there may be a built-in downward bias when there is a
controlling shareholder. Our results in relation to the price-to-book ratio are
therefore more tentative than in relation to ROA, ROE, and sales growth.
POLICY IMPLICATIONS
IF INSTITUTIONS DO, INDEED, PERFORM USEFUL MONITORING of both corporate
managers and controlling shareholders, this lends weight to the arguments of
Black,
77
 Roe,
78
 Macintosh
79
 and others that legal restraints on institutional
activism tend to impede effective monitoring. In Canada, there are a variety
of such restraints, many of which could be lessened or removed without
adverse consequences.
80
 There are also a number of ways in which regulation
could be enhanced to promote institutional activism (such as introducing
confidential voting), and to reduce the extent to which institutions can be
co-opted into voting with management. Our results should encourage policy
makers to consider both removing legal restraints that are not cost-effective,
and introducing measures designed to facilitate the institutional role in
corporate governance.
Do any policy implications derive from our results in relation to
controlling shareholders? While the results appear to be consistent with our
hypothesis that controlling shareholders both monitor managers and engage
in redistribution, the evidence is far from conclusive. Moreover, as indicated
above, Morck & Stangeland find evidence that not all types of controlling
shareholders have the same effect on firm value. The controlling shareholder
data must be further stratified to determine the effect of different types of
controlling shareholders on value. Thus, we do not draw any specific policy
conclusions.
However, supposing that controlling shareholders do indeed engage in
redistribution of wealth from non-controlling interests, this is not necessarily a
prescription for any particular market intervention (such as strict control of
related-party transactions). As long as shareholders are fully cognizant of the
danger of controlling shareholder predations, and stock prices are adjusted
accordingly, shareholders of controlled companies should earn normal returns.
332
INSTITUTIONAL AND CONTROLLING SHAREHOLDERS
Arguably, there is a danger that a non-controlled firm may be acquired
by a controlling shareholder, inflicting unanticipated losses on shareholders.
There appear to be two mitigating factors, however. In an economy in which
control is the rule rather than the exception, it is possible that a discount is
built into the stock price of non-controlled companies to reflect this danger.
Further, not all types of control transactions present the same risk to share-
holders. An any-or-all takeover offer presents the least risk, given that share-
holders can fully cash out at a premium if they choose, and avoid the risk of
subsequent looting or expropriation.
A partial takeover apparently presents somewhat greater risks. The
implicit back-end price (i.e., the market-trading price after the event) is
almost invariably less than the front-end price. Some have argued that this
effectively coerces shareholders into tendering.
81
 If there is a danger of redistri-
bution following the acquisition of control, this will magnify the extent of the
coercion. The takeover may succeed, even though shareholders as a whole are
worse off as a result, and the controlling shareholder may thus profit by
purchasing control and adopting a strategy of redistributing assets in its favour.
A private change of control, where redistribution ("looting") is the motive for
the transaction, may similarly reduce shareholder wealth with no efficiency
benefits.
We note, however, that the magnitude of the redistribution necessary to
make a looting strategy profitable is necessarily large, and would likely attract
legal (or administrative) intervention. We also note that the U.S. evidence
casts doubt on the frequency with which control is purchased for the purpose
of effecting redistribution. In addition to Holderness & Sheehan's evidence
(above), Barclay & Holderness find that private changes in control tend to
result in significant benefit to minority shareholders.
82
 There is also evidence
tending to refute the theory that partial takeovers are coercive.
83
 Further
investigation of the role of controlling shareholders in Canadian markets is
clearly in order.
333
MACINTOSH & SCHWARTZ
ENDNOTES
1 Adolf A. Berle, Jr. & Gardiner C. Means, The Modern Corporation and
Private Property, New York: Macmillan Co., 1932.
2 Michael Jensen & William Meckling, "Theory of the Firm: Managerial
Behaviour, Agency Costs, and Ownership Structure," journal of Financial
Economics, 3, 1976, p. 305.
3 A notable exception is Demsetz, who hypothesizes that the optimal own-
ership structure is determined endogenously by a variety of factors which
may vary from firm to firm and from industry to industry. Consequently, in
his view, there should be no relation between ownership structure and firm
performance. See Harold Demsetz, "The Structure of Ownership and the
Theory of the Firm," Journal of Law & Economics, 26, 1983, p. 375.
4 See, e.g., Michael C. Jensen, "The Takeover Controversy: Analysis and
Evidence", in Knights, Raiders and Targets: The Impact of the Hostile
Takeover, edited by John C. Coffee, Jr. et al., New York, Oxford University
Press, 1988, p. 314.
5 Stulz provided the theoretical basis for the view that as managerial owner-
ship increases, firm value will decrease owing to the diminished likelihood
of a takeover bid. In Stulz' model, as management ownership increases, the
premium that an acquirer must pay to purchase control increases, while
the probability of an offer occurring decreases. At 50 percent managerial
ownership, the probability of a hostile bid is reduced to zero, and a
minimum firm value is achieved. See Rene Stulz, "Managerial Control of
Voting Rights, Financial Policies, and the Market for Corporate Control,"
journal of Financial Economics, 20, 1988, p. 25. Morck et al. hypothesize (as
in Jensen & Meckling, supra, note 2) that increasing managerial owner-
ship will align managerial and shareholder interests, while also (as in
Stulz) reducing the probability of a takeover bid. See Randall Morck,
Andrei Shleifer & Robert W. Vishny, "Management Ownership and
Market Valuation," Journal of Financial Economics, 20, 1988, p. 293.
6 A pithy summary of the theory and evidence may be found in Randall K.
Morck, "On the Economics of Concentrated Ownership", paper prepared
for Canadian Corporate Governance: An Interdisciplinary Perspective,
C.D. Howe Institute, February 10-11, 1994.
7 Studies on this point are conveniently summarized in Vijay Jog & Ajit
Tulpule, "Control and Performance: Evidence from the TSE 300", in this
volume.
8 Randall Morck, Andrei Shleifer & Robert W. Vishny, "Management
Ownership and Market Valuation," Journal of Financial Economics, 20,
1988, 293.
9 J. J. McConnell & H. Servaes, "Additional Evidence on Equity Ownership
and Corporate Value," Journal of Financial Economics, 27, 1990, p. 595.
334
INSTITUTIONAL AND CONTROLLING SHAREHOLDERS . .
10 Karen Hopper Wruck, "Equity Ownership Concentration and Firm Value,"
Journal of Financial Economics, 23, 1988, p. 3.
11 See, e.g., Michael J. Barclay & Clifford G. Holderness, "Private Benefits
From Control of Public Corporations" Journal of Financial Economics, 25,
1989, p. 371; Michael J. Barclay & Clifford G. Holderness, "Negotiated
Block Trades and Corporate Control" journal of Financial Economics, 46,
1991, p. 861.
12 Morck et al. supra, note 4, at 309-10 interpret this result in the following
manner:
It suggests that outside board members, like officers, respond to financial
incentives and contribute more to corporate wealth as their ownership stakes
rise. In addition, it suggests that outside board members are capable of becom-
ing entrenched. Since outside board members are less likely than top officers
to enjoy corporate perks, such entrenchment perhaps takes the form of
unchecked deployment of corporate wealth into projects that the board, but
not necessarily the market, considers desirable. Finally, this interpretation
suggests that for outside board members, as well as for top officers, the
convergence-of-interests effect again dominates at very high ownership levels.
13 The demise of the junk-bond market also had a significant effect on the
ability of acquirers to complete hostile acquisitions successfully, although
at the time of writing, this market is staging a powerful resurgence.
14 The use of poison pills is policed by the courts, limiting the managers'
freedom to abuse the pill to entrench themselves. While pills are thus
imperfect entrenchment devices, the evidence suggests that managers can
indeed use poison pills to service their own, rather than shareholders' best
interests. See Jeffrey G. Macintosh, "Poison Pills in Canada: A Reply to
Dey and Yalden," Canadian Business Law Journal, 17, 1991, p. 323; Jeffrey
G. Macintosh, "The Poison Pill: A Noxious Nostrum for Canadian
Shareholders," Canadian Business Law Journal, 18, 1989, p. 276; Jeffrey G.
Macintosh, "Are Poison Pills Good for Shareholders? The Empirical
Evidence," Corporate Governance Review, 6 (6), 1994, p. 2.
15 See those articles cited ibid.
16 See, e.g., 347883 Alberta Ltd. v. Producers Pipelines Ltd. (1991), 80 D.L.R.
(4th) 359 (Sask. C.A.); Remington Energy Ltd. v. Joss Energy Ltd., unreported,
Alberta Court of Queen's Bench, per Fraser J., Dec. 17, 1993; Re MDC
Corporation and Regal Greetings & Gifts Inc., Ontario Securities
Commission, Oct. 13, 1994; Re Lac Minerals Ltd., Ontario Securities
Commission, Oct. 13, 1994.
17 Randall Morck & David A. Stangeland, "Corporate Performance and
Large Shareholders," Working Paper No. 4-94, Institute for Financial
Research, Faculty of Business, University of Alberta.
18 Indeed, this classification is an oversimplification, since there will often be
a divergence of incentives even within these classes. Public and private
335
pension funds, for example, face rather different incentives to get involved
in matters of corporate governance.
19 These have been well developed in the extensive U.S. literature dealing
with institutional shareholders, and are summarized in Jeffrey G. Macintosh,
"Institutional Investors and Corporate Governance in Canada", paper
prepared for Canadian Corporate Governance: An Interdisciplinary
Perspective, C.D. Howe Institute, February 10-11, 1994 (Canadian
Business Law Journal, forthcoming).
20 Morck et al., supra, note 5, at 310.
21 See, e.g., Clifford P. Holderness &. Dennis P. Sheehan, "The Role of
Majority Shareholders in Publicly Held Corporations: An Exploratory
Analysis," Journal of Financial Economics, 20, 1988, p. 317.
22 This is done in the Canada Business Corporations Act, R.S.C. 1985, c. C-44
(CBCA), for example, in connection with an amalgamation (s. 183(3)),
sale of all or substantially all the assets of the corporation (s. 183(5)), and
continuance in another jurisdiction (s. 188(4)).
23 This is done in the CBCA, for example (with some limitations), in con-
nection with an amalgamation (s. 183(4)), sale of all or substantially all
the assets of the corporation (s. 189(6)), and amendments to the articles of
incorporation (s. 176).
24 Related parties include directors, officers, and significant shareholders.
25 See generally "Note on Statutory and Judicial Voting Entitlements," in
Cases and Material on Partnerships and Canadian Business Corporations,
edited by Jacob S. Ziegel et al., Toronto: Carswell, 1995, 3rd edition.
26 See generally Jeffrey G. Macintosh, "Minority Shareholder Rights in Canada
and England: 1860-1987," Osgoode Hall Law Journal, 27, 1989, p. 561.
27 See Philip Anisman, "The Commission as Protector of Minority
Shareholders," in Special Lectures of the Law Society of Upper Canada, 1989:
Securities Law in the Modern Financial Marketplace, Toronto: Richard De
Boo, 1989, p. 451; Jeffrey G. Macintosh, "Corporations," in Law Society of
Upper Canada, Special Lectures, 1990: Fiduciary Duties,Toronto: Thomson
Professional Publishing Canada, 1990, p. 189.
28 See, e.g., Ronald Gilson, "The Case Against Shark Repellant Amendments:
Structural Limitations on the Enabling Concept," Stanford Law Review, 34,
1982, p. 775.
29 Supra, note 6.
30 Supra, note 17.
31 R. Daniels & P. Halpern, "The Canadian Quandary: Accounting for the
Survival of the Closely Held Corporation", paper prepared for Canadian
Corporate Governance: An Interdisciplinary Perspective, C.D. Howe
Institute, February 10-11, 1994, (Canadian Business Lau> Journa!, forth-
coming).
32 Holderness & Sheehan, supra, note 21.
33 In a window beginning 20 days prior to the trade and ending ten days after
336
MANCINTOSH& SCHWAETZ
INSTITUTIONAL AND CONTROLLING SHAREHOLDERS
the trade, Holderness & Sheehan found a statistically significant increase
in the value of market-traded equity of 12.8 percent. Ibid.
34 Controlling shareholders might use their powers of control to force managers
to choose accounting procedures that reduce reported income, in order to
reduce return on assets and return on equity. They might do so to conceal
the true magnitude of the firm's profits, in order to facilitate redistribution
to themselves from non-controlling interests. This suggests that the
presence of a controlling shareholder may indeed negatively affect return
on return on assets and return on equity. We think this unlikely, however.
As an empirical matter, the market would appear to adjust in an unbiased
manner to changes in accounting methods. If this is the case, then
influencing the selection of accounting methods will not enable control-
ling shareholders to conceal their predations.
35 R. Morck & D. Stangeland, "Corporate Performance and Large
Shareholders: An Empirical Analysis," supra, note 17; Morck, et al. supra,
note 5. See also B. Johnson, R. Magee, N. Nagarajan, & H. Newman, "An
Analysis of the Stock Price Reaction to Sudden Executive Deaths:
Implications for the Managerial Labour Market," Journal of Accounting &
Economics, 7, 1985, p. 151.
36 Morck & Stangeland, supra, note 17.
37 Holderness &. Sheehan, supra, note 21.
38 A control shareholder "appoints" management indirectly, by virtue of the
ability to elect a majority of directors. The directors formally appoint the
officers.
39 Holderness &. Sheehan, supra, note 21.
40 Bernard S. Black, "Shareholder Passivity Reexamined," Michigan Law
Review, 89, 1990, p. 520. In the Canadian setting, see Macintosh,
"Institutional Investors and Corporate Governance in Canada," supra,
note 19.
41 James A. Brickley, Ronald C. Lease, & Clifford W. Smith, Jr., "Ownership
Structure and Voting on Anti-takeover Amendments," Journal of Financial
Economics, 20, 1988, p. 267.
42 Agrawal & Mandelker, "Large Shareholders and the Monitoring of
Managers: The Case of Antitakeover Charter Amendments," Journal of
Finance & Quantitative Analysis, 25, 1990, p. 143.
43 Ibid.
44 McConnell & Servaes, supra, note 9. See generally Bernard S. Black, "The
Value of Institutional Investor Monitoring: The Empirical Evidence,"
U.C.L.A. Law Review, 39, 1992, p. 986.
45 Macintosh, supra, note 19.
46 Ibid.
47 Ibid. This may well be due to the movement of bank balances into money
market mutual funds, rather than a movement of diversified mutual funds
into non-equity instruments.
337
H
48 See Macintosh, "Institutional Investors and Corporate Governance in
Canada," supra, note 19; Jeffrey G. Macintosh, "The Role of Institutional
and Retail Shareholders in Canadian Capital Markets," Osgoode Hall Law
Journal, 32, 1994, p. 371; R.J. Daniels & E.J. Waitzer, "Challenges to the
Citadel: A Brief Overview of Recent Trends in Canadian Corporate
Governance," Canadian Business Law Journal, 23, 1994, p. 23; K.E.
Montgomery, "Survey of Institutional Shareholders," Corporate Governance
Review, ("Survey") 4:4, 1992, p. 5; Kathryn E. Montgomery, "Shareholder
Activism in Canada: A Survey of Institutional Shareholders,"
("Shareholder Activism"), Working Paper Series No. NC 92-014-B,
Western Business School, University of Western Ontario.
49 Ibid.
50 Public pension funds will wish to avoid the spotlight in order to avoid a
political backlash. See, e.g., Mark J. Roe, "A Political Theory of American
Corporate Finance," Columbia Law Review, 91, 1991, p. 10.
51 See Macintosh, supra, note 48.
52 See Macintosh, ibid.
53 See Macintosh, supra, note 48, for a number of proposals for reform of
Canadian legislation to permit a more activist role.
54 John Pound, "Proxy Contests and the Efficiency of Shareholder
Oversight," Journal of Financial Economics, 20, 1988, p. 237.
55 Brickley, Lease &. Smith, supra, note 41.
56 See e.g., Michael E. Porter, "Capital Choices: Changing the Way America
Invests in Industry." Journal of Applied Corporate Finance, 5:2, (1992):4.
57 With respect to U.S. data see, e.g., Office of the Chief Economist,
Securities and Exchange Commission, "Institutional Ownership, Tender
Offers, and Long-Term Investments," April 19, 1985. With respect to
Canadian data, see Lewis D. Johnston & Bohumir Pazderka, "Firm Value and
Investment in R&D," Managerial and Decision Economics, 14, 1993, p. 15.
58 See Morck &. Stangeland, supra, note 17, although see also Holderness &
Sheehan, supra, note 21.
59 Macintosh, supra, note 48.
60 See National Policy 39.
61 The Vickers database lists the class of shares held, but does not indicate
whether the class is voting or non-voting. Knowing the identity of the
share class held by institutional investors, it was possible to determine the
voting power of the shares. However, it was not feasible within the time
and resource constraints of our study.
62 In future iterations of this study, we hope to segregate the sample according to
the voting power of the institutional holdings, in order to determine which
holdings supply institutions with the greatest leverage over management.
63 The greatest source of managerial sensitivity to share price comes not from
options or bonuses, but from managerial shareholdings. See Michael C.
Jensen & Kevin J. Murphy, "Performance Pay and Top Management
338
MACINTOSH & SCHWARTZ
INSTITUTIONAL AND CONTROLLING SHAREHOLDERS
Incentives," Journal of Political Economy, 98, 1990, p. 225. While Jensen &
Murphy demonstrate that management remuneration is not highly sensitive
to firm performance, managerial share holdings may well give managers a
good incentive to pay attention to share price.
64 In the case of Canada Malting, two corporate shareholders (John Labatt
and Molson Brewing) each held just under 20 percent of the common
shares. While there is evidence that these shareholders have in fact acted
jointly in relation to some corporate decisions, we were not prepared to
assume that they would always do so. Indeed, a disagreement between the
two shareholders would result in the votes of one effectively cancelling the
votes of the other. Thus, we classified this company as not controlled.
65 There is evidence suggesting that smaller firms have higher risk adjusted
returns than larger firms (the "small firm effect"). There is also evidence
that such firms have lower price-to-book ratios. See, e.g., Avner Arbel,
"Generic stocks: An old product in a new package." Journal of Portfolio
Management, Summer 1985, p. 4. In turn, there is evidence that price-to-
book ratios are predictive of shareholder returns. See generally Eugene F.
Fama, "Efficient Capital Markets: II," Journal of Finance, 46, 1991, p. 1575.
We thought it prudent to control for the effect of size on the various
accounting measures of profitability discussed below.
66 See generally E. Lindenberg & S. Ross, "Tobin's Q Ratio and Industrial
Organization," Journal of Business, 54, 1981, p. 1.
67 Stephen A. Ross & Randolph W. Westerfield, Corporate Finance, St. Louis:
Times Mirror/Mosby College Publishing, 1988, p. 41.
68 See, e.g., ]. Lakonishok, A. Shleifer, R. Thaler & R. Vishny, "Window
Dressing by Pension Fund Managers," American Economic Review, 81,
1991, p. 227.
69 Lakonishok, ibid.
70 Ibid.
71 Note, however, that the study was unable to reject the null hypothesis of
no difference in the selling intensity of losers as between quarters 1 to 3
and quarter 4 (the t-statistic for difference in means was only 1.75). See
ibid, at 230.
72 Ibid, at 230.
73 See J. Lakonishok, A. Shleifer & R.W. Vishny, "The Impact of
Institutional Trading on Stock Prices," Journal of Financial Economics, 32,
1992, p. 23 (finding that institutional investors tend to engage in contrarian
trading).
74 Ibid, at 231.
75 In this case, the t-statistic for the difference in means of a statistic reflecting
purchasing intensity was 5.47, allowing rejection of the null hypothesis of
no difference in means. It should also be noted that while there was some
evidence that the funds slowed their sales of winners, this did not achieve
statistical significance. Ibid, at 230.
339
MACINTOSH & SCHWARTZ
76 Vijay M. Jog & Allan L. Riding, "Market Reactions of Return, Risk, and
Liquidity to the Creation of Restricted Voting Shares," Canadian journal of
Administrative Sciences, March 1989, p. 62.
77 Supra, note 40.
78 Supra, note 50.
79 Supra, note 48.
80 Ibid.
81 See generally Macintosh, "The Poison Pill: A Noxious Nostrum for
Canadian Shareholders," supra, note 14.
82 "Negotiated Block Trades and Corporate Control," supra, note 11.
83 The evidence is summarized in Macintosh, "The Poison Pill: A Noxious
Nostrum for Canadian Shareholders," supra, note 14.
340
Michel Patry & Michel Poitevin
Department of Economics Department of Economics
Ecole des HEC and CIRANO Universite de Montreal, CRDE and CIRANO
Why Institutional Investors are not
Better Shareholders
INTRODUCTION
T
HE GROWING INTEREST IN GOVERNANCE REGIMES and their effects on the
performance of firms has generated a widespread and interesting debate —
with institutional investors often at the centre of that debate. Voices are being
heard that call for banks, mutual funds and pension funds to play a more active
and effective role in monitoring managers, even to the point of guiding
enterprises on some issues.
The apparent success of Japanese and German firms has fuelled this
interest in governance regimes and firm performance and has led many
observers to suggest that Canada and the United States should consider imple-
menting changes to their system(s) of corporate governance that would
encourage long-term partnerships at the price of liquidity. Many scholars
1
 have
argued that the weak role played by institutional investors in North America
is attributable to legal and institutional barriers. Our approach is somewhat
different in that we concentrate on two determinants of institutional
investors' activism: their own governance and organizational structure.
Our focus is particularly on mutual funds and pension funds and we ask:
do managers of mutual funds and pension fund managers have an incentive to
monitor and influence the management of the firms in which they invest? Can
they effectively monitor and influence corporate policy? What cost-benefit
relationship supports their decisions? Finally, how can we make institutional
investors more active partners?
To answer these questions, we first present an overview of the role and
informational structure of financial contracts. We then examine how the exit
and voice mechanisms function to control managerial behaviour. A brief outline
of the importance of institutional investors follows. We continue with an
examination of the governance and internal organization of mutual funds and
pension funds and follow with an analysis of their consequences on the level of
monitoring and influence on activities. Finally, we offer a critical assessment of
some of the propositions designed to improve the governance of Canadian firms.
34 /
PATRY & POITEV1N
THE NEED FOR CORPORATE GOVERNANCE
T
HIS SECTION PROVIDES AN OVERVIEW OF THE ROLE and functions of financial
markets in modern capitalistic economies. It has four parts. First, we
examine the role played by financial markets in modern economies. The
second part is devoted to financial market imperfections and presents some
evidence suggesting these imperfections are real. The third part focuses on the
"institutional" response to these imperfections, namely the emergence of
contracts. The last part examines the governance of financial contracts and its
related objectives and problems.
THE ROLE OF FINANCIAL MARKETS
FIRMS AND HOUSEHOLDS EVOLVE in an economic environment that is inherently
uncertain. These agents have strong incentives to insure risks as well as to
smooth fluctuations associated with their income. Insurance needs arise when
unexpected losses are possible; smoothing needs arise when income is expected
to fluctuate over time. For example, a firm may want some insurance against
foreign exchange risk if it is doing business in a foreign country; it may also want
to smooth its income if its line of business fluctuates seasonally. On a personal
level, a household may wish to insure itself against potential losses related to
accidents or unemployment, while at the same time seeking to smooth its
income over the period when all members are working and into retirement.
In an economy with complete contingent markets, the demand for
insurance and smoothing could be fully satisfied. Because of prohibitive trans-
action costs, however, many contingent markets do not exist. It can be argued,
therefore, that financial markets have arisen as a substitute for absent contingent
markets. If agents cannot transfer goods and services across all zones and states
of the world, as they would with complete contingent markets, sophisticated
financial securities will allow them to transfer income over time and through-
out the world.
If financial markets merely provide insurance and income smoothing,
why are they so often associated with growth and investment? It is easy to
reconcile this view of financial markets with the view presented above.
Suppose a firm wants to undertake a large and costly investment. Without
financial markets, the firm would have to obtain its financing through
entrepreneurial funds, or through credit from its suppliers. In both cases, this
would impose large risks on the financing party. Well-functioning financial
markets spread such risks across a large number of investors. This diversifica-
tion effectively reduces the financial cost of the investment, and therefore
promotes investment and growth. This explains why sophisticated financial
markets are a primary source of growth and investment.
As with most markets, financial markets have developed a structure
where a layer of intermediaries links the buyers and sellers of securities.
342
WHY INSTITUTIONAL INVESTORS ARE NOT BETTER SHAREHOLDERS
Traditional theories of financial intermediation argue that these inter-
mediaries allow the matching of diverse term horizons (some buyers may
prefer short-term securities, while sellers may be more interested in long-term
securities); and pool a large number of small buyers to accommodate the needs
of large sellers. We now argue that the role of financial intermediaries becomes
crucial when financial markets are plagued with imperfections.
IMPERFECTIONS OF FINANCIAL MARKETS
A FINANCIAL SECURITY IS A PROMISE TO PAY some predetermined financial
return in the future in exchange for an immediate payment. Some securities,
like bonds or loans, promise a fixed specified return, while others, like shares,
promise an unspecified dividend stream. Still others are more complicated and
may promise to return combinations of these basic securities, and occasionally
even returns based on these basic securities (derivatives).
What would constitute a reasonable expectation if financial markets
were complete and perfect? If financial markets provide smoothing and insurance,
agents' profits or consumption should be fairly constant over time, or at least
should be perfectly correlated because all idiosyncratic risks would be diversi-
fied and aggregate risks would be shared by all agents. Yet, even a cursory
examination of macro-economic data reveals that agents' idiosyncratic risks
are not perfectly diversified.
Tables 1 and 2 show the sources and uses of funds of non-financial
corporations in Canada and in the United States. Examination of the tables
suggests that financial markets do not provide complete insurance or smoothing.
Two features shown by the tables deserve comment. First, the reliance on
internal funds for investment increases during recessions. Second, financial
slack, measured by the change in marketable securities, increases during boom
periods. We argue that such features should not arise in the presence of perfect
financial markets.
Suppose that corporate profits are lower during recessions than during
boom periods. Perfect financial markets should then direct the flow of funds
toward firms during recessions, and toward investors during booms. The data
clearly contradict this assumption. Firms use more internal funds during reces-
sions and increase their financial slack during booms — exactly the opposite of
what perfect financial markets should achieve. This shows that agents cannot
fully insure themselves and smooth their consumption. They must rely on
internal funds when money is tight, and build up financial slack in good times,
rather than reimburse investors for the financing provided in bad times.
The sources of imperfections in financial markets are the investor's lack
of information and the manager's lack of commitment to the investor. These
problems arise as a result of the separation of ownership and control. Recent
theory emphasizes three main problems: moral hazard, adverse selection, and
imperfect commitment.
343
PATRY & POITEV1N
TABLE 1
SOURCES AND USES OF FUNDS OF NON-FINANCIAL CORPORATIONS IN CANADA
SOURCES
YEAR
1969
1970*
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980*
1981
1982*
1983
1984
1985
1986
1987
1988
1989
1990*
1991
1992
Note:
Source:
TOTAL
USES
CAPITAL FINANCIAL
($ BILLIONS) INTERNAL (%)
11.1
10.3
14.3
14.8
19.7
28.1
24-1
27.6
30.1
42.0
57.2
60.4
75.2
32.1
40.4
59.9
66.0
72.7
84.0
93.2
105.9
81.4
69.8
66.2
* Trough of the recession.
60.7
68.3
52.7
58.8
53.3
42.4
55.0
60.8
56.2
43.9
40.6
44.4
29.7
58.3
77.0
64.5
67.1
56.7
59.0
55.5
48.3
54.1
58.9
63.0
EXTERNAL (%) EXPENDITURES (%) ASSETS (%)
39.3
31.7
47.3
41.2
46.7
57.6
45.0
39.2
43.8
56.1
59.4
55.6
70.3
41.7
23.0
35.5
32.9
43.3
41.0
44.5
51.7
45.9
41.1
37.0
Tables A.2 13-2/04 and E.3-2/04 Sources and Uses of Funds of Non
87.4
91.5
77.8
83.6
79.9
73.0
70.7
76.9
64.2
52.5
60.5
62.1
59.1
79.2
79.0
69.3
71.5
68.0
68.0
74.8
80.8
85.5
88.4
82.5
Financial Corporations,
12.6
8.5
22.2
16.4
20.1
27.0
29.3
23.1
35.8
47.5
39.5
37.9
40.9
20.8
21.0
30.7
28.5
32.0
32.0
25.2
19.2
14.5
11.6
17.5
OCDE Financial Statistics 1977 and 1994.
We highlight these information and commitment problems by means of
the following simple example. A financial investor is looking for profit oppor-
tunities, and an entrepreneur possesses a project which is potentially lucrative.
The entrepreneur has private information about the quality or profitability of
his project, which depends on how much effort he will exert, among other
things. In addition, the life of the project is long and the entrepreneur may
need refinancing at some point in the future if initial profits are low.
MORAL HAZARD
SINCE MANAGERIAL EFFORT IS NOT DIRECTLY OBSERVABLE by the financier, the
entrepreneur's compensation is necessarily based on a number of noisy signals
344
WHY INSTITUTIONAL INVESTORS ARE NOT BETTER SHAREHOLDERS
TABLE 2
SOURCES AND USES OF FUNDS OF NON-FINANCIAL CORPORATIONS
IN THE UNITED STATES
YEAR
1948
1949*
1950
1951
1952
1953
1954*
1955
1956
1957
1958*
1959
1960*
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970*
1971
1972
1973
1974
1975*
1976
1977
1978
1979
1980*
1981
1982*
1983
1984
1985
1986
1987
Notes:
Source:
TOTAL
($ BILLIONS)
29.4
20.5
42.6
36.9
30.2
28.6
29.8
53.4
45.1
43.5
42.2
56.6
48.2
55.8
60.6
68.5
74-2
92.7
99.0
94.9
114.0
116.0
101.8
127.4
153.4
215.2
179.0
155.3
214.6
259.3
314.1
326.0
324.8
375.8
298.5
420.3
492.6
459.2
492.2
474-1
SOURCES
INTERNAL (%)
67.0
97.6
43.4
56.4
74.5
78.0
81.9
56.0
66.7
73.6
12.1
64.3
74.5
66.1
71.3
68.6
70.5
63.8
63.9
67.7
57.7
56.2
61.7
58.6
56.3
25.0
49.9
80.4
66.2
63.7
58.0
60.6
61.6
63.7
81.2
68.0
68.3
76.7
72.6
74.4
EXTERNAL (%)
33.3
2.4
56.3
43.6
25.8
21.7
18.1
43.8
33.5
26.4
27.5
35.7
25.7
33.9
28.7
31.5
29.5
36.2
36.1
32.3
42.4
43.8
38.3
41.4
43.7
56.4
47.9
19.6
33.8
36.3
42.0
39.4
38.4
36.3
18.8
32.0
31.7
23.3
27.4
25.6
TOTAL
($ BILLIONS)
25.6
18.4
40.4
37.9
30.0
28.5
28.1
49.1
41.1
40.0
38.5
52.1
41.8
50.7
56.2
60.3
64.9
83.4
92.0
87.6
106.2
115.0
97.9
121.8
145.1
189.7
191.1
153.4
210.4
242.2
324.7
368.1
342.1
383.6
303.5
385.8
502.7
435.3
454.3
436.6
USES
CAPITAL
EXPENDITURES (%)
80.9
81.0
59.4
80.7
84-7
91.9
82.9
66.2
90.5
89.3
72.2
72.9
90.4
72.0
77.9
73.8
77.2
73.9
81.8
81.3
71.0
72.4
80.9
69.9
65.5
62.7
72.5
73.2
74.6
74.2
66.8
64.7
71.2
74.7
84.5
70.2
73.7
78.6
73.0
82.7
FINANCIAL
ASSETS (%)
19.5
19.0
40.6
19.5
15.3
8.1
17.4
33.6
9.7
10.5
28.1
27.3
9.3
28.0
22.2
26.0
23.0
26.1
18.2
18.7
29.0
27.6
19.1
30.1
34.5
37.3
27.5
26.8
25.4
25.8
33.2
35.3
26.8
25.3
15.5
29.8
26.3
21.4
27.0
17.3
* Trough of the recession.
Data from the Board of Governors of the Federal Reserve System (1990)
345
PATRY & POITEV1N
of his effort, such as profits, sales, growth in sales or profits, output, market
share, etc. This raises the problem of moral hazard, which flows from a deficient
internal structure that fails to provide the manager with appropriate incentives
to maximize the value of the assets under his control. As a result, the manager
may pursue goals other than wealth-maximization, such as personal prestige or
the accumulation of perquisites.
ADVERSE SELECTION
RECALL THAT THE INHERENT PROFITABILITY of a project, or the quality of an
entrepreneur are not easily or directly observable. This is the source of an
adverse selection problem, which arises when the firm's decision-makers have
private information about the quality of different projects, their competitors,
the technology, or even about their own ability to manage the assets they control.
The manager's problem, then is to convince financial investors to provide the
necessary funding to undertake production. The financier's problem is to offer
a contract that will induce entrepreneurs to reveal their private information
and to select valuable projects on this basis.
IMPERFECT COMMITMENT
A PERFECT RISK-SHARING OR SMOOTHING ARRANGEMENT between an
entrepreneur and a financier requires the development of a long-term relation-
ship between the two parties. The dynamics of such a relationship bring about
some problems of their own. A perfect arrangement necessitates a strong
commitment by both parties to co-operate during all future periods. Such co-
operation requires that the firm reimburse financiers in good times and that
financiers provide the firm with the funding it needs in bad times.
But a long-term, dynamic contract is complex and cannot cover all
contingencies. Contracting parties are generally unable to describe explicitly
all future contingencies, or even to anticipate all possible events. For these
reasons, dynamic contracts are incomplete - and it is this incompleteness that
makes the governance of the contract an important factor in the success of the
relationship.
The incompleteness of contracts in dynamic relationships also generates
the need for punctual or selective interventions by the financier. For the same
reasons that contracts are incomplete, however, these interventions are far
from trivial. For example, the financier may want to intervene when the firm
is in genuine difficulty, or when the entrepreneur appears to be doing a poor
job. But again, lack of information may preclude the financier from drawing
the correct inference. Financial statements give at best a noisy and imperfect
picture of a firm's financial health and of managerial performance.
346
WHY INSTITUTIONAL INVESTORS ARE NOT BETTER SHAREHOLDERS
THE GOVERNANCE OF FINANCIAL CONTRACTS
IF THE FINANCIER CANNOT RELY SOLELY on financial statements, what other
tools may be used to draw the correct inference about managerial performance?
One possibility is to monitor the firm's performance closely. Monitoring
implies the gathering of information about the firm's external as well as its
internal economic environment. The financier should learn about the firm's
main competitors, the technology they use, the technologies that are available,
the characteristics of customers, the future prospects of the industry, the
research and development activities that are likely to affect the product market,
the internal structure of decision-making, and the quality and competence of
the firm's managers, among other things. Without precise information about
these elements, the financier can hardly evaluate managerial performance.
Monitoring activities have the following characteristics. First, they are
costly: to gain a good understanding of a firm's economic environment consumes
resources. (Note, however, that monitoring activities are subject to economies
of scale.) Second, the result of monitoring is noisy and almost never produces
a precise answer to the question: is the firm performing as it should? Finally,
the information gathered has a public-good aspect. Once produced, it is
difficult to keep other investors from taking advantage of it. The decision to
monitor or not depends on cost and expected benefits. Even when the costs of
monitoring are easy to determine, the benefits still depend on what gain is
expected from the information produced. There are two general mechanisms
(or strategies) at an investor's disposal to discipline a non-performing firm: the
investor can sell his/her shares (exit), or the investor can try to influence
management (voice). We turn to a discussion of these two mechanisms in the
next section.
Monitoring activities are further complicated by the "two-tier agency"
problem, another aspect of financial markets that has not been discussed so
far. In modern economies, financiers are often agents themselves, i.e., they act
as intermediaries between buyers and sellers of financial securities.
Institutional investors are a perfect example of such intermediaries. In a two-
tier agency structure, the incentive problems that exist between the interme-
diary and the firm's manager also exist between the primary financial investors
and the intermediary. The contract between the intermediary and the
investors should therefore provide the intermediary with the necessary incen-
tives to behave in the investors' interests. We therefore ask: "Who shall moni-
tor the monitor?".
Alternatively, a disgruntled financier could sell all shares in the
entrepreneur's project, thus exiting the market. In both cases, the financier
needs reliable information to determine a course of action.
347
PATRY & PO1TEVIN
CORPORATE GOVERNANCE: EXIT AND VOICE STRATEGIES
W
E HAVE SO FAR ARGUED THAT AGENCY PROBLEMS in financial markets are
the consequence of informational and commitment problems. The
incompleteness of contracts draws our attention to their governance. This
section surveys a number of corporate governance strategies.
EXIT MECHANISMS
AN EXIT STRATEGY CONSISTS OF SELLING THE SHARES of a non-performing firm
and investing the proceeds elsewhere. Financially exiting a firm is often
associated with having a short-run perspective, or behaving "myopically". We
point out that this is not necessarily the case. Suppose a monitoring investor
judges that a firm is not performing satisfactorily because of problems deriving
from moral hazard or adverse selection. One obvious way of transmitting this
information is to liquidate the firm's shares. This affects share prices and thereby
transmits some information to the firm's board as well as to other potential
investors.
Exit may therefore be a valid market response to agency problems if it
triggers a response from the board, from other investors, or from a raider, that
induces managers to return to a wealth-maximizing policy. However, this
mechanism is not perfect. First, while share-price movements may communicate
some information about a firm's under-performance, they do not inform other
investors and board members about the nature of an agency problem. Second,
share-price movements caused by factors unrelated to firm performance may
trigger a market response when none is desirable. Managers are frequently
heard to complain about the "short-run" bias of the market. Sometimes money
managers also voice the same complaint.
One way in which an exit strategy can produce a corrective response, if
adopted by a significant number of shareholders, is through a takeover: some
investors exit while others enter. The incoming investors acquire relevant
information about a firm's agency problems and decide to take it over. (They
must have some private information as to the cause of under-performance in
order to profit from their takeover.) If the incoming investors can induce
incumbent shareholders to tender their shares at a price that leaves with them
a portion of the gains from reorganizing the enterprise, they will proceed with
the reorganization.
This strategy is not always successful, however. For one thing, it is
plagued by a free-rider problem, as shown by Grossman & Hart (1980). Since
each individual shareholder has an interest in being the last to sell, the initia-
tor of a takeover may have difficulty convincing incumbent shareholders. This
results in the "raider" having to offer a high price to all the firm's shareholders,
thus reducing his own gain. The incentive to monitor and initiate takeovers is
thus reduced.
348
WHY INSTITUTIONAL INVESTOR S ARE NOT BETTER SHAREHOLDERS
According to this logic, there may be too few takeovers. But even when
they occur, they may not be efficient. As Shleifer & Summers (1988) argue, a
takeover may be profitable only because the raider breaks implicit contracts
between the firm and its partners - workers, long-term suppliers, etc. Since
implicit agreements may be ex ante efficient, takeovers may then have a lower
social value than expected, because they create a commitment problem.
The ultimate desirability of a takeover rests on its ability to solve the
moral hazard or adverse selection problems that are at the source of inefficien-
cy. In the case of moral hazard, success depends on the extent to which the
raider can restructure the incentives facing incumbent managers. Jensen
(1988, 1989) argues that many management buyouts in the United States did
just that by increasing the participation of management teams in equity. The
same results can be obtained by reorganizing a company, by having it focus on
its core business, and by divesting it of the units its management team cannot
run competitively, thus changing the incentive structure facing managers.
Bhagat, Shleifer & Vishny (1990) and Patry & Poitevin (1991) found evidence
in both the United States and Canada that is consistent with that hypothesis.
Conversely, a raider can alleviate adverse selection problems by replacing the
management team. Notice, however, that in any case, a successful takeover
requires an extensive monitoring operation.
Exiting customers are another indicator of a firm's poor performance.
Although a declining market share is a reliable signal of poor performance, it
does not provide information about the nature of the agency costs.
Nevertheless, financiers often base their financing decisions on product market
data.
The market for managers is more closely related to our analysis. An
active market for managers provides incentives for managers to perform
adequately in the expectation that they may raise their market value.
However, we believe that this market does not work as efficiently as may be
thought. It suffers from the same deficiencies as financial markets. If financiers
must incur great costs to evaluate a firm's performance, it is likely that participants
in the market for managers will have the same problem. Even though they
constitute another potential group of monitors of the firm's performance, they
are not likely to invest as much in monitoring activities as financiers since
they are not likely to have as much at stake as financial investors.
VOICE MECHANISMS
VOICE MECHANISMS ACTIVELY SEEK TO INFLUENCE the actions of firms. Voice
mechanisms can be informal, as when money managers discuss corporate strategy
with a management team, or they can be very formal, as in a proxy fight.
Voice can be exercised through an active presence on the board of
directors. Directors have access to privileged information about the firm and
its economic environment and are required to take an active part in the firm's
349
PATRY & POITEVIN
strategic planning. Through its compensation committee, the board is also
responsible for structuring adequate compensation packages for corporate
managers. The board is also a key instrument in solving moral hazard problems.
A presence on the board therefore reduces monitoring costs by providing
access to inside information.
On the negative side, board members are frequently "captured" by
management. For example, certain directors may lean toward management's
point of view more often than they should. This is particularly likely in
Canada where ownership is fairly concentrated, the supply of potential directors
relatively small, and the number of interlocking directorships important.
Institutional investors also tend to be afraid that rules concerning insider trading
and conflict of interest will limit their investment options. We return to this
point later in our study.
A proxy fight is a more spectacular voice mechanism. The initiator of a
proxy fight must gather sufficient votes to ensure his point gains a majority. If
successful, a proxy fight can result in a significant change in the way in which
a firm is run. A proxy fight is, however, a strenuous voice mechanism in which
the investor begins at a clear disadvantage compared with the manager(s).
The logistics of proxy fights are weighted in favour of management, which
suggests that some reforms might be helpful here.
Although the benefits of voice mechanisms are relatively well under-
stood, their costs are often understated. Both an active presence on the board,
and a proxy fight can be fairly costly, in addition to the monitoring costs that
we have already discussed. Indeed, some proxy fights have been shown to cost
millions of dollars. Furthermore, an important element of voice mechanisms is
that they have a strong public-good characteristic in that all shareholders
benefit without having to pay any of the cost. This may be why we do not
observe enough "voice". Another implicit cost related to proxy fights is the
breach of trust that they may cause. For example, management may be reluctant
to lay off a group of workers toward which it considers it has a responsibility
based on a long-term relationship. A proxy fight can avoid this problem with-
out implicating management. The cost of doing so is that it may be more difficult
in the future to build the necessary trust for long-term relationships. A similar
argument has been used in our discussion of takeovers.
Finally, we note another voice mechanism on the part of creditors which
is often overlooked in the literature. Debt covenants are certainly part of the
governance package which disciplines managers by placing limits on managerial
discretion over sets of strategic decisions. These are useful to contain moral
hazard problems, but are not especially helpful in cases of adverse selection
where managerial quality is the real problem.
350
WHY INSTITUTIONAL INVESTORS ARE NOT BETTER SHAREHOLDERS
INSTITUTIONAL INVESTORS IN CANADA
C
ORPORATE AND PUBLIC PENSION FUNDS, mutual funds, banks and near
banks, insurance companies, and public and private endowments are
institutional investors. In the United States, at the end of 1990, these
investors held over US$ 6 trillion, or 45 percent, of all financial assets.
Comparing this figure with their 21 percent share in 1950 illustrates their
indisputable rise. In the equity market, which is of particular importance to us,
the share held by institutional investors rose from 23 percent in 1955 to 53
percent in 1992. American institutional investors own over US$ 1 trillion of
the equity in the United States and 90 percent of this is held by corporate,
public, and union pension funds (Lakonishok et al., 1992). This means that
pension funds control 47 percent of all U.S. equity and explains why pension
funds have been at the forefront of the movement toward institutional
activism.
The position of Canadian institutional investors is similar. For instance,
at the end of 1993, the book value of financial assets under the control of
trustee-administered pension funds, mutual funds, insurance companies, banks
and near banks had swollen to $1,283 billion, of which $142 billion was equity
(Table 4). Canadian public and private pension funds together controlled
book value assets of nearly $250 billion and book value stocks of $70 billion.
Note that in Canada the share of institutional investors in total equity is only
35 percent, compared to 53 percent in the United States. Nevertheless, these
institutional holdings are substantial.
This last figure appears more impressive, in fact, when the relative lack
of liquidity of the Canadian equity market is taken into account. Fowler &
Rorke (1988) estimate that a mere 5.3 percent of the stocks traded on the
Toronto Stock Exchange can be said to be widely traded. Most stocks (59.4
percent) on the TSE are traded in thin markets, which increases the price of
exit. Daniels & Waitzer (1994, p. 33), quoting The Globe and Mail, estimate
that 50 percent to 60 percent of the shares of widely held companies that are
traded in deep markets are held by institutional investors. Here, the size and
power of institutional investors in the United States as well as in Canada have
increased substantially over the past 30 years.
In the United States this increase has been accompanied by a call for a
more active role for institutional investors in corporate governance issues. As
in the United States, the growing importance of institutional investors in the
equity market has been accompanied in Canada by an awareness of corporate
governance problems and the potential benefits shareholders could derive
from a louder institutional "voice". The wave of mergers and takeovers that
swept Canada in the 1980s, and the ensuing attempts of corporate managers to
adopt poison pill strategies, prompted an interest in corporate governance and
in the protection of minority shareholders. Examination of the TSE Report on
Corporate Governance, of the Corporate Governance Standards of the Pension
35?
PATRY & PO1TEVIN
Investment Association of Canada and of the Corporate Governance Guidelines
of some major institutions, like the Caisse de depot et placement du Quebec,
shows that the issues of managerial compensation and the independence and
effectiveness of the board are receiving considerable attention. Fairvest Inc.
has developed a score card that permits institutional investors to assess more
precisely the corporate governance quality of Canadian corporations.
Nonetheless, few poison pill battles have been successfully fought by
Canadian institutional investors. A remarkable exception is the battle over
Labatt's poison pill. Generally, however, very little Canadian research has
been done on compensation packages; even if the Caisse de depot et place-
ment has at times been considered aggressive in its defense of minority share-
holders (as in the battle over the adoption of a poison pill at Inco), no major
institution in Canada is seen as the CalPERS of the North. As Daniels &
Waitzer (1994) put it: "In many ways, institutional activism in Canada is still
very much in a nascent stage." (p. 33). Should we ask them to do more?
The Berle & Means (1932) model of the corporation, in which control
and ownership are divorced, has led many to believe that managers now
effectively control America's corporations. Since someone must watch the
managers, and since individual shareholders do not have the incentive to do the
watching, then why not have the institutional shareholders do it? This is the
claim we critically assess in this study.
The split between ownership and control in the large American and
Canadian corporations is not necessarily a reality. For instance, Demsetz &
Lehn (1985) compute the total percentage of stock controlled by the five most
important shareholders of the Fortune 500 corporations in the United States.
This amounts, on average, to 25 percent (compared to 33 percent in Japan
[Prowse, 1991]). Whether such an ownership structure induces efficient over-
sight is one thing; arguing that no individual or coalition of shareholders has a
focused interest in monitoring the managers is quite another. It does not ring
true.
In Canada, any split between ownership and control is the exception
rather than the rule because Canadian corporations are more tightly controlled
than their American counterparts. Table 3 shows that, in 1994, 60 percent of
the Financial Post 500 largest Canadian non-financial corporations were wholly
owned or effectively controlled by a single shareholder. Only 79 of the 500
companies in the Financial Post 500 list, a mere 16 percent, could be consid-
ered widely held. The equivalent percentage for the Fortune 500 firms is esti-
mated at 63 percent (Daniels & Macintosh, 1991)
The importance of family-controlled businesses in Canada and of wholly
foreign-owned Canadian subsidiaries explains this characteristic of the
Canadian corporate landscape. Another explanation is the relative size of
Canadian firms. The median number of employees of the top 200 industrial
firms in Canada was 4,938 in 1993, compared to 10,136 for the Fortune 500
firms in the United States.
2
 As a firm grows in size, concentration of owner-
352
WHY INSTITUTIONAL INVESTORS ARE NOT BETTER SHAREHOLDERS
TABLE 3
OWNERSHIP CONCENTRATION (BY SALES OR REVENUE) OF THE FINANCIAL POST 500
PUBLIC CORPORATIONS
NUMBER PERCENT
Wholly owned or 50%+
by one shareholder (or government) 93 36.6
Two largest owners together > 34% 60 23.6
One owner > 15% or two largest
owners together own 20% < X < 34 36 14-2
Largest owner between 10% and 15%
but second < 10% 11 4.3
Widely held (including cooperatives
and memberships) 54 21.3
Total 254 100.0
ALL
NUMBER PERCENT
300 60.6
67 13.5
36 7.3
13 2.6
79 16.0
495 100.0
Source: The Financial Post 500, 1994.
ship results in increased firm-specific risk for the investor (Demsetz, 1993).
Hence, there is a tradeoff between the social benefit of concentrated owner-
ship (which reduces the incidence of moral hazard) and the cost of bearing
firm-specific risks.
Whatever the reasons, the fact is that ownership in Canadian corporations
is much more concentrated than it is in U.S. corporations. Combined with the
smaller percentage of equity in the hands of Canadian institutions (recall that
the percentage of total equity owned by institutional investors in the United
States is roughly twice that in Canada) this translates into a reduced potential
for Canadian institutional investors.
Another singular aspect of the Canadian corporate landscape which has a
bearing on the potential role of institutional investors is the high level of cross-
ownership. Daniels & Macintosh (1991) contend that of"... the top one hundred
most profitable companies in Canada in 1987, close to 45 percent held 10 per-
cent or more of the voting shares of another company on the list" (p. 888).
Berkowitz, Kotowitz & Waverman (1977) find that the 361 largest enterprises in
Canada had stakes in 4,944 other firms. Through cross-ownership and inter-
linked directorships, a few large groups have spun an intricate web of relationships
that allows these groups to exercise extensive control over the largest enterprises
in Canada. Compared to the Japanese keiretsu system of cross-ownership, however,
Canadian market ownership appears very diffused. Viner (1988) estimates that
65 percent of the stock of the companies listed on the Tokyo stock exchange is
held by keiretsu members (cited in Coffee, 1991, p. 1,296).
353
PATRY & POITEVIN
This networking effect has many implications for the potential role of
institutional investors. First, it is a source of the relative lack of liquidity in the
equity market in Canada. This raises the cost of exit and might lead one to
think that, as in Japan and Germany where liquidity is lower and intercon-
nectedness higher than in the United States, Canadian institutional investors
would be induced to play a more active role. Second, interlocking director-
ships create circumstances in which managers can assist each other to
entrench themselves further, which makes disciplining corporate managers
much more difficult since no one wants to "rock the boat". Third, the extensive
control and power wielded by a few large groups or families increases the
severity of the penalty that a disgruntled management could impose on an
unsettling institutional investor. (This last point is discussed more fully later
in our study.)
THE GOVERNANCE O F MUTUAL FUNDS
AND PENSION FUNDS
B
ANKS, MUTUAL FUNDS, CORPORATE AND PUBLIC PENSION FUNDS all tend to
be large institutions with diffuse ownership. They should not be expected
to behave like individual investors. Their internal structures and governance
must be examined to understand the extent to which they might be expected
to become involved in the governance activities of the firms in which they
invest. If the managers of an institution are themselves engaging in wealth-
reducing activities, or cannot motivate their employees and those (money
managers) to whom they delegate the management of funds to monitor and
discipline corporate managers, there is a moral hazard problem at the institu-
tional investor level. In addition, if institutional investors cannot discriminate
between good and bad money managers, there is an adverse selection problem. In
other words, agency problems cannot be solved simply by allocating the respon-
sibility for active monitoring of corporate managers to a class of agents (financial
intermediaries) since they are themselves plagued with information problems.
Hence, the classic question "Who shall monitor the monitor?" resurfaces.
We concentrate on two types of institutional investors in Canada -
mutual funds and pension funds. Together, they own almost 25 percent of all
equity in Canada (Table 4) and 69 percent of all the equity controlled by
institutional investors. What incentives do they have to monitor and discipline
corporate managers? How are they organized? Who does the monitoring?
THE GOVERNANCE OF MUTUAL FUND S
SINCE MUTUAL FUNDS POOL FINANCIAL RESOURCES from hundreds of thousands
of individual sources of capital, controlling fund managers poses many serious
problems that are analogous to those of controlling corporate managers. First,
354
WHY INSTITUTIONAL INVESTORS ARE NOT BETTER SHAREHOLDERS
TABLE 4
INSTITUTIONAL FINANCIAL ASSETS AND CANADIAN SHARES
DECEMBER 31, 1993 ($ MILLIONS)
TYPE OF INSTITUTION CANADIAN STOCKS
(BOOK VALUE)
FINANCIAL ASSETS
(BOOK VALUE)
ASSETS OF TOP TEN
(MARKET VALUE)
Pension Funds
a
 68,864 249,542
Public 39,872
b
 160,132
Private 2 8,992
b
 101,226
Mutual Funds
0
 28,899 111,192
Insurance Companies
0
 21,430 158,219
Banks and Near-Banks
0
'
d
 12,731 722,582
Caisse de Depot et placement
6
 9,890 41,685
Institution-owned 141,814
Total Shares Owned 406,216 >
116,988
f
lll,487
f
36,411
79,5088
147,388
h
Notes:
a
 Quarterly Estimates for Trusteed Pension Funds, Statistics Canada cat. 74-001 and
Benefits Canada, April 1994 for Top 10 concentration.
" Canadian Stocks of all pension funds distributed at pro ram of total stock holdings.
c
 System o/National Accounts, Statistics Canada cat. 13-214 and CANSIM Database.
Includes Trust companies and Mortgage Loans.
e
 Statistiques Financieres 1993, Caisse de depot et placement du Quebec.
Excludes pension funds under control of the Caisse de depot.
8 Benefits Canada, April 1994.
h
 The Financial Post 500, 1994.
1
 To avoid double-counting, we have subtracted from 517,408 the shares issued by mutual funds.
there is the collective-action problem of monitoring the management: no single
investor has the incentive to incur the full cost of assessing the quality of the
job done by the specialists to whom the responsibility of managing the funds
was delegated, while sharing the benefits of a better performance.
Second, what motivates the individual investor to delegate responsibility
(the root of the agency problem at hand) is the fact that the agent-manager
has (presumably) a superior ability to choose stocks. Therefore, if individual
investors are to second-guess the experts and do the analytical work, they
might as well do the investing themselves.
Third, the agency relationship gives rise to adverse selection and moral
hazard complications. On the one hand, adverse selection shields poor
management. On the other hand, moral hazard implies that money managers
may increase the risk of their portfolio(s) in order to boost returns to the
marginal investors (the so-called "bait and switch" tactic). The seriousness of
this (moral hazard) problem depends on two factors: the relative ease with
which investors can determine risk levels, and the predisposition of money
managers not to engage in such behaviour (integrity).
355
PATRY & POITEVIN
As a result of these informational frictions, the problem of motivating
the agent-manager for the individual investor is solved by observing the
performance of the agent. Since it is difficult to separate good performance
from good luck, an individual investor takes two steps: he chooses a fund based
on its medium- to long-run performance (thus smoothing out streaks of good
or bad luck) and he compares the performance of this fund to those of the
other funds available on the market, presumably controlling for risk level.
Using inter-temporal and inter-agent comparisons to assess the relative perfor-
mance of an agent can be very efficient, particularly when there are many
agents (Sappington, 1991).
Most Canadian mutual funds are open-ended, in which individual
investors hold a pro rata ownership of the pooled resources. An individual
investor in an open-ended mutual fund can insist at any time that the fund
buy back his/her share in the fund at its net asset value. (No such obligation
binds the closed-end mutual fund). As a result, a dissatisfied investor in an
open-ended fund is likely to sell his/her shares. Even if the costs of exit are not
nil (because of fees and search costs), exit is less costly than the influence
costs the investor would have to bear in order to bring about a change in the
fund's policy. When many investors flee, the managers of the fund must neces-
sarily sell the stocks owned by the investors. Liquidity is therefore central to
both the investors and the funds. Given the large number of competing funds
in Canada, exit mechanisms should function well.
3
Other mechanisms also constrain managers of mutual funds. Regulations
with respect to information provided to investors, the assets the fund can
invest in, and the role, composition, and fiduciary duties of the board also
limit mutual fund managers. Conversely, the market for corporate control is
ineffective and so certain forms of compensation for managers are also regulated.
THE GOVERNANCE OF PENSION FUNDS
PENSION FUNDS ARE STRIKINGLY DIFFERENT FROM mutual funds in one crucial
respect. Workers have very little control over the financial resources they invest
in pension funds. The main instrument of control is indirect and relies on the
monitoring capability of the pension funds committee. This voice mechanism
works in a very imperfect manner. In the first place, workers should delegate
representatives, thus creating an agency relationship, to monitor the allocation of
funds. Finally, individual workers are plagued by a collective action problem and,
as a result, the disciplining potential of workers is seriously impaired.
The individuals who invest in pension funds are in a much more vulnerable
position than those investing in mutual funds. Their voice is muted, their influence
is weakened by a collective action problem, and exit is costly, if not impossible.
Under defined-benefits pension plans, which represent 90 percent
4
 of
Canadian pension plans (the remaining 10 percent are defined-contribution
plans), the sponsor - a private corporation or the government - promises
356
WHY INSTITUTIONAL INVESTORS ARE NOT BETTER SHAREHOLDERS
employees a stream of future benefits that do not depend on the fund's
performance. Furthermore, the sponsor must compensate any shortfall, and
this becomes the first claim on the corporation in the case of bankruptcy.
(However, the sponsor is also entitled to any surplus above the actuary-
calculated liabilities of the fund.) There is therefore a possibility that the
sponsoring organization may have an intrinsic interest in maximizing the
return on the funds and therefore in taking whatever action, including over-
seeing corporate managers, that fosters that goal.
Unfortunately, the fact that the sponsor is the residual claimant for
defined-benefits plans does not, in itself, guarantee that the sponsor will aim
at maximizing the risk-adjusted return on the assets, nor does it signify that
the sponsor will monitor corporate managers. For one thing, the property
rights on the surplus are not well-defined, and so workers argue that in some
circumstances they should share in the spoils. Also, in the private sector, a
large surplus may increase the likelihood of a takeover - a prospect incumbent
managers do not relish. In addition, the delegation of the fund's management
to the treasurer's office and to external money managers is likely to give rise to
other moral hazard, adverse selection, and commitment problems (see next
section). Finally, Canadian corporate pension fund managers often find them-
selves embroiled in serious conflicts of interest that pit their fiduciary duty
against the business interests of the corporation that employs them.
SOME EVIDENCE ON PERFORMANCE
Do MONEY MANAGERS PURSUE WEALTH MAXIMIZATION? There is a growing body
of evidence that returns on the assets in the corporate pension fund segment
of the money management industry fall below those of a market index and of
mutual funds. This may signify that managers of pension funds manage their
equity more conservatively than the average stock market investor. This is a
testable empirical proposition. However, pension funds can modify their overall
risk by changing their asset mix. Conversely, poor performance may simply
mean that corporations have more complicated objectives than maximizing
the funds' risk-adjusted return, or that they botch the job, or both!
Lakonishok, Shleifer & Vishny (1992) analyze the performance of 769
all-equity corporate pension funds in the United States and find that, on average,
the representative (equity) fund under-performs the Standard 6k Poors 500
Index by 1.3 percent per year, before subtracting management fees. This is
consistent with the findings of Beebower & Bergstrom (1977) for the 1966-1975
period; with those of Brinson, Hood 6k Beebower (1986) for the 1974-1983
period; and with the results reported in Malkiel (1990) for the 1975-1989
period. Similar results for Canada are reported by Taylor (1995) who concludes
that ". . . for periods ranging from one to eight years, the annualized return of
the index has outperformed the median manager by between 0.10% and
3.10%, before management fees" (p. 25).
357
PATRY & POITEVIN
The evidence relating to the performance of mutual funds is much less
conclusive. While the classic study by Jensen (1968) and recent refinements
by Grinblatt & Titman (1989) and Connor & Korajczyk (1991) cast some
doubt on the capability of mutual funds to outperform the market, some recent
evidence suggests that this might be the case. Ippolito (1989) finds that mutual
funds outperform the market by 80 basis points, once management fees and
load charges have been netted. Results similar to Ippolito's for Canada are
obtained by Berkowitz 6k Kotowicz (1993) who conclude that the "... funds in
our sample [roughly 40 mutual funds] have outperformed the market by a very
significant margin" (p. 863).
The evidence thus suggests that corporate pension funds generate
systematically lower returns. Pensions and Investments (cited in Lakonishok et
al., 1992) indicates that mutual funds had a mean return 2.3 percent higher
than that of pension funds over the 1971-1980 period — lower than those of
mutual funds, but (more damningly) lower than the return on an index of the
market. By adopting a passive investment strategy (indexing), pension funds
can earn a higher return on their equity portfolio and cut costs, since typical
fees for indexed funds in Canada average 43 basis points below those for
actively managed funds (Taylor, 1995, p. 26).
What about the performance of public pension funds? These are the
largest of all the institutional players, and they are the ones who appear to
have been most active in the area of corporate governance. Public pension
funds manage nearly 8 percent of all Canadian equity. Each of the two largest
public pension funds, the Ontario Teachers' Pension Fund Board and the
Quebec Public Employees Fund, manages as many assets as the ten largest
private funds.
5
Reviewing the evidence on the performance of public pension funds
in the United States, Mitchell & Hsin (1994) conclude that "... yields on
public pension fund assets have frequently been low, with public plans earning
rates of return substantially below those of other pooled funds and often
below leading market indices" (p. 3). Interestingly, their data (Table 3, p.
28) show that in 10 of the 17 years for which comparisons can be made,
large Canadian private pension funds outperformed U.S. state and local
pension plans. The authors suggest that one possible cause for this under-
performance is the responsiveness of the staff managing public pension
funds to political pressures. Romano (1993) also concludes her analysis of
public fund activism in the United States by warning that "... public pension
funds face distinctive investment conflicts that limit the benefits of their
activism. Public fund managers must navigate carefully around the shoals of
considerable political pressure ... ." (p. 795). Romano also finds that
"[f]unds with more politicized board structures perform significantly more
poorly than those with [a] more independent board" (p. 852).
358
WHY INSTITUTIONAL INVESTORS ARE NOT BETTER SHAREHOLDERS
THE INTERNAL ORGANIZATION OF DECISION-MAKING:
MUTUAL FUNDS
WE NOW TURN OUR ATTENTION TO THE ORGANIZATION of mutual and pension
funds. These institutional investors entrust money managers with the respon-
sibility of choosing stocks. Two factors - the nature of the contract offered to
money managers and the method by which pension funds select money
managers - have an effect on the level of monitoring and influence exerted by
institutional investors.
Generally, contracts between shareholders and money managers are
regulated. In Ontario, for example, mutual funds that want to offer their
money managers performance-fee contracts must obtain the approval of the
Ontario Securities Commission. Performance-fee contracts link a money
manager's compensation to the return on the portfolio managed. The OSC
checks, among other things, that the benchmark defined in the contract is
deemed "satisfactory".
Such outcome-based contracts have many advantages. For one thing,
they mitigate moral hazard problems by requiring that the money manager
bear a percentage of the cost of any relaxation of his/her effort. Another
advantage of these contracts is the possibility of signalling. When managers
have an informational advantage and are allowed to offer various contracts,
good money managers can signal their superiority by selecting incentive
contracts that shift more risk and profit to them.
Mutual funds might therefore be expected to rely on performance-fee
contracts inasmuch as regulatory authorities permit. But this expectation is
contradicted by the evidence in both Canada and the United States, where
similar constraints apply. Indeed, the vast majority of mutual fund money
managers are compensated on an asset-based scheme. Discussions with funds
managers suggest that less than 10 percent of assets in Canada are managed
under performance-fee contracts. The typical scheme provides for a fee equal
to a declining percentage of the funds managed. Even in the more permissive
context of the United States, Bailey (1990) mentions the lack of interest in
performance-fee contracts.
This finding is somewhat puzzling. A possible explanation is offered by
Berkowitz & Kotowicz (1993) for mutual funds. They suggest that asset-based
contracts create an indirect link between compensation and performance.
They assume that "investors believe superior management is possible and
better performing funds can be identified on the basis of their past perfor-
mance". It follows that the demand for better-performing funds will increase,
thus endowing more capital to superior managers. Given that performance is
difficult to observe because of the noisiness of the market, some inertia will
characterize the evolution of market shares. In this scenario, sustained superior
performance is rewarded with an expanded asset base, and higher compen-
sation. Their empirical results support their thesis about an indirect link
359
PATRY & PO1TEVIN
between performance and compensation. Hendricks, Patel & Zeckhauser
(1993) also find evidence supporting the (short-run) persistence of superior
performance.
6
Given that money managers compete in a world resembling that
described by Berkowitz & Kotowicz (1993), their investment behaviour will
critically depend on the horizon they expect investors will use to allocate capital
among them. If this horizon is quite short (a few quarters or at most a few
years) money managers should be expected to focus on short-term equity
returns. Holding onto stocks that offer a high potential over the long run, if it
does not translate into short-term returns, is self-defeating. If, on the other
hand, investors are characterized by a long-term horizon, money managers will
be induced to adopt the same (long-term) horizon.
The shorter the horizon, the more liquid money managers will want to
be. They will also be less inclined to spend time and energy to influence
corporate managers. Coffee (1991) contends that "most mutual funds are
active traders and would hesitate to make any investment the liquidation of
which would require a significant block discount" (p. 1318).
To conclude on mutual funds' incentive structure, it appears that money
managers are compensated mostly on the basis of the assets they manage. This
contract structure can, if investors use past performance as an indicator of
future performance, create a link between effort and compensation, thus
alleviating the moral hazard problem. On the other hand, the uniformity of
contracts in the market impedes the signalling strategy of the best money
managers; and, competition for a share of investable funds depends critically
on the effective horizon characterizing investors and mutual funds managers.
THE INTERNAL ORGANIZATION OF DECISION-MAKING:
PENSION FUNDS
OUR DISCUSSION OF THE INTERNAL ORGANIZATION of pension funds draws
heavily on the analysis of Lakonishok, Shleifer & Vishny (1992). Three
important corporate decisions must be made. They concern the use of internal
versus external management of the funds, the portion of the funds to be
invested in indexed funds, and the selection of money managers. What
governs these decisions, and what are the incidences on the behaviour of the
agents doing the investing, the monitoring and, at times, the influencing of
corporate managers?
Table 5 shows that in 1993 close to 60 percent of the top 100 pension
funds in Canada were managed in-house. Of the 40 percent that were managed
externally, 14 percent were by balanced funds managers and 26 percent by
specialized money managers. We can also see that public pension funds rely
more heavily on internal management and specialized money managers. This
appears to be consistent with the perception that public pension funds have
been more active.
360
WHY INSTITUTIONAL INVESTORS ARE NOT BETTER SHAREHOLDERS
TABLE 5
ASSET MANAGEMENT BY TOP 100 CANADIAN PENSION FUNDS (%)
1990
a
PRIVATE & PUBLIC
In-house Management
Balanced Fund Managers
Specialized Managers
Total
Note:
 a
 Top 80 funds only.
Source: Benefits Canada Survey 1991
50.1
27.4
22.5
100.0
, 1994.
1993
PUBLIC PRIVATE
73.3 31.1
7.7 28.7
19.0 40.2
100.0 100.0
PRIVATE & PUBLIC
59.7
14.5
25.8
100.0
Lakonishok et al. (1991) claim that treasurers can be expected to be
biased against both internal management and indexing. The rationale is that
treasurers are torn between two goals: they want to increase the resources
under their control but at the same time want to bear as little risk as possible.
They want to take charge but do not want to be held responsible for the bad
returns on the assets with which they are entrusted. The solution, according to
Lakonishok et al., is to allocate as many funds as possible externally, but to adopt
an active management policy, which calls for constant review, evaluation and
selection of the money managers they hire. Choosing money managers and
allocating the available funds between them becomes the treasurer's most
important job.
As is the case for mutual funds, pension funds compensate money managers
on an asset basis. An increase in a money manager's wealth is directly related to
the extent of the increase in the size of his asset base. To make his case and
convince treasurers to allocate a greater share of the funds to him, he will claim
that he can outperform other money managers as well as the market. (However,
the claim that money managers, on average, can beat the market is very
debatable.) Now the money manager's track record plays a determinant role.
How will treasurers evaluate that track record? What horizon will they favour?
Much evidence, both anecdotal and empirical, points in the direction of
a very short-term bias. Coffee (1991, p. 1,325) reports on a recent survey
indicating that at least 90 percent of pension funds reviewed the performance
of the outside money managers they employed at least quarterly, with a mere
3 percent doing annual reviews! The problem with a short-term horizon is
that performance is very elusive. Although Lakonishok et al. (1992) find some
long-term consistency, they see very little short-term (annual or quarterly)
consistency. Bauman & Miller (1994) mention several recent analyses that
demonstrate the instability of performance rankings over time.
Failure to recognize the elusiveness of performance has three pernicious
effects. Consider first the adverse selection bias it insinuates in the market for
361
PATRY & POITEVIN
money managers. Many funds that use recent performance to select the "best"
money managers are led to "hire high, fire low". They drop money managers
who are about to recover and hire those who are bound to fall from their
pedestal. The results of Lakonishok et al. (1992) indicate that this might well
be the case, at least in the United States . The authors even speculate that the
"strategy of switching to the good past performers may not be a bad one" (p. 372).
Hence, a treasurer using the wrong time horizon will systematically pick
money managers who will underperform.
Second, money managers know that performance is elusive, and so they
develop strategies to differentiate their products and justify their short-term
performance. That way, they can always come up with a good story to explain
any short-term disappointing performance. This leads to window-dressing by
money managers. Lakonishok, Shleifer, Thaler & Vishny (1991) and Benartzi
& Thaler (1992) find evidence that money managers get rid of badly performing
stocks at the end of the year and buy into trendy stocks, even if it means
sacrificing performance, in order to be in a better position to impress sponsors.
Mitchell & Hsin (1994, p. 5) also report that money managers who are subject
to frequent performance reviews may sacrifice long-term goals for short-term
objectives.
The third consequence of a failure to look ahead is a high(er) turnover
rate of funds. The elusiveness of performance induces sponsors to reallocate
funds among money managers too often and encourages money managers to
trade too much. Regarding the last point, Lakonishok et al. (1992) conduct an
interesting experiment: they freeze the holdings of money managers for six or
twelve months and compare the returns earned by the frozen portfolios to the
actual returns. This leads them to conclude that "trades made by the funds
were counterproductive, costing an average forty-two basis points relative to a
portfolio frozen for six months and seventy-eight basis points relative to a
portfolio frozen for twelve months" (p. 354).
As a result of the selection and evaluation process developed by pension
fund managers, money managers are pressured to distort their investment
strategies, to trade too actively, and to trade stocks that they should not trade
in order to provide their sponsors with "schmoozing" — good stories that help
explain poor performance. This means either that sponsors are behaving
myopically and do not know that quarterly returns are very noisy or that they
know but prefer to accept lower returns and schmoozing to straight high(er)
returns.
Lakonishok et al. (1992) claim that the second argument is true. They
point out that if pension fund managers wanted straight performance they
would invest heavily in indexed funds, which have outperformed the equity
component of pension funds for many years. Instead, claim the authors,
corporate treasurers prefer to diversify across money managers, thus achieving
some averaging effect. From the vantage point of the treasurer, the problem
with an indexing strategy is that it considerably reduces the work of the treasurer.
362
WHY INSTITUTIONAL INVESTORS ARE NOT BETTER SHAREHOLDERS
As with mutual funds, pension funds delegate investment decisions to
money managers. Like mutual funds, they split investable funds between
active management and passive management, in the form of indexed funds.
Although pension funds involve much lower liquidity constraints than
mutual funds (recent work by Berkowitz & Logue, 1987, finds the average
annual turnover rates of equity for pension funds to be 61 percent compared
to 76 percent for mutual funds), we find reason to doubt that money managers
are chosen on the basis of long-term performance. As Lakonishok et al. (1992)
put it, "... sponsors clearly reallocate funds in response to past performance,
and because consistent performance is fairly elusive there is tremendous
turnover at the top in terms of industry leadership and market share" (p. 364).
Finally, since workers can only "vote with their feet", and even then only to a
limited extent, the relatively poor performance of pension funds can continue
(despite poor equity management) without fear that exiting capital will trigger
a reaction.
Having argued that institutional investors themselves suffer from internal
agency problems, we now move on to discuss how these problems interact
with corporate governance activism.
THE ROLE OF INSTITUTIONAL INVESTORS IN
RESOLVING MARKET IMPERFECTIONS
A LL MONEY MANAGERS HAVE AN INCENTIVE to monitor the stocks they manage
jtVand the companies that issue them. Monitoring results in one of three
possible courses of action: more monitoring, dumping the stock (exit), or an
attempt to influence corporate management (voice). The first two options are
clear-cut. The third, that of exerting influence, covers a wide range of possible
action including: informal discussions with managers; seeking the support of
other institutional investors or of a large shareholder; sponsoring a shareholder
resolution; or engaging in a proxy fight (Pozen, 1994; Rock, 1991, p. 453); or
becoming a relational investor (one who holds a large stake in the company,
has representation on the board, and commits for the long term) (Ayres &
Cramton, 1994).
THE COST-BENEFIT ANALYSIS FOR THE SPONSOR
ASSUME THAT INSTITUTIONAL INVESTORS MAXIMIZE risk-adjusted returns.
Under this maintained hypothesis, they will decide what action to take based
on the highest net present benefit-to-cost ratio. For instance, even if some
forms of activism appear to be profitable, the benefit-to-cost ratio of, say,
tendering shares in a takeover may be even greater. Since the costs and benefits
of each course of action vary considerably and are contingent on the business
context, we can make only fairly general and speculative comments here.
363
PATRY & POITEVIN
It is well recognized that, by influencing managerial policy, a shareholder
produces a public good: all other shareholders benefit from the disciplining
action, regardless of their participation in the provision of the activity. This is
the source of a fundamental free-rider problem: apathy may be rational for a
shareholder who does not expect his contribution to be pivotal.
Both the likelihood that a shareholder's contribution is pivotal and the
share of the public good one shareholder can appropriate are proportional to
his stake in the equity of the corporation. Thus, increasing the concentration of
share ownership by encouraging or allowing institutional shareholders to hold
larger stakes should induce them to adopt a more active strategy.
Concentration also increases the cost of exit since the sale of large blocks of
shares is bound to have a depressive effect on the stock price.
Alternatively, free-riding is a rational strategy for a small institutional
shareholder and in circumstances where one shareholder or a small group of
shareholders is in a position of control. This is typical in Canada.
For institutional shareholders who are not rationally apathetic, the decision
to influence management and the situations which provoke them to do so
depend on the cost of organizing a disciplinary move as well as on the cost of
possible retaliation. These costs, in turn, depend on both the nature and the
subject of the intervention.
Process and procedural reforms — such as regulating the size and compo-
sition of boards (by having a majority of outside directors on the board, ensuring
that compensation and other key committees are staffed by outside directors,
etc.), limiting the tenure of directors and the number of boards that a director
can serve on, and separating the functions of chairman and chief executive
officer, etc. — impl y lower search and monitoring costs than most other
interventions, such as replacing the management team or influencing corporate
strategy. Furthermore, large institutional investors with stakes in many widely
held corporations can reap economies of scale by using the information they
produce to influence the incentive structure of many enterprises. Finally, an
important institutional investor can reasonably share its expertise with other
large institutional investors to obtain their support. The organization of the
pension fund industry in Canada (which is concentrated around the top 20
public and private investors) is conducive to the building of such coalitions.
Much the same reasoning applies to informal initiatives, such as writing
explanatory letters to the company's CEO or directors, or holding discussions
with the firm's management.
There are four reasons why this mode of intervention is less likely to be
fruitful when firm-specific issues are at issue. First, micro-management requires
a level of effort and a depth of analysis that are lacking in most pension funds
or mutual funds. Without such detailed knowledge, an institutional investor
cannot determine what action it should take. Before endowing pension funds
with the mission to straighten out the governance of Canadian firms, we must
ensure that these institutions have the experience and the necessary know-
364
WHY INSTITUTIONAL INVESTORS ARE NOT BETTER SHAREHOLDERS
how to get the job done. In our opinion, very few pension funds are properly
equipped. The fact that public pension funds have often lead the corporate
governance movement is consistent with that opinion: public pension funds
have more resources than other institutional investors because they are much
larger and because they manage a larger proportion of their assets in-house.
And even then, few have ventured into micro-management.
Second, micro-managing does not confer the same economies of scale on
the investor. The knowledge required to appraise a particular restructuring, for
instance, may have little utility for another firm in another sector. Third, such
moves are likely to be perceived negatively by corporate managers, and this
increases the likelihood of retaliation. A pension fund manager may fear that
the enterprise it is trying to micro-manage will retaliate by depriving the
pension fund's sponsor of future business, or by cutting out the sponsor from
informal information flows. The same rationale applies to banks, insurers, and
so on. This retaliatory threat is a private cost that is likely to be borne by the
institutional entrepreneur and is particularly credible in the Canadian context,
which is characterized by a high level of ownership concentration and by the
existence of a dense network of firms and directors.
Finally, while having a representative on the board is a key factor in the
ability to micro-manage or monitor business policy, few institutional investors
will accept the private costs of sending a representative to the board. Having a
representative on the board is costly and increases the liability of the institution
— important private costs that cannot be shared with other shareholders. First,
there is the opportunity cost of delegating an officer or a representative of the
institution to the board. Then consider the increased liability that the insti-
tution must face. Since board members cannot fully insure themselves against
all risks (Daniels & Waitzer, 1994), the decision to have a representative on
the board increases the risk of the institution. This delegation also limits the
ability of the institutional owner to trade because of insider trading provisions.
The same provisions force the institution to comply with extra disclosure
rules. In addition, the institutional investor may find itself in a position of
conflict of interest if holding a large equity stake, for instance, can be deemed
a breach of fiduciary duty. Further analysis of this issue would take us into legal
territory,
7
 but we believe a case can be made that these costs and additional
liabilities will tend to tip the balance toward other modes of intervention.
The incentives of institutional investors to invest in generic research to
ameliorate the governance of specific firms or industries are further reduced by
the public-good aspect of the information they might produce. Once it is
known that a particular public pension fund or a large mutual fund devotes
resources to these ends, their decisions and the actions they undertake will
signal to other investors the opportunity they have identified.
365
PATRY & POITEVIN
AGENCY PROBLEMS: MONITORING BY PLAN ADMINISTRATORS
AND MONEY MANAGERS
ASSUMING THAT IT is IN THE INTEREST OF A GIVEN institutional investor to
influence the management or the policies of some companies it invests in, will
the administrators of the funds, the employees and the money managers
behave in ways that foster that interest? One response could be that, were it in
the interest of mutual funds or pension funds actively to engage in influence
activities, they would adopt an appropriate incentive structure. Alternatively,
one may study the present organization and ask: within that context, will
agents behave in the principal's interest?
We argue that institutional investors do not necessarily maximize risk-
adjusted returns. First, the drive to maximize profits and to undertake the influ-
ence activities can be muted because of governance defects. Second, the dele-
gation of management to administrators and money managers may also be
plagued by moral hazard and adverse selection problems.
CONSEQUENCES OF MORAL HAZARD
WHEN A MONEY MANAGER IS HIRED by a mutual fund, the typical fee-contract
states that there is a duty to perform the normal proxy activity, to vote in the
best interest of the fund, and to submit regular reports to an advisor. A money
manager receives less than 1 percent per year on the assets he/she manages.
Such fees do not cover the costs of most actions aimed at effecting a change in
corporate governance (Pozen, 1994, p. 144). As a result, if money managers in
this segment expend resources to monitor firms, they will generally refrain
from taking steps that imply heavy costs, such as filing and preparing proxy
materials, writing to all shareholders and hiring consultants and lawyers. In
addition, the advisor has little information to evaluate the governance input
of the managers, which is the root of the moral hazard problem.
Since the actions of the money manager are hidden, and since monitoring
and influencing imply costs but generate returns that are not easily observable
by the advisor, the money manager will be tempted to cut costs and to minimize
the effort to influence management. This problem is compounded if the horizon
of the money manager is short, since he/she will then discount any future
benefits heavily. Finally, there are good reasons to believe that mutual fund
money managers will favour exit over voice. This is evidenced by the higher
turnover rate of equity in mutual funds. This preference for liquidity is linked
to the redeemable nature of the claims on mutual funds and reinforced by the
dominant exit strategy of individual investors in pension funds.
Although pension funds value liquidity much less and claim to focus on
the long-term (Bauman & Miller, 1994, p. 32), we have seen that fund admin-
istrators may have problems inducing their money managers to adopt a long-
term horizon. Management fees paid in Canada are also somewhat lower than
366
WHY INSTITUTIONAL INVESTORS ARE NOT BETTER SHAREHOLDERS
those paid in the United States and the United Kingdom (30 basis points
compared to 35 and 39 )
8
 and this means that managers have few resources to
commit to monitoring (although we do not understand why the circumstances
are different in the United States and the United Kingdom). In addition,
pension fund money managers appear to be rewarded not only for their stock
market performance, but also for their capability to produce other goods, such
as schmoozing.
CONSEQUENCES OF ADVERSE SELECTION
SPONSORS APPEAR TO LACK A VALID SCREENING STRATEGY. Although this issue
deserves much more research, there is evidence of a very high turnover of
money managers and of some "hire high, fire low" patterns. It is possible that
treasurers are satisfied with the process because it provides them with the
scapegoats they seek.
9
Activists like Black (1992), who claims that "... if monitoring is profitable,
financial institutions ... will find ways to pursue that profit" (p. 876), call for
the selection of managers on the basis of their level of monitoring effort. We
consider this option to be interesting but it raises other questions.
First, sponsors would have to be convinced that this is a good thing, and
that it should be taken into account when evaluating money managers (which
does not now appear to be the case). Second, money managers will react to
such contractual changes by signalling their concern and demonstrating the
diligence and seriousness with which they take their fiduciary duties. Unless
good signals can be devised, the outcome may just be an increase in schmoozing.
The desirability of including formal measures of monitoring in the process of
evaluating managers depends on the noisiness of the signals. Another possible
drawback might be increased pressure on money managers by corporate
managers to support the sponsor's business, thus exacerbating conflicts of
interest.
The apparent high value that sponsors place on the recent track record
of the managers they hire is yet another indication that the market for outside
money managers is, perhaps, subject to adverse selection problems. Since
performance is elusive and inconsistent, and since the benefits of sound
corporate governance are long-term in nature, money managers will find it
difficult to extract a higher compensation in exchange for a promise of higher
quality. Lakonishok et al. (1992) confirm that conjecture (p. 372). They also
state that "perceptions of the qualities of individual firms vary widely over
time and across customers" (p. 364). In other words, reputations are difficult to
develop and maintain and this limits the role of such reputations as bonding
devices. Reputational effects are therefore not very likely to help sponsors
choose good managers.
367
PATRY & POITEVIN
POLICY ANALYSIS AND CONCLUSIONS
EXIT AND VOICE INNOVATIONS
THIS SECTION PRESENTS A CRITICAL REVIEW of many of the propositions found
in the literature aimed at enhancing the role of institutional investors in the
governance of firms. The propositions intended to improve exit and voice
mechanisms can be regrouped into two classes. First, those that limit the
investment strategy of funds managers; second, those that address directly the
monitoring problem facing investors.
RESTRICTIONS ON INVESTMENT STRATEGY
HlRSCHMAN (1970) HAS SUGGESTED THAT ONE WAY to enhance the effectiveness
of voice is to restrict exit. This raises the question of whether authorities
should encourage the adoption of restrictions to institutional investors' strategy
in order to force money managers into a more active role. Exit could be
precluded by the rules governing the fund (as the limit on foreign ownership
already does), or its cost could be increased through some form of regulation
(such as a "flip-tax" on capital gains) or, quite simply, by imposing some form
of indexing on a percentage of the funds, thus blocking to some extent the
exit option.
Underlying these reform proposals is the idea that institutional investors
can be "forced" to allocate more resources to monitoring and influencing
activities. These reform proposals regard institutional investors as "black boxes"
and, it is argued, by constraining their choice sets, they can be induced to invest
more in corporate governance. We have argued to the contrary, i.e., that the
governance and internal organization of investors also matter.
We disagree with such reform proposals and suggest that they are unwar-
ranted for financial reasons. First, it is not clear that investors exit too often or
that exit is ineffective when compared to voice. Those advocating such
restrictions have yet to demonstrate the relative ineffectiveness of exit in the
Canadian context since Canada has sophisticated and well-functioning capital
markets, similar to those in the United States and the United Kingdom.
Second, it must be stressed that the effectiveness of voice also depends
on the existence of a credible exit threat. The voice of CalPERS was amplified
by the risk of not only the shareholder resolutions, but also massive exit - with
its damaging consequence on stock price. By further constraining the exit
strategy of Canadian institutions, we could also lower the efficiency of their
voice mechanism.
Third, the narrowness of the Canadian equity market, its relative lack of
liquidity, and the closet indexing that results from sponsors' diversification
across money-manager types and through the indexing of active money
managers who want to lock in their superior performance, all lead us to believe
368
WHY INSTITUTIONAL INVESTORS ARE NOT BETTER SHAREHOLDERS
that unless radical regulations were adopted, little change in behaviour would
result from increased indexing per se. In fact, large Canadian institutional share-
holders are already locked into the relatively narrow Canadian equity market
with the result that the Ontario Teachers Pension Fund, the Caisse de Depot et
placement, the Ontario Municipal Employees Retirement Board (OMERS), and
other similar large institutional investors already hold relatively large stakes in
the top Canadian firms. Ownership concentration in Canada is higher than in
the United States and liquidity is lower, but Canadian institutional investors'
behaviour has paralleled that of their American counterparts. Again, we take
this as evidence that unless very radical changes are contemplated, institutional
investors are not likely to become more active.
Deliberately restraining the liquidity of Canadian institutional investors
has costs of its own. For example, the overall efficiency of Canadian financial
markets would decline, and the informational content of share prices would be
adversely affected. In efficient financial markets, movement in share prices
transmits some (noisy) information from informed investors to uninformed
ones. This transfer of information improves the functioning of financial markets
and contributes indirectly to the effectiveness of corporate governance. The
monitoring of informed investors conveys information through prices to financial
markets. The market then disciplines those firms that are not performing
satisfactorily by increasing their financial costs.
The informational content of share prices provides a delicate balance
between the provision of incentives for firms to perform and the provision of
incentives for investors to monitor. Formally, as Grossman & Stiglitz (1976)
show, in equilibrium, the proportion of monitoring financiers is such that they
receive a fair financial compensation for their monitoring costs. Potential
monitors, therefore, take into account the noisy leakage of the information
they may acquire; they also take into account the fact that if they do not monitor,
share prices will still transmit some information. We believe that these consid-
erations are crucial and should be addressed when evaluating any restriction
on the investment policy of institutional investors.
One such restriction which already generates serious by-products is the
regulation on foreign ownership. In Canada, the Income Tax Act imposes a 20
percent ceiling on the amount of the equity (at book value) that a corporate
pension fund can hold in foreign securities. By concentrating the investments
of Canadian pension funds, it can be argued that fund managers are induced to
improve the corporate governance of Canadian firms.
We believe, however, that this effect is likely to be small. For as we have
seen, despite the relatively higher concentration of ownership in Canada, the
behaviour of Canadian institutional investors does not differ markedly from
that of their U.S. counterparts. On the other hand, this restriction imposes
two significant costs on the Canadian economy.
First, whatever corporate governance gains are generated by the restriction
should be carefully weighted against the cost to Canadian workers of limiting
369
PATRY & POITEVIN
the diversification of their pension funds. This cost is likely to be important
for two reasons: Canadian market capitalization is only a small percentage of
world equity, and Canadian equity is biased towards natural resources industries.
Second, consider the implicit cost induced by the conflicts of interest
spawned by this regulation. Large Canadian institutional investors are major
shareholders of almost all corporate Canada and their compliance with insider
trading laws and conflict-of-interest rules poses serious questions. When an
important institutional investor, such as the Ontario Teachers Pension Fund,
also assumes the role of dealmaker in assisting takeovers, the ethical, legal, and
governance issues become even more complicated.
10
 Relaxation of the foreign
ownership regulation would immediately alleviate both these costs, while at
the same time reinforcing the effectiveness of the exit strategy by giving it
more credibility.
To reiterate, the removal of the foreign equity ownership regulation
would, in our opinion, have little detrimental effect on the governance of
Canadian firms. In fact, it might add power to the disciplinary role of exit.
IMPROVEMENTS IN MONITORING INCENTIVES
REFORMERS CLAIM THAT NEW INSTITUTIONS could provide stronger monitoring
incentives. Two such proposals encourage the creation of a class of professional
monitors and relational investing. Both suggestions contain interesting ideas,
but we point out some deficiencies that are bound to impair their functioning.
Gilson & Kraakman (1991) recently proposed that a (new) class of
professional monitors who would act on five or six boards and whose task
would be to monitor management might be an appropriate solution to many
corporate governance problems. These monitors would be remunerated by the
firms themselves but, more importantly, would have the incentive to monitor
by reputation effects on a newly created market for these monitors. Remember,
however, that corporate governance imperfections are caused primarily by
informational problems. By viewing corporate governance against this back-
drop, it becomes apparent that the addition of a class of professional monitors
does not address the problems at the heart of the corporate governance debate.
First, all the problems of the market for managers would be replicated in
the market for monitors. How would the market evaluate these monitors? It
would have to use the same performance measures as those used for managers.
In our opinion, this market would have no advantage in generating reputation
effects strong enough to improve monitoring and, consequently, strong enough
to improve managerial effort. Second, we believe that the problem of board
capture by management - although attenuated — would likely resurface with
this class of professional monitors because they would join the network of
managers and insiders. In the absence of clear signals on their efficiency,
problems of moral hazard will plague the outside monitors. Clubability would
also be a factor, particularly in the small Canadian market. Third, since the
370
WHY INSTITUTIONAL INVESTORS AR E NOT BETTER SHAREHOLDERS
quality of monitoring is hardly observed, we should expect professional
monitors to be engaged in a certain amount of schmoozing in order to convince
their sponsors that they diligently accomplish their function. Finally, we note
that there are at present no legal or institutional barriers to prevent the
emergence of such a class of outside directors. This may be an indication that
the innovation is not profitable.
Relational investing has also received some attention recently.
Relational investing consists of having long-term minded investors who
commit to buy and hold significant blocks of a firm's shares. Relational
investors participate actively in the strategic decision-making of the firm.
Ayres & Cramton (1993) discuss the advantages of relational investing in
detail.
First, they claim it provides strong incentives to monitor and influence
since the investor can appropriate a large share of the gains from such activities.
Repeated play between the investor and management is assumed to enforce
the reputation effects necessary to induce monitoring by the investor. Second,
relational investing is said to alleviate moral hazard problems more effectively
than takeovers, to the extent that it insulates management from inefficient
takeover attempts (see our earlier discussion). Finally, they argue that a
relational investor can put adequate pressure on managers while internal or
external markets may apply too much, especially when there is a short-term
bias in these markets. As evidenced by the weak market for corporate control
in Europe and Japan, a long-term investor is certainly less likely to respond to
short-term movements in profitability or share prices if those movements are
transitory or not agency related.
Nonetheless, relational investing has problems of its own. First, it is
important that the investor commit to hold a large percentage of the firm's
shares. How can any investor, particularly an institutional investor, commit to
such a strategy? Even pension funds, which assign a relatively low weight on
liquidity, would be hard pressed to resist a takeover bid above current market
value. Premiums paid to target shareholders in North America normally average
over 30 percent and have at times been very large. Second, relational investing
tilts the incentive-diversification tradeoff in favour of more incentives for
monitoring and less portfolio diversification. Most institutional investors, and
pension funds in particular, will shrink from such a commitment. Once again,
nothing prevents institutional investors from pursuing that strategy now. It
therefore appears that this tradeoff is well understood by investors and that
they find micromanaging too costly in terms of reduced risk diversification and
liquidity.
DISCLOSURE RULES INNOVATIONS
THIS STUDY FOCUSES ON ASYMMETRIC INFORMATION and agency problems to
explain the inefficiencies found in the design of financial contracts and in the
371
PATRY & POITEVIN
corporate governance process. The first step in alleviating informational problems
is to monitor a firm's activities. Any improvement in the output of monitoring is
likely to improve corporate governance. We suggest some innovations with
respect to disclosure rules that would contribute to this improvement.
Recently, Quebec and Ontario proposed regulations requiring firms to
disclose the compensation package of their five most highly paid managers.
Such information is useful in analyzing the incentives afforded to managers,
and is therefore a valid input in monitoring activities that assess moral hazard
and adverse selection problems. We firmly support this innovation. In fact, we
believe this regulation should be extended to include managers of mutual
funds and pension funds.
These funds have their own corporate governance problems that spill
over onto the governance of the firms in which they invest. Disclosing the
compensation packages of fund managers can only help the corporate gover-
nance process. Investors will learn how to utilize such information and this is
likely to trigger exit and/or voice reactions. This should help "monitor the
monitor".
At present, money managers' contracts are based almost exclusively on
the size of the assets they manage. In a static context, such compensation is
likely to provide low-powered incentives. In a dynamic context, as we have
seen, money managers may have an incentive to increase their asset base to
increase their next period remuneration. Therefore, it is plausible that
appropriate incentives are provided by these compensation packages.
We believe that better disclosure rules may attract investors' attention to
the problem of designing a compensation scheme that motivates fund managers.
For example, we have argued that, in a number of cases, money managers will
try to maximize the expected return for a chosen level of risk. However, expected
return is not the only characteristic that interests small investors. They are also
interested by the level of risk. Since risk is unobservable to investors, managerial
compensation could be made contingent on indices of a fund's "riskiness", using
a measure such as the historical variance of the fund's returns. Finely tuned
compensation packages could induce money managers to provide investors with
their desired risk-return profile. Another possible innovation would consist, for
investment funds, to disclose the B value of their portfolio.
Another disclosure rule innovation that could reduce monitoring cost
would be to modify accounting standards in order to include more economically
relevant data in financial statements. Several areas might be considered:
mandatory reporting by firms of market shares, product quality, productivity
levels, labour and customer satisfaction, quality control indices, etc. These
performance indices, if reported for the firm and the industry as a whole, and
for, say, a five-year period, would constitute a useful yardstick in the assessment
of the performance of a management team.
Economic information would complement traditional accounting and
financial data. One disadvantage of such indices may be a weakening of the
372
WHY INSTITUTIONAL INVESTORS ARE NOT BETTER SHAREHOLDERS
strategic position of some firms that compete in a global environment with
other firms that are not subject to such disclosure rules. The scope of these
measures must therefore be assessed carefully before being implemented.
CONCLUSIONS
CLEARLY, THE POTENTIAL FOR INSTITUTIONAL ACTIVISM is much lower in
Canada than in the United States. Fewer companies are widely held in
Canada, which makes for fewer pivotal institutional investors. For most
investors, most of the time, a passive attitude is rational. We also believe that
common sense points towards public and private pension funds as the most
promising would-be activists, monitors and influence seekers.
We argue that the weaknesses in the governance of pension funds must
be dealt with before any significant improvement will be seen in the internal
organization of those funds and, consequently, before any dividend that might
accrue from the improved governance of Canadian firms can be reaped.
Given that pension funds may play a more prominent role in the
governance of Canadian firms in the near future, we believe that generic
issues, of the process and procedural types, are most likely to emerge. We
also conjecture that pension funds will prefer informal modes of intervention
to formal modes, and conciliatory approaches to corporate governance issues
to adversarial approaches. Progress could be made quickly if the largest funds
developed ways and means to coordinate their behaviour. A detailed analysis
of the equity portfolios of the 20 largest pension funds, for example, could shed
light on this question.
We also offer some tentative comments on the desirability of a few
regulatory innovations against a rather bleak portrait of proposed changes to
corporate governance in Canada. It is our contention that fine-tuning financial
regulation cannot bring significant changes to governance behaviour. In many
ways, the Canadian context is significantly different from the American
context; despite this, the behaviour of Canadian and American institutional
investors is strikingly similar.
One may point to the Japanese and German regimes, which are markedly
different, but there is no conclusive evidence that either the Japanese or the
German system of governance is superior. Claims of superiority are generally
based on the belief that Japanese and German regimes favour long-term
relationships and relational investing, and avoid "short-termism". For example,
there is evidence that "liquidity constraints" are more important for North
American firms.
11
 This suggests that internal financing is cheaper than external
financing for these firms, which may mean that capital markets are plagued by
informational problems.
But the liquidity constraints facing North American firms can be
interpreted as optimal constraints imposed on management teams to prevent
them from making non-profitable investment. Jensen's (1986) theory of free
373
PATRY & POITEV1N
cash flows makes this point clear. Cash-constrained firms may be more efficient
since they cannot invest in negative-value ventures. Whether this argument
can then be extended to Japan and Germany, thereby claiming that managerial
capture of investors allows them to invest too much (namely in non-profitable
projects) is an empirical question. These ideas have not been proven formally
in the literature nor have they been verified empirically. They are, at present,
only speculation. However, they certainly deserve more attention before
making radical policy changes to the Canadian system.
Moreover, there is recent evidence of moral hazard and adverse selection
problems in the Japanese and German regimes. For example, the deregulation
of Japanese financial markets, which has facilitated the issue of corporate
bonds, has induced many Japanese firms to move away from keiretsu financing
toward North American-type financial contracting. This is certainly at odds
with the belief that Japanese financial contracting is superior. Also, recent
evidence of the capture of banks by management in Germany shows that the
German system is not collusion-proof nor, indeed, as efficient as it was
thought to be.
12
Our knowledge at present does not allow us to draw any firm conclusions
regarding the relative efficiency of the Japanese and German models vis-a-vis
the North American model. Without such evidence, we suggest that major or
radical changes to our present system of corporate governance would be totally
unwarranted. Furthermore, assuming the superiority of the corporate governance
regimes of Japan and Germany, we doubt that they could be easily cloned or
imitated; institutions are imbedded in history.
Conversely, we have also argued that many regulations aimed at reducing
the information asymmetry between investors and money managers should
improve the efficiency of the corporate governance regime in Canada.
On the one hand, we contend that some reforms could produce social
benefits, but we also suggest that they would likely be marginal. On the other
hand, major changes should be scrutinized for their important ramifications on
the efficiency of Canadian financial markets. Those markets are embedded in
a complex web of beliefs, attitudes and values and have served us very well.
We should therefore ensure that important regulatory changes do not unwit-
tingly reduce their efficiency or alter their fundamental value.
374
WHY INSTITUTIONAL INVESTOR S ARE NOT BETTER SHAREHOLDERS
ENDNOTES
1 See Black (1992) for an ardent plea in favour of expanding the role of
institutional investors in the United States.
2 Fortune and Canadian Business, 1993.
3 In Canada, as of January 31, 1995, 1077 mutual funds are listed in the Bell
Charts of the Financial Times of Canada. In addition, entry is rather easy
and has occurred, particularly by banks, following deregulation of the
financial sector. Foreign competition is increasing. Information on the
performance of each fund is widely available, with newspapers, specialized
newsletters, and other sources regularly reviewing the performance of all
funds.
4 Statistics Canada, Trusteed Pension Funds, 1992.
5 The Financial Post, Canada's Largest Corporations, 1994.
6 "A strategy of selecting, every quarter, the top performers based on the last
four quarters . . . can significantly outperform the average mutual fund,
albeit doing only marginally better than some benchmark market indices"
(Hendricks, Patel & Zeckhauser, 1993, p. 122).
7 See Black (1990) and Daniels &. Waitzer (1994) for a discussion of these
questions.
8 Greenwich Associates, Survey of Canadian Pension Funds, 1994.
9 Why do firms allow their treasurers to entrench themselves like that?
10 "Pension Fund Power." The Globe and Mail, Saturday, June 17, p. Bl and B3.
11 Bascunan, Poitevin 6k Garcia (forthcoming), for instance, verify if there is
any connection between the investment level of firms and their cash flow.
They find that Canadian and American firms are liquidity constrained, while
Japanese and German firms are not.
12 The Economist (January 29, 1994) notes that some German firms, like
Siemens, Daimler-Benz, and Metallgesellschaft have been slower than
American firms to spot problems. In the same article, a German analyst
and lawyer blames the bad investments of many large German enterprises
on the weak supervision of the banks.
375
PATRY & POITEVIN
BIBLIOGRAPHY
Ayres, I. and P. Cramton. "An Agency Perspective on Relational Investing", Working
Paper No. 105, John Olin Program in Law and Economics, Stanford Law School, 1993.
Bailey, J. V. "Some Thoughts on Performance-Based Fees." Financial Analysis Journal, (July-
August 1990) :31-40.
Bascunan, M-, M. Poitevin and R. Garcia. "Information asymetrique, centrainte de liquidite
et investissement: une comparaison internationale." L'Actualite economique.
(Forthcoming).
Bauman, W. S. and R. E. Miller. (1994) "Can Managed Portfolio Performance Be
Predicted?" The Journal of Portfolio Management. (Summer 1994):31-40.
Baumol, W. J. The Economics of Mutual Fund Market: Competition versus Regulation. 1990.
Beaudry, P. and M. Poitevin. "Competitive Screening in Financial Markets when
Borrowers can Recontract." Review of Economic Studies, 62 (1995):401-23.
Beebower, G. and G. Bergstrom. "A Performance Analysis of Pension and Profit-Sharing
Portfolios: 1966-1975." Financial Analysts Journal, 33 (1977):31-42 .
Berkowitz, S. A and D. Logue. "The Portfolio Turnover Explosion Explained." Journal of
Portfolio Management. (Spring 1987).
Berkowitz, M. K. and Y. Kotowitz. "Incentives and Efficiency in the Market for
Management Services: A Study of Canadian Mutual Funds." Canadian Journal of
Economics, XXVI, 4 (1993).
Berkowitz, M. K., Y. Kotowitz and L. Waverman. "Enterprise Structure and Corporate
Concentration: A Technical Report", Ottawa, Ministry of Supply and Services, Canada,
1977.
Berle, A. A. Jr. and G. Means. The Modem Corporation and Private Property. New York:
Harcourt, Brace and World Inc., 1932.
Bhagat, S., A. Shleifer, and A. Vishny. "Hostile Takeovers in the 1980s: The Return to
Corporate Specialization." Brookings Paper on Economic Activity, 1990.
Black, B. S. "Shareholder Passivity Reexamined." Michigan Law Review, 89 (1991):520-607.
. "Agents Watching Agents: The Promise of Institutional Investor Voice." UCLA
Law Review, 39 (1992):811-93.
Black, F. and M. Scholes. "The Pricing of Options and Corporate Liabilities." Journal of
Political Economy, 81 (1973):637-54.
Brinson, G., L. R. Hood, and G. Beebower . "Determinants of Portfolio Performance."
Financial Analysts Journal, 43 (1986):39-44-
Coffee, J. C. "Liquidity versus Control: The Institutional Investor As Corporate Monitor."
Columbia Law Review, 91, 6 (1991).
Connor, G. and R. Korajczyk. "The Attributes, Behavior, and Performance of U.S. Mutual
Funds." Review of Quantitative Finance and Accounting, 1 (1991):5-26.
Daniels, R. and J. Macintosh. "Toward a Distinctive Canadian Corporate Law Regime."
Osgoode Hall Law Review, 29 (1991).
Daniels, R. and E.J. Waitzer. "Challenges to the Citadel: A Brief Overview of Recent
Trends in Canadian Corporate Governance." Canadian Business Law Journal, 23
(1994):23-44.
Demsetz, H. "Financial Regulation and the Competitiveness of the Large U.S.
Corporation." Federal Reserve Bank of St. Louis Bulletin, March 1993.
Demsetz, H. and K. Lehn (1985) "The Structure of Corporate Ownership: Causes and
Consequences." Journal of Political Economy, December 1985:1155-77.
376
WHY INSTITUTIONAL INVESTORS ARE NOT BETTER SHAREHOLDERS
Fowler, D. J. and C. H. Rorke. "Insider Trading Profit on the Toronto Stock Exchange."
Canadian Journal of Administrative Science (1988).
Gilson, R. J. and R. Kraakman. "Reinventing the Outside Director: An Agenda for
Institutional Investors." Stanford Law Review, 43 (1991):863-906.
Grinblatt, M. and S. Titman. "Mutual Funds Performance : An Analysis of Quarterly
Portfolio Holdings." Journal of Business, 62 (1989):393-416.
Grossman, S. and O. Hart. "Takeover Bids, The Free Rider Problem and the Theory of the
Corporation." Bell Journal of Economics, 11 (1980):42-64.
Hendricks, D., J. Patel, and R. Zeckhauser. "Hot Hands in Mutual Funds: Short-Run
Persistence of Relative Performance, 1974-1988." Journal of Finance. (March 1993):93-130.
Hirschman, A. O. Exit, Voice and Loyalty, Responses to Decline in Firms, Organizations, and
States. Cambridge, MA: Harvard University Press, 1970.
Ippolito, R. "Efficiency with Costly Information: A Study of Mutual Fund Performance,
1965-1984." Quarterly journal of Economics, 104 (1989):l-23.
Jensen, M. C. "The Performance of Mutual Funds in the Period 1945-1964." Journal of
Finance, 23 (1968):389-416.
. "Agency Costs of Free Cash Flow, Corporate Finance, and Takeovers", American
Economic Review, 76 (1986):323-29.
. "Takeovers : Their Causes and Consequences." Journal of Economic Perspectives, 2,
1 (1988):21-48.
. "Eclipse of the Public Corporation." Harvard Business Review, 67 (September-
October 1989):61-74.
Lakonishok, J., A. Shleifer, and R. Vishny. "The Structure and Performance of the Money
Management Industry." Brookings Papers on Economic Activity: Microeconomics 1992,
(1992):339-91.
Lakonishok, J., A. Shleifer, R. Thaler, and R. Vishny. "Window Dressing By Pension Fund
Managers." American Economic Review. (May 1991):227-31.
Mitchell, O. S. and P. L. Hsin. "Public Pension Governance and Performance," Working
paper No. 4632. NBER, January, 1994.
Patry, M. and M. Poitevin. "Hostile Takeovers: The Canadian Evidence," in Corporate
Globalization through Mergers and Acquisitions. Edited by L. Waverman. Calgary:
University of Calgary Press, 1991.
Pozen, R. C. "Institutional Investors: The Reluctant Activists." Harvard Business Review,
(January-February 1994):140-49.
Rock, E. B. "The Logic and (Uncertain) Significance of Institutional Shareholder
Activism." The Georgetown Law Journal, 79 (1991):445-506.
Romano, R. "Public Pension Fund Activism in Corporate Governance Reconsidered."
Columbia Law Review, 93, 4 (1993).
Sappington, D .E. "Incentives in Principal-Agent Relationships." Journal of Economic
Perspectives, 5, 2 (1991):45-66.
Shleifer, A. and L. Summers. "Breach of Trust in Hostile Takeovers." in Corporate
Takeovers: Causes and Consequences. Edited by A. J. Auerbach. Chicago: University of
Chicago Press, 1988.
Shleifer, A. and R. Vishny. "Large Shareholders and Corporate Control." Journal of Political
Economy. (1986):461 -88.
Taylor, K. "Return on the Index Fund." Benefits Canada. (February, 1995):25-26.
Wang, W. Y. "Corporate versus Contractual Mutual Funds: An Evaluation of Structure and
Governance." Washington Law Review, 69 (1994).
377
This page intentionally left blank 
Stephen Foerster "1 f\
Western School of Business -L \J
University of Western Ontario
Institutional Activism by Public Pension
Funds: The CalPERS Model in Canada?
INTRODUCTION
I
NSTITUTIONAL INVESTORS INCLUDE PUBLIC PENSION FUNDS, private (corporate)
pension funds, mutual funds and insurance companies. According to Fare's
(1990) survey of chief executive officers of major U.S. companies, public
pension fund managers were perceived to be the most activist group of share-
holders. Given the many similarities and close proximity to Canada, public
pension funds in the United States are a natural focal point for an investigation
of investor activism (although, on a worldwide basis, the United Kingdom is also
viewed as a world leader with respect to addressing corporate governance issues).
The California Public Employees' Retirement System (CalPERS) is the
largest public pension fund in the United States, with assets of around $US 80
billion (approximately $C 112 billion) and is also regarded as the most active
institutional investor in the United States. CalPERS has been involved in
proxy fights since 1985. It also has a policy of targeting poorly performing
companies and attempting to improve them by meeting with management,
issuing shareholder resolutions, or voting against their boards of directors.
Every year since 1991 the fund has issued private early warnings to targeted
companies and has subsequently gone public with a revised (and usually
shorter) list of companies which have not addressed its concerns raised in
private meetings. A recent study by Nesbitt (1994) of 42 campaigns conducted
by CalPERS between 1987 and 1992 found that targeted companies under-
performed the S&P 500 Index by 60 percent prior to CalPERS' involvement,
then outperformed the index by 40 percent subsequent to CalPERS' involvement.
Currently in Canada (and other countries outside the United States),
there appear to be few (if any) public pension funds that approach the level of
activism displayed by CalPERS. An important question for Canadian
managers and regulators is why public funds in Canada have not been as
visibly active as CalPERS. There are at least four different hypotheses that
may explain this phenomenon.
379
100
FOERSTER
One possible explanation has to do with regulatory differences between
the two countries. A second possible explanation is that, in Canada, third-
party or intermediary groups such as Fairvest Securities Corporation or the
Pension Investment Association of Canada (PIAC) might be playing the
activism role. A third possible explanation is the often-observed lead-lag
relationship between U.S. and Canadian innovations or the sheer size differences
between the largest U.S. and Canadian funds. A fourth possible explanation is
the matter of style: Canadian fund managers may be more reluctant to engage
in overt "American-style" confrontations.
If the first explanation is valid, then regulatory changes may accelerate the
movement to CalPERS-style activism in Canada. If the second explanation is
valid, then these third-party groups will continue to grow in stature. If the third
explanation is valid, then examining the CalPERS model may serve as a useful
benchmark for the direction of institutional investor activism in Canada as
Canadian pension funds continue to grow and as Canadian fund managers
become more aware of the potential rewards of institutional activism. If the
fourth explanation is valid, then we should not expect to detect any visible
change in the way public pension fund managers approach activism issues. Of
course, a combination of these and/or other explanations is also possible.
The purpose of this study is to examine these four hypotheses in the
context of corporate governance and institutional activism, primarily based on
interviews with major Canadian public pension fund managers. Specifically,
the study has the following objectives.
• Provide a detailed framework of the CalPERS model of investor
activism.
• Assess the "evolution" of activism in Canadian public pension funds
(relative to the CalPERS model) and determine what corporate
governance issues are important to public pension fund managers.
• Provide regulators with a sense of how current regulations are
viewed by institutional investors.
• Provide corporate directors with a sense of what to expect in the
future in terms of institutional investor activism.
The study is organized as follows. The next section reviews public pension
fund activism in North America and examines the role of public pension
funds as activists. This is followed by a description of the CalPERS model and
an examination of some potential future trends related to the involvement of
CalPERS in corporate governance issues. The next section describes the
survey methodology used in the study and this is followed by the results of
interviews with Canadian public pension fund managers. The final section
provides a summary and conclusions.
380
INSTITUTIONAL ACTIVISM BY PUBLIC PENSION FUNDS: THE CALPERS MODEL
BACKGROUND
THE CANADIAN EXPERIENC E
ONE OF THE FIRST MAJOR PUBLIC INSTANCES of institutional activism in Canada
occurred in 1986 when members of the Billes family (founders of Canadian
Tire) struck a deal — later overturned by the Ontario Securities Commission -
to sell 49 percent of the company's voting shares to a group of Canadian Tire
dealers for a large premium over the market price, excluding from the deal
many large institutional holders of Canadian Tire non-voting shares (see
Stoffman, 1990 for a discussion).
Another major example of institutional activism was Inco's successful
introduction of Canada's first poison pill in 1988, an action vehemently
opposed by the Caisse de depot et placement du Quebec (which held 3 per-
cent of the stock) and other institutions. While poison pills or shareholder
rights plans come in a variety of forms, most provide special rights to a company's
shareholders in the event of an unwanted takeover attempt by another company.
For example, existing shareholders may be entitled to purchase shares in their
own company at a deep discount, thus diluting the value of a potential acquirer's
stake. In practice, poison pills are not triggered but instead force potential
acquirers to negotiate with the board. Critics claim this gives the board - not
the shareholders - veto power over any offers, while proponents argue that
poison pills allow the board to negotiate a higher price. In Canada, unlike the
United States, companies are required by provincial law to have their poison
pills ratified by shareholders. Canadian poison pills generally provide for
"permitted bids" to be put to shareholders for approval and if approved the
pill is not triggered. Inco's success was partially attributable to the linking of the
pill to a special $10 dividend. As of 1995, poison pill proposals remained an
important issue, with over 20 new proposals planned (see Mackenzie, 1994).
During the 1990s, numerous large pension funds have become more
actively involved in corporate governance issues, largely because of their growing
size and presence in equity markets — which makes it more difficult for them
simply to sell their shares if they are dissatisfied with management. In addition
to the Caisse, funds such as the Ontario Municipal Employees Retirement
System (OMERS), the Ontario Teachers' Pension Plan Board (Ontario
Teachers), CN Investments and other large funds have gained reputations in
the investment community for their involvement in corporate governance
issues, although this involvement is not always visible. A survey of Canadian
institutional shareholders by Montgomery (1992) indicated perceptions
among a vast majority (over 80 percent) of respondents that not only has
Canadian institutional shareholder activism increased in the past, it is likely
to continue increasing in the future.
Many of the issues that institutions have become involved in are structural
(rather than related to day-to-day management), such as dual-class shares,
various takeover defenses, executive compensation, and the size and composition
381
FOERSTER
of the board of directors (see Macintosh, 1993, for a detailed examination of
the role of Canadian institutional investors). In addition to institutional
investors themselves, intermediaries such as Fairvest, PIAC, as well as the fund
manager Jarislowsky, Fraser have acted on behalf of a larger number of
institutional investors.
Fairvest Securities Corporation is a securities firm which acts exclusively
as an agency trader for institutional investors. Formerly known as Allenvest
(named after its founder, Bill Allen), it has played an active role in corporate
governance issues. It is viewed by some institutions as providing a "conscience"
for shareholders. For some institutions which do not wish to play a visible role,
Fairvest provides a veil and yet allows them involvement in governance issues.
Fairvest keeps institutions informed through its publication, Corporate
Governance Review. Fairvest recently provided a submission to the Committee
on Corporate Governance of the Toronto Stock Exchange.
The Pension Investment Association of Canada (PIAC) is another inter-
mediary. It represents over 100 Canadian pension funds with combined assets
of over $250 billion. Its purpose is to provide a forum for debate and resolution
of issues facing pension funds. The role of PIAC is viewed as largely educational.
PIAC has also distributed corporate governance standards to its members.
There are many aspects of the Canadian market which restrict the
potential involvement of institutional investors (Montgomery & Leighton,
1993) such as the prevalence of dual-class share structure (over 200 TSE firms)
and the dominance of controlling shareholders (i.e., defined as those with at
least 20 percent of the voting power), estimated at 70 percent of the 100
largest public companies in the early 1990s, but thought to have declined
somewhat. Nonetheless, corporate governance remains a dominant issue,
particularly surrounding the release of the TSE's Dey Report on Corporate
Governance (see an overview by Star, 1994, and critiques by Thain, 1994, and
Macintosh, 1995).
THE U.S. EXPERIENCE
INVOLVEMENT OF CANADIAN INSTITUTIONAL INVESTORS in corporate governance
issues parallels many recent developments in the United States. During the
1990s, a new form of activism has replaced takeovers in the market for corporate
control (see Pound, 1992, and Pozen, 1994, for some recent examples).
Institutionally-led initiatives through proxy votes have gained increasingly
larger support over the years (Silverstein, 1994). Not surprisingly, publi c
pension funds are leading the way in terms of activism (see a survey of U.S.
institutional investors by Useem et al, 1993) partly because of their size, but
also because of their longer investment horizon and few cases of potential
conflicts of interest (e.g., if Company XYZ, which sells widgets, has a large
private pension fund owning shares in Company ABC, XYZ wil l be hesitant to
vote against ABC's management if ABC is a current or potential purchaser
382
INSTITUTIONAL ACTIVISM BY PUBLI C PENSION FUNDS: THE CALPERS MODEL
of widgets). The growth and concentration of institutional ownership in the
United States has been incredible. Between 1985 and 1990, institutional
holdings in the top 25 U.S. stocks grew from 36 percent to 46 percent
(Chernoff, 1993). While institutional shareholders as a group now own over
50 percent of the stock in U.S. companies, it is estimated that pension funds
alone will own one-third of all stock by the year 2000 (Cordtz, 1993).
Institutional shareholders have recently used their power to assist in removing
CEOs at such large, well-known (and often under-performing) U.S. companies
as American Express, General Motors, and Digital Equipment (Stewart,
1993). CalPERS has played a leadership role in many of these initiatives.
PUBLIC PENSION FUNDS AS ACTIVISTS
WHILE THERE ARE MANY REASONS WHY PUBLIC PENSION FUNDS may be natural
leaders in institutional activism — their size, their long-term horizon, and fewer
potential conflicts of interest - public pension funds also face potential problems.
As noted by Murphy & Nuys (1994), incentives and governance structures of
some public pension funds may provide only weak incentives to increase the
value of the funds. For example, they argue that with defined benefit plans,
pre-specified future retirement benefits are promised to public employees with
guarantees (ultimately) by state taxpayers who are the true residual claimants,
yet most boards of trustees of these plans have a fiduciary responsibility to the
beneficiaries (the retired public employees) and not to the residual claimants
(the taxpayers). They hypothesize that boards are less interested in generating
high returns and are more interested in avoiding paying high management
fees and avoiding taking large risks. They note that public pension fund managers
often receive low salaries and few financial incentives, and face constraints on
their ability to be effective monitors of the companies whose shares they own.
Fund managers often have limited expertise about specific companies and also
face ownership restrictions. However, the study noted that CalPERS and some
other funds have been able to circumvent some restrictions by hiring certain
key employees such as the Chief Investment Officer as outside consultants.
A recent study by Wahal (1995) noted that some public funds such as
CalPERS have been fairly successful in gaining acceptance for shareholder pro-
posals and persuading firms to adopt changes. However, Wahal questions the
long-term effect on stock prices of targeted companies. For the complete sam-
ple of stocks, mean "excess" performance is generally positive in the three
post-target years (particularly for pre-target underperformers) but not significantly
different from the market-adjusted returns. However, some results in the same
study (described below) suggest CalPERS may be one public pension fund that
has had some effect.
383
FOERSTER
THE CALPERS MODEL
I
T IS WORTHWHILE NOT ONLY TO DESCRIBE the CalPERS model of activism, but
also to trace its evolution and project likely trends. This overview will allow
Canadian fund managers to benefit from the CalPERS learning experience.
BACKGROUND
As OF 1993, CALPERS WAS THE THIRD LARGEST PENSION FUND in the world
(behind the New York fund TIAA-CREF and the Netherlands fund Algemeen
Burgerlijk), and was the largest public pension fund in the United States. It
served more than one million members: retirees from various local, county,
and state governments. The fund was expected to continue to grow in the
near future, since although it paid out more than $US 8 million each day in
benefits, it also had over $US 15 million left each day for investment (see
Ybarra, 1995, for a recent description of CalPERS and a history of its involve-
ment with corporate governance issues).
As of 1993, CalPERS owned $US 22.4 billion in U.S. equities - over
1,000 U.S. companies with an average stake of approximately 0.6 percent of
each company. Forty-one percent of assets were in fixed income, 40 percent
were in equities (roughly two-thirds domestic and one-third or $US 8 billion
in international equities), and the remainder in real estate, private equity,
mortgages and cash. Approximately 80 percent of the equities were passively
managed internally, while the remainder were managed by approximately 20
outside managers.
THE HISTORY OF CALPERS ACTIVISM
THE ROOTS OF CALPERS' INVOLVEMENT IN ACTIVISM can be traced to the
corporate merger and acquisition activities of the early 1980s. In order to maintain
control, management of several companies resorted to anti-takeover devices
such as greenmail, scorched earth policies, and poison pills. Numerous studies have
shown that these techniques often lead to a decrease in the company's stock
price. Many of these stocks were held by pension funds such as CalPERS.
Consequently, as early as 1985, CalPERS found itself involved in corporate
governance issues, primarily through proxy battles at annual meetings.
According to CalPERS representatives, one of the major reasons for CalPERS
involvement was the unwillingness of corporate executives to communicate in
any manner, even by simply returning phone calls.
In 1986, Jesse Unruh, a member of the CalPERS investment committee,
helped form the shareholder rights organization known as the Council of
Institutional Investors (CII). In the same year, Richard Koppes was hired as
general counsel. In 1987, Dale Hanson was hired as Chief Executive Officer.
In that year, CalPERS began issuing shareholder resolutions at annual meetings.
The main focus of these resolutions was an attempt to eliminate poison pills
384
INSTITUTIONAL ACTIVISM BY PUBLIC PENSION FUNDS: THE CALPERS MODEL
recently instituted by various companies. According to CalPERS representatives,
shareholder resolutions were used because, at the time, this was the only tool
CalPERS had at its disposal.
These early CalPERS campaigns met with considerable resistance from
management and were largely unsuccessful (although in 1988, CalPERS
achieved its first shareholder victory over management opposition - an anti-
greenmail proposal). At the time, there were also Securities and Exchange
Commission (SEC) regulations (subsequently changed in 1992) which restricted
communication among investors, making it difficult for large institutional
investors to organize a joint effort on corporate governance issues. In addition,
the only communication with management was at annual meetings. However,
a major difficulty for CalPERS occurred in cases where it tried to chastise a
company (about to introduce a poison pill) which had also achieved record
profits and displayed a rising stock price.
In late 1989, with the assistance of outside consultants, CalPERS
changed strategies. In a document titled "Why Corporate Governance?"
CalPERS articulated its corporate governance philosophy. The focus was no
longer on specific management actions, such as the enactment of poison pills,
but rather on promoting structural changes aimed at improving performance.
For example, the fund proposed to advocate stronger boards with indepen-
dent directors and more qualified members. In addition, CalPERS emphasized
the importance of attempting to maximize its investment return and taking a
long-term investment outlook.
CalPERS continued to meet with considerable resistance during the early
phase of this new approach. Companies accused it of not understanding the
companies or their industries. CalPERS further inflamed corporations when, in
1990, it sent out a questionnaire to directors of S&P 500 companies asking
their views of the fund's new corporate governance emphasis.
In 1990 and 1991, CalPERS met with other members of the CII and
developed a list of ten principles of corporate governance, which was published
in the Harvard Business Review (The Working Group on Corporate
Governance, 1991). These principles included improving the board of directors
by having outside directors evaluate the performance of the CEO, screening
board candidates in a more effective manner, and having institutional share-
holders acting more as owners rather than investors (for a critique of these
principles, see Wharton, Lorsch & Hanson, 1991).
CalPERS attempted to emphasize the importance of dialogue between
management and institutional investors. According to Dale Hanson, former
CEO of CalPERS, it was important for management to know its top 20 share-
holders when a company was doing well so that those same shareholders would
be more understanding in bad times (Chief Executive, 1994). Although
CalPERS attempted a "kinder, gentler" approach in 1992 by scheduling
informal meetings without the threat of shareholder resolutions, this was
largely unsuccessful and CalPERS decided once again to submit shareholder
resolutions.
385
FOERSTER
In October 1992, largely due to CalPERS involvement, the SEC changed
the rule which restricted the ability of large institutions to communicate with
one another. Disinterested persons not seeking proxy authority were exempted
from proxy requirements and thus were permitted to communicate freely with
other persons or through the media. This rule change made it easier for large
institutions to plan voting strategies on key issues.
In 1992, and again in 1994, CalPERS commissioned studies by Wilshire
Associates to assess the effectiveness of their corporate governance program
(see Nesbitt, 1992, 1994, and a summary in The Economist, 1994). As noted
above, the first study of 42 CalPERS' campaigns between 1987 and 1992
found that targeted companies underperformed the S&P 500 Index by 60 per-
cent prior to CalPERS' involvement, then outperformed the index by 40 per-
cent subsequent to CalPERS' involvement. Furthermore, the second study
indicated that since 1990 CalPERS campaigns produced even more dramatic
results than the earlier campaigns aimed at specific management actions. This
study provided an important cost-benefit analysis of activism. According to
both studies, based on the average CalPERS' holding per stock of approximately
$US 35 million, the cumulative outperformance amounted to an average
gain in excess of the market of about $US 2.9 million per company or $US 137
million for all such targeted companies, well in excess of the total annual
administrative costs estimated at $US 500,000. A recent paper by CalPERS
general counsel and senior staff counsel, Koppes & Reilly (1994), referred to
this evidence as the "missing link" connecting the corporate governance
approach of CalPERS with the improvement of a firm's stock price. Further
evidence of the connection between corporate governance and performance
was found in an important study by the Gordon Group for CalPERS, which
highlighted the beneficial results from active investment strategies (Gordon &.
Pound, 1993).
A recent study by Wahal (1995) examined a number of active U.S.
pension funds and found that only firms targeted by CalPERS experienced a
significantly positive stock price reaction around the target date. Furthermore,
WahaPs results highlight the importance (in terms of effect on shareholder
wealth) of performance-based targeting, which appears to be the key contributor
to any "excess" stock price reaction.
The most recent public action by CalPERS involved the promotion of
the "GM Guidelines" - 28 principles related to corporate governance issues
developed by General Motors in March 1993. The principles examined selection
of the board chair and CEO, use of board committees, size of the board, mix of
inside and outside directors, criteria for selecting new directors, evaluation of
performance, succession planning, and other corporate governance issues. In
the fall of 1994, CalPERS circulated copies of the guidelines to 200 of the
largest U.S. companies, asking them to adopt similar principles. Follow-up
letters were issued to non-respondents warning of public grading of responses
(or non-responses). The grading scheme developed by CalPERS was as follows:
386
INSTITUTIONAL ACTIVISM BY PUBLIC PENSION FUNDS: THE CALPERS MODEL
A+ for an excellent response, A for a good response, B for a good response but
requiring more information, C for needing more information, D for missing
the point, F+ for a brush off, and F for no response. In October 1994, Business
Week published the results (Dobrzynski, 1994b). Subsequently, numerous
companies which had received a failing grade of F (for non-responses) were
re-graded after sending CalPERS their responses. In December 1994, CalPERS
expanded the survey to the next 100 largest companies. By that time, 83 percent
of the 100 largest companies had responded and almost half of these had
either completed, or were in the process of completing a governance self-
evaluation (i.e., received grades of either A or B).
THE CALPERS MODEL AS OF 1995
IN OCTOBER 1994, CALPERS TRUSTEES HIRED as CEO former assistant executive
officer for investment operations, James Burton, to replace Dale Hansen (who
had resigned from CalPERS in order to join an investment firm). Burton had
been involved in corporate governance issues for two years prior to the arrival
of Hanson. Subsequently, Burton designated general counsel Koppes with the
additional title of deputy chief executive and turned over corporate governance
responsibility to him. Although less visible than Hanson, Koppes was already
heavily involved in corporate governance.
The current version of the CalPERS model is a carefully timed, finely
tuned approach (see Hanson, 1993, for a recent overview of the process).
Around June of each year, an outside consulting firm receives a list of the
approximately 1,000 U.S. companies that form CalPERS internally managed
portfolio. The consulting firm assists with the initial screening by ranking the
companies based on their return to shareholders over the most recent complete
five-year period (it also reports the most recent six-month results).
Performance relative to each firm's industry is also examined (based on 2-digit
and 3-digit SIC codes), as are year-by-year results.
CalPERS then focuses on the bottom quartile - approximately 250 companies
— based on each firm's five-year absolute performance. These companies are
examined by CalPERS in greater detail. Other performance measures are
calculated, such as return on assets, return on equity, return on sales, profitability,
and payout ratios. In addition, any recently announced restructuring plans or
special circumstances (such as change in management) are noted, along with
the number of shares held by CalPERS.
CalPERS' own industry analysts examine these companies and also rely on
other independent reports such as Value Line. A brief assessment of each company
accompanies the report. This process usually takes place during July and August.
Several screens are put in place to reduce the bottom quartile list to the
"Failing 50" list. Companies such as utilities in heavily regulated industries
are eliminated. If an entire industry is doing poorly, most firms within this
industry might be eliminated. Firms owned by foreign parents are not targeted;
387
FOERSTER
nor are firms with either a high level of insider ownership (generally above 30
percent) or a low level of institutional ownership (generally less than 20 percent).
These last screens are aimed at increasing the odds of realizing improved
performance. Companies that have undergone recent restructuring are some-
times eliminated, if the restructuring is judged to be a serious effort and in the
best interest of the company.
Around September, the investment committee meets to finalize the
"Failing 50" list. In addition, based partly on recommendations from CalPERS'
Investment Office, the committee chooses ten of the 50 companies to target
for improved performance.
The worst ten companies, as well as the next-worst 20 to 25 companies
are sent letters - to the Board Chairs - indicating the amount of stock held by
CalPERS and a note of dismay concerning the company's performance. Firms
are asked to respond to specific problems and are asked what actions are being
taken to improve performance. In addition, meetings (especially with the
independent board members) are requested with all of the ten worst per-
former companies.
During the subsequent months, if the stock performance of the next-
worst firms improves dramatically, they may be removed from the watch list. If
the ten worst companies are not responsive to CalPERS suggestions, they are
then targeted for shareholder resolutions in the upcoming annual meeting
(typically around February or March). In late January or early February,
CalPERS makes public its list of the ten worst targeted companies. For example,
the 1995 list included First Mississippi, U.S. Shoe, Jostens, Boisie Cascade,
Melville, K Mart, Navistar International, Zurn Industry, and Oryx Energy.
CalPERS attempts to meet with these companies and resolve issues before
each company's annual meeting. In some cases, companies remain on the
CalPERS list for several years, while other companies are removed from the list
after agreeing to CalPERS' requests. For example, the list of ten worst in 1994
was reduced to just two companies (Dobrzynski, 1994a; and Anand, 1994).
FUTURE TRENDS AND ISSUES
CALPERS IS PART OF A WAVE OF INCREASING INSTITUTIONAL ACTIVISM which
shows no sign of abating. There appear to be several reasons for increased
investor activism by U.S. public pension funds (such as CalPERS). Given the
sheer size of many public pension funds, simply voting with their feet if they are
not satisfied with a particular equity investment is not always a viable option for
the fund managers since the cost of selling is quite high. Part of the increase in
investor activism can also be attributed to a 1988 U.S. Labor Department ruling
which stated that pension funds must vote the shares in their portfolio for the
exclusive benefit of plan members (rather than abstaining or automatically
voting with management). Numerous issues and trends will continue to shape
the evolution of the CalPERS model.
388
INSTITUTIONAL ACTIVISM BY PUBLI C PENSION FUNDS: THE CALPERS MODEL
The Media
Some of CalPERS' success in corporate governance issues is attributable to the
power and influence of the U.S. media. Recall that the average holding by
CalPERS is only around 0.6 percent of outstanding shares. However, its influence
is magnified by the media's attention to such actions as its annual "hit list"
and its recent letter regarding the GM principles. As long as CalPERS is able
to garner this public attention, it will continue to have influence.
New Screens
The key to CalPERS recent success appears to have been its ability to link cor-
porate governance issues directly with poor long-term performance by
companies. However, a recent additional criteria to the investment "screen" is
more controversial. A study for CalPERS by the Gordon group uncovered a
positive correlation between financial performance and workplace practices
such as job rotation, job training, and self-management teams (see a discussion
by Birchard, 1994). Time will tell whether this new screen is a natural extension,
attempting to link governance issues with performance, or whether it represents
a venture into social issues which may not relate to the "bottom line".
Indexing
Indexing — buying portfolios of stocks which replicate well-known indexes
such as the S&.P 500 - is still (relatively) far more prevalent in the United
States than in Canada. For pension funds, indexing raises an interesting issue
concerning the "prudence" responsibility of pension funds. Koppes & Reilly
(1994) argue that indexing simply for the purpose of diversification may not
be an adequate test of prudence, and pension funds may have a duty to monitor
actively such "passive" funds. They suggest that to screen the index based on
some form of corporate governance model may be more appropriate. This
relates to the previous issue of developing an appropriate screening technique
- one which is linked to the performance of the firm.
Global Diversification
As it increasingly diversifies internationally, CalPERS is taking a more active
role in governance issues in foreign countries. As of 1993, CalPERS had
investments of over $US 500 million in each of France, Germany, Italy, Japan,
Spain and the United Kingdom, and over $US 300 million in Australia,
Canada and the Netherlands. Mainly in the "study phase" now, CalPERS has
an international consultant and is examining the possibility of more participation
in international proxies.
389
FOERSTER
Relationship Investing
The most interesting prospect for CalPERS is an increased emphasis on relation-
ship investing which is not necessarily linked to corporate governance, but is
aligned in philosophy. Relationship investing is defined by CalPERS representa-
tives as "the conscious decision to own a large block of common or preferred
stock in a single corporation, coupled with a commitment to active management
over a long-term holding period" (Koppes & Reilly, 1994, pp.7-8). How the
notion of "active management" evolves will ultimately be the key to the success
or failure of this approach.
SURVEY METHODOLOG Y
G
IVEN AN UNDERSTANDING OF THE CALPERS MODEL, we are now in a position
to examine the administration of public pension funds in Canada. First, a
brief description of the survey methodology: between January and March 1995,
in-person interviews were conducted with two groups — executives and managers
(hereafter collectively referred to as fund managers) of a small number of major
Canadian public pension funds, including the Caisse de depot et placement du
Quebec, Ontario Teachers, OMERS and another large fund (which wished to
remain anonymous); and a representative of Fairvest Securities.
Fund managers were asked to respond to a series of questions dealing
with four general areas: their portfolio composition and equity investment
philosophy; their views on institutional activism in Canada; the CalPERS
model of institutional activism; and their advice to CEOs, directors, and
managers of Canadian companies. Many of the questions were intentionally
open-ended, allowing fund managers to express their views on a range of issues.
Given the concentrated nature of Canada's public pension funds, the
sample size is understandably quite small. Nonetheless, the fund managers who
participated in the survey were among the largest fund managers in Canada
and represented, collectively, over $90 billion in assets. According to a recent
survey of pension funds (Williams, 1994), these assets represent approximately
one-third of the total assets of Canada's 100 largest pension funds.
CANADIAN PUBLIC PENSION FUNDS
THE TEN LARGEST PUBLIC CANADIAN PENSION FUNDS, according to Benefits
Canada (Williams, 1994), were (in order): Ontario Teachers, Quebec Public
Employees (managed by Caisse de depot et placement du Quebec), OMERS,
Ontario Public Service, Alberta Public Sector, CN Railways, Ontario
Hospitals, Ontario Hydro, BC Municipal, and BC Public Service. Numerous
fund managers among this top-ten group were interviewed in order to compare
and contrast the degree of institutional activism in Canada relative to the
CalPERS model and to provide a basis for accepting or rejecting the four
390
INSTITUTIONAL ACTIVISM BY PUBLIC PENSION FUNDS: TH E CALPERS MODEL
hypotheses (mentioned earlier) concerning why Canadian funds tend not to
be as visibly active as CalPERS.
PHILOSOPHY AND STYLE
PHILOSOPHY AND STYLE WERE EXAMINED in order to determine whether certain
styles tended to be associated with a more active stance by fund managers.
Managers were presented with the following definitions of styles, as derived
from Macintosh (1993): "passive investment management" was defined as
selling an investment when dissatisfied with management of the company,
while "active investment management" was defined as attempting to alter
management's plan of action by voting against management, participating in a
proxy contest, or attempting to influence management by other means.
Managers were asked to rank their funds' philosophy on a scale of 1 to 5, with 1
representing "totally passive" and 5 representing "totally active". In addition,
managers were asked to describe their fund's equity investment style and
philosophy in general terms.
"Active" public pension funds displayed a wide variety of management styles.
Some funds were very definite in describing their own style in a particular
manner, for example, as being very much "bottom-up" or "value" oriented,
attempting to identify good companies in attractive industries. In contrast,
other funds employed a variety of styles, including indexing, growth-oriented,
value-oriented, small-cap, etc. Some funds viewed these multiple styles as a
form of diversification. In some cases, this variety of styles was achieved
through the use of external managers. There was no apparent relationship
between a fund's self-assessment of activism and the focus of its style.
Interestingly, some funds distinguished between their current approach to
activism (in the middle of the range) and their target (near 5), suggesting a
trend toward increased activism.
One fund raised an important distinction between reactive and proactive
activism. An example of reactive activism was voting against a poison pill
proposal - reacting to an issue raised by the management and/or board of a
corporation. In contrast, proactive activism involved approaching a corporation
and trying to enact change before an issue reached the point of a vote at an
annual meeting. One fund manager viewed most Canadian public funds — with
the exception of the Caisse - as engaging predominately in reactive activism.
Fund managers claimed to be long-term investors, yet were also very
cognizant of shorter-term performance.
Most funds claimed to be long-term investors - some even likened their style to
that of U.S. investor Warren Buffet, focusing on ownership rather than trading.
391
FOERSTER
Target holding periods often exceeded two years, and some cases were cited
where stocks were held for more than 20 years. Funds cited the high cost of
selling stocks, particularly those in the bottom 200 of the TSE 300 Index. This
reluctance to sell without creating a large price impact was also a reason cited
for the increased activism by the funds.
Paradoxically, most funds admitted to tracking their portfolio's perfor-
mance on a quarterly, monthly, and even daily basis. In addition, many fund
managers admitted to being compensated for beating the index (usually the TSE
300) over much shorter periods than their investment horizons. One fund
manager emphasized this dilemma between short-term and long-term horizons.
VIEWS ON REGULATION
MANAGERS WERE ASKED TO COMMENT ON THEIR VIEWS regarding any regulatory
restrictions, such as the Foreign Property Rule and the Pension Benefits Act.
Prior to 1990, funds were restricted in terms of the amount of foreign assets —
ten percent (of book value) - in which the fund could be invested without
incurring penalties. This amount was gradually increased to 20 percent in
1994- In contrast, there are no such restrictions in the United States. There
was even talk leading up to the February 1995 federal budget that foreign
property restrictions might be tightened again (to less than 20 percent).
In addition, prior to the early 1990s, funds covered under the Canada
Pension Benefits Standards Act (as well as many similar provincial Acts), were
restricted in terms of the types of eligible Canadian companies in which they
could invest. For example, funds were required to restrict investments to
companies which had paid dividends in four of the last five years, including
the most recent year. Recent changes in Canadian and most provincial Acts
now focus on the "prudence" test: for example, establishing and adhering to
investment policies, standards and procedures that a reasonable and prudent
person would apply in respect of a portfolio of investments to avoid undue risk
of loss and obtain a reasonable return. However, numerous Acts still restrict
the amount (percentage) of stock a fund can own.
While funds preferred less regulation to more, they did not envision a
complete removal of restrictions, e.g., related to foreign assets.
Fund managers viewed any constraints as restrictions on potential investment
returns and welcomed the move to the prudence test. Not surprisingly, funds did
not wish to see any additional restrictions, and some funds cited the strong
performance (generally) of Canadian pension funds as a reason for no additional
regulations.
The issue of the foreign property rule brought a wide range of responses
and often the strongest (i.e., most emotional) response of any topic discussed.
While most managers expressed a desire for an increased foreign asset level,
392
INSTITUTIONAL ACTIVISM BY PUBLIC PENSION FUNDS: THE CALPERS MODEL
there was also an opinion expressed by one fund that Canada has a right to
limit foreign investment of pension funds. However, another fund manager
called it an "absolutely foolish rule that flies in the face of what capital markets
are all about" and argued that placing walls around Canada will only serve to
impoverish our nation. This manager challenged the government to present a
solid economic rationale for the rule. One fund manager argued for the
removal of the rule by noting that the use of derivative securities makes the
rule a false barrier. Another fund manager referred to restrictions as ludicrous
and a form of foreign exchange controls that will only continue to result in
lost opportunities for Canadian pension stakeholders.
Some fund managers cautioned that legislators (and bureaucrats) are
always looking for perceived problems and solutions, while these policy solutions
were often worse than the symptoms. Other managers commented that cur-
rent restrictions were simply some bureaucrat's dream. One manager lamented
the fact that Canada did not have a single regulatory agency as in the United
States; and another manager praised the form of regulation that requires firms
to disclose more information to shareholders.
CORPORATE GOVERNANCE ISSUES
MANAGERS WERE PRESENTED WITH A LIST OF corporate governance issues and
asked to comment on their perceived importance, with 1 representing "not
important" and 5 representing "very important." Issues listed included poison
pills, executive compensation, independence of directors, staggered boards,
number of board members, and minority shareholder rights. As well, managers
were asked to comment on any additional corporate governance issues which
they felt were important. Results are summarized in Table 1.
TABLE 1
PERCEIVED IMPORTANCE OF CORPORATE GOVERNANCE ISSUES
BASED ON A SURVEY OF CANADIAN PUBLIC PENSION FUND MANAGERS
Independence of Directors 4-5
Poison Pill Plans 43
Executive Compensation 3.8
Minority Rights 3.7
Number of Board Members 3.3
Staggered Boards 3.0
Notes: 1 = not important.
5 = extremely important.
393
FOERSTER
The most important corporate governance issue was the perceived lack
of independence of many directors, followed by poison pill plans.
Coinciding with the draft of the TSE report on corporate governance, the
independence - or more specifically, the perceived lack of independence - of
directors of Canadian corporations was viewed as the major corporate gover-
nance issue. One fund manager singled out the major Canadian banks as an
example of the lack of truly independent outside directors.
The second most important issue dealt with poison pill plans.
Interestingly, some pension fund managers professed an evolution in thinking
since Inco implemented the first such pill in Canada in 1988. Many funds
routinely voted against poison pills as a matter of principle, but some have
recently sought improvements in pill plans in exchange for supporting such
plans. Others claimed that almost without exception they continued to vote
"no". One fund manager was surprised at the large number of planned poison
pills during 1995 after the recent first defeat of a poison pill (Labatt's).
Some fund mangers commented on a key difference in Canada versus the
United States: the relative prevalence of closely held firms. CalPERS screens
out most companies with at least 30 percent inside ownership, although
according to CalPERS representatives, in some cases companies with up to 40
percent inside ownership would be considered. The primary reason given for
this screen was to increase the chances of success for any shareholder resolutions.
In Canada, since many firms are closely held, an investigation was conducted
to determine the effect of this CalPERS screen on Canadian firms. The TSE
has recognized the prevalence of control blocks and has incorporated it into
the calculation of the benchmark TSE 300 index. The TSE Review provides a
list of these 300 stocks, including the "available float" on which the relative
weights on the TSE 300 composite index are calculated. As of January 1995,
among the TSE 300 companies, 48.0 percent had 100 percent available float
(i.e., non-control block shares), 58.7 percent had at least 80 percent available
float, 65.4 percent had at least 70 percent available float, 73.7 percent had at
least 60 percent available float, and 82.0 percent had at least 50 percent available
float. Thus, according to the CalPERS screen, between two-thirds and three-
quarters of TSE 300 companies would be eligible for targeting or, conversely,
one-quarter to one-third would not. Canadian public pension fund managers
recognized this reality, but many simply viewed it as something they were
required to live with. A related control issue deals with unequal or subordinate
voting rights shares. As a rule, most fund managers recommended voting
against any proposals that would authorize issues of new common stock that
had unequal or subordinate voting rights. Thus fund managers appeared to be
more concerned with ensuring that ownership concentration did not increase
beyond existing levels.
Fund managers also described other corporate governance issues which
they viewed as important. These included the perceived lack of planning in
many companies facing Chief Executive Officer succession.
394
INSTITUTIONAL ACTIVISM BY PUBLIC PENSION FUNDS: TH E CALPERS MODEL
Several fund managers indicated they maintained proxy voting guide-
lines which they viewed as public documents and sent to most large Canadian
corporations. Many of these guidelines were based strictly on the economic
impact of voting for or against a particular proposal.
COMPARISONS WITH CALPERS
REACTION WAS MIXED IN TERMS OF EVALUATING the CalPERS model of
activism. Some applauded the emphasis on long-term performance and its
proactive stance. The CalPERS approach was generally viewed as constructive,
but not the only viable approach.
Some suggested that targeting only a dozen firms and exerting pressure
for change would not have a major effect on the North American economy,
although perhaps it would have a small effect on CalPERS' performance.
There was concern that companies' fear of antagonism might cause them to
perceive shareholders as a nuisance or might create an "us versus them" mentality
- similar to management/union struggles - rather than both parties working
together to improve share value. Cases were cited of Canadian firms which
underperformed both the TSE and their industry, yet managers of these
companies were considered to be dedicated individuals.
Another manager suggested that up to half of the public pension funds
in Canada were reluctant to get involved in corporate governance issues. Fund
executives, rather than fund analysts, were most concerned about corporate
governance issues. Analysts were viewed as being more concerned with keeping
their contacts at various firms and did not wish to "rock the boat".
Several fund managers suggested that the best way to influence manage-
ment was through private conversations rather than through proxy statement
battles. Fund managers indicated they were usually just a phone call away from
CEOs — usually having calls returned within fifteen minutes — and this was
often an effective route. One fund manager cited a case of attempting to
approach management, but being rebuffed. After creating an alliance with
another group, the fund was able to attract the attention of the board and positive
changes were put in place. The fund manager noted that management took
full credit for the changes. Unlike many CalPERS cases, the confrontation
between institutional investors and management was not reported in the
media, and the fund manager was happy to let management take the credit so
long as it had a positive effect on the bottom line of the fund.
Nonetheless, fund managers also supported vigorous action on the part of
public funds in the (rare) case of real management abuses, provided the "quiet
persuasion" route has been exhausted. Examples were cited of public advertise-
ments which were placed in The Globe and Mail when fund managers wanted to
make public a particularly important point. However, these last resort measures
were in contrast to the annual public lists of their targeted companies supplied
by CalPERS. We can now re-examine the four hypotheses with respect to why
Canadian public pension funds are less visibly active than CalPERS.
395
FOERSTER
Canadian pension funds tend to engage in corporate governance activism
in a less visible fashion than major American counterparts such as
CalPERS primarily because of style differences.
The first possible explanation was regulatory differences between the two
countries. Fund managers generally did not view regulatory differences as that
great and not in areas that might impact on activism. The second possible
explanation was that, in Canada, third-party groups such as Fairvest or the
Pension Investment Association of Canada (PIAC) might be playing the
activism role. PIAC was viewed primarily as an educational vehicle, and while
Fairvest was viewed as providing a voice for funds, it was not seen as precluding
funds from taking an activist role. While the third explanation — the lead-lag
between the United States and Canada cannot be eliminated, the fourth
explanation appears to be most likely. Fund managers often differed in their
views concerning corporate governance issues, yet presented a consistent picture
of the "Canadian way" of dealing with management of companies — non-
confrontational, except in unique circumstances.
One fund manager used the analogy of owners (shareholders), directors,
and managers all in a boat together trying to navigate through difficult waters.
There were enough external factors trying to sink the boat and it would be easier
to survive if the three parties got along rather than screaming and shouting at
one another. Another fund manager stressed the importance of keeping low
visibility by referring to a fundamental law of physics: for every action, there is
an equal and opposite reaction. The more a fund visibly pressed for change, the
more resistant management would become, in order not to lose face in public. A
third fund manager simply described Canadian fund managers as being more
conservative and cautious than American counterparts.
ADVICE FOR CEOs
FINALLY, MANAGERS WERE ASKED TO COMMENT on what advice they would offer
to CEOs, directors, and managers of publicly-traded Canadian companies in
order to prepare them for possible increases in institutional activism. Many
fund managers viewed corporate managers in Canada as generally doing a
good job managing their companies. Some fund managers felt that institutions
were erroneously cast in the media as always being the "saints" while managers
were always the "sinners". Nonetheless, there were a few notable cases
(particularly in the United States) where management was clearly incompetent
and solidly entrenched.
Improve investor relations.
Fund managers felt corporate managers should be responsive to their share-
holders and listen to their ideas, particularly if shareholders have knowledge or
396
INSTITUTIONAL ACTIVISM BY PUBLI C PENSION FUNDS: THE CALPERS MODEL
experience to offer. An example cited was a firm which held an annual dinner
with representatives of its five largest shareholders. While the intent was
certainly not to obtain insider information, the exchange of ideas was viewed
as a positive experience.
One fund manager described the key as accountability of management
and boards and better communication among all three parties: shareholders,
managers and boards.
Improve the composition of boards.
The main issue appeared to be the lack of independence of board members.
Fund managers sought the removal of management-appointed members with
little company or industry knowledge or experience. In rare cases, fund managers
indicated they sought board representation themselves.
One fund manager stressed the importance of separating the CEO from
the board chair (a model more common in the United Kingdom than in most
other countries). The primary argument was that a bit of "tension" (i.e., the
CEO having to report to someone else) is a "good thing".
Balance the short-term and long-term prospects of the firm.
Fund managers recognized the dilemma faced by CEOs who were being pushed
by analysts to improve short-term performance, while at the same time trying
to build a long-term strategy. Some fund managers felt the pension funds
themselves should be more vocal in stating they are prepared to take a longer-
term outlook for a firm's prospects, giving boards the freedom to select good
managers and providing them with the time to succeed. Nonetheless, over the
longer term, this flexibility should be related to performance, or fund managers
will demand the removal of under-performing management. Fund managers
recognized that performance relative to a peer group (such as an industry) was
an appropriate way to judge a firm's performance.
SUMMARY AND CONCLUSIONS
T
HIS STUDY INVESTIGATED WHY PUBLIC PENSION FUNDS in Canada have not
been as visibly active as the California Public Employees' Retirement
System (CalPERS) in the United States. Primarily based on interviews with
major Canadian public pension fund managers as well as with CalPERS
representatives, this study provided a detailed framework of the CalPERS
model of investor activism, described how Canadian public pension fund
managers viewed this model, examined what corporate governance issues were
important to Canadian public pension fund managers, and provided corporate
directors with a sense of what to expect in the future in terms of investor
397
FOERSTER
activism. There are several lessons to be learned from this study: lessons for
pension fund managers, regulators and corporate managers.
The results of this study are not to suggest that all Canadian public
pension funds should try to be just like CalPERS, but rather that they should
benefit from many of the CalPERS' experiences, particularly the relatively
recent focus on company performance. Pension fund managers must do their
homework: know what questions they wish to ask managers and articulate what
it is they are looking for. They must do research on companies, especially
examining performance (total shareholder return) and any factors related to per-
formance. They must recognize that they are partners with managers and boards.
Regulators should resist any temptation to install additional regulations
which might restrict public pension fund ownership, and should, in fact, consider
removing regulations. Specifically, funds should be allowed to maximize
returns to their stakeholders based on desired risk levels and utilize an
unconstrained global opportunity set of securities. In addition, any regulations
that restrict public fund mangers from being effective monitors of corporations
should be removed. For example, public pension funds should be allowed to
own a larger stake in any company in which they invest.
Corporate managers should be concerned with improving shareholder
return, and should recognize emerging evidence linking performance with
corporate governance issues. In particular, corporate managers should consider
recommendations — made by public pension funds as well as others - for making
boards more effective. Communication with shareholders should also be
strengthened.
While governance issues will continue to dominate the concerns of
public pension funds (and other institutional investors), and while these funds
will become increasingly active and continue to search for ways to improve
the long-term performance of Canadian firms, they will do so in a less visible
fashion than major American counterparts such as CalPERS. Nonetheless,
Canadian pension funds have much to learn from the U.S. experience, as do
Canadian managers and regulators.
398
INSTITUTIONAL ACTIVISM BY PUBLIC PENSION FUNDS: THE CALPERS MODEL
ACKNOWLEDGEMENTS
T
HE FINANCIAL SUPPORT PROVIDED by the Financial Research Foundation of
Canada is gratefully acknowledged. I also wish to acknowledge with
thanks the research assistance of Rod Graham and Dave Thomas. My thanks
to Carl Bang for useful comments and to the discussants and seminar participants
at Industry Canada's Conference on Corporate Decision-Making in Canada. I
also wish to thank Karin Estes and Maureen Reilly from the California Public
Employees' Retirement System and representatives of the Canadian public
pension fund industry (some of whom wished to remain anonymous) for agreeing
to be interviewed for this project. Any errors are solely attributable to the author.
BIBLIOGRAPHY
Anand, V. "CalPERS Gunning for Poor Performers." Pensions & Investments. (January 24,
1994):4,80.
Birchard, B. "The Call for Full Disclosure." CFO. (December 1994):31-42.
The California Public Employees' Retirement System. "Why Corporate Governance?"
(November?, 1989):1-21.
Cordtz, D. "Corporate Hangmen." Financial World. 162, 7 (March 30, 1993):24-8.
Chernoff, J. "Institutions Strengthen Grip on Companies." Pensions & Investments. (March
8,1993):!, 34.
Chief Executive. "The New Governance Paradigm." (April 1994):40-54.
Dobrzynski, J. H. "Cool it! It's the CalPERS Cops." Business Week. (January 31, 1994a):6.
. "An Inside Look at Calpers' Boardroom Report Card." Business Week. (October
17, 1994b):186~88.
The Economist. "Reluctant Owners." (January 29, 1994):16-7.
Gordon L. A. and J. Pound. "Active Investing in the U.S. Equity Market: Past Performance
and Future Prospects." The California Public Employees' Retirement System. (January
ll,1993):l-43.
Hanson, D. M. "Much, Much More than Investors." Financial Executive. (March/April
1993):49-51.
Koppes, R. H. and M. L. Reilly. "An Ounce of Prevention: Meeting the Fiduciary Duty to
Monitor an Index Fund through Relationship Investing." Working Paper, The
California Public Employees' Retirement System. (March 25, 1994):l-56.
Macintosh, J. "If it Ain't Broke...." Canadian Investment Review. (Winter 1994/95):37-9.
Macintosh, J. G. "The Role of Institutional and Retail Investors in Canadian Capital
Markets." Osgoode Hail Law journal, 31, 2 (1993):371-472.
Mackenzie, W. "Corporate Governance News." Corporate Governance Review, 9, 1
(December 1994/January 1995):10-12.
Montgomery, K. E. "Shareholder Activism in Canada: A Survey of Institutional
Shareholders." National Centre for Management Research and Development Working
Paper NC 92-014-B. Western Business School, University of Western Ontario, 1992.
399
FOERSTER
Montgomery, K. E. and D. S. R. Leighton. "The Unseen Revolution is Here." Business
Quarterly. (Autumn 1993):39-46, 48.
Murphy, K. J. and K. V. Nuys. "State Pension Funds and Shareholder Inactivism." Working
Paper. Harvard Business School, 1994.
Nesbitt, S. L. "Rewards from Corporate Governance." California Public Employees'
Retirement System. (February 12, 1992):l-5.
. "Long-term Rewards from Shareholder Activism: A Study of the CalPERS
Effect." journal of Applied Corporate Finance, 6 (1994): 75-80.
Pare, T. P. "Two Cheers for Pushy Investors." Fortune. (July 30, 1990):95, 98.
Pound, J. "Beyond Takeovers: Politics Comes to Corporate Control." Harvard Business
Review. (March-April 1992):83-93.
Pozen, R. C. "Institutional Investors: The Reluctant Activists." Harvard Business Review.
(January-February 1994):140-49.
Silverstein, K. "Pension Funds Increase Presence in Corporate Boardrooms." Pension World.
(May 1994):4-5.
Star, M. G. "Measuring up in Canada." Pensions & Investments. (April 18, 1994):22.
Stewart, T. A. "The King is Dead." Fortune. (January 11, 1993):34-7,40.
Stoffman, D. "Look who's Calling the Shots." Canadian Business. (July 1990):45-7.
Thain, D. "The TSE Corporate Governance Report: Disappointing." Business Quarterly.
(Autumn 1994):77-86.
Useem, M., Bowman, E. H-, Myatt, J. and C. W. Irvine. "US Institutional Investors look at
Corporate Governance in the 1990s." European Management Journal, 11,2 (1993):175-89.
Wahal, S. "Pension Fund Activism and Firm Performance." Working Paper, The Kenan-
Flagler Business School. The University of North Carolina, 1995.
Wharton, C., J. Lorsch and L. Hanson. "Advice and Dissent: Rating the Corporate
Governance Compact." Harvard Business Review. (November-December 1991):136-43.
Williams, P. "Benefits Canada Top 100 Pension Funds." Benefits Canada. (April 1994):41-61.
The Working Group on Corporate Governance. "A New Compact for Owners and
Directors." Harvard Business Review. (July-August 1991): 141-43.
Ybarra, M. "Money Talks." California Lawyer. (February 1995):50-5, 90.
400
Ben Amoako-Adu & Brian F. Smith
School of Business and Economics
Wilfrid Laurier University
11
Outside Financial Directors and
Corporate Governance
INTRODUCTION
A
S CORPORATE LONG-TERM STRATEGISTS FOCUS THEIR ATTENTION on global
competition, environmental concerns and corporate restructuring, the
role and effectiveness of the board of directors in positioning the company to
meet these challenges is being re-examined. Weitzer (1991) states that the
goal of corporate governance is to realign the interests of shareholders, managers
and other stakeholders of the corporation so as to enhance the long-term
growth of the company. Directors should, therefore, represent stakeholders'
interests, monitor the actions of management and provide advice, which
should enhance corporate value (Gillies, 1992).' Mace (1972) argues that the
role of corporate boards is not well understood and that there appears to be a
gap between what boards are expected to do and what they actually do.
Although he states that boards provide advice, exercise discipline on executives
and act in crisis periods, his interviews with board members reveal that CEOs
exert significant influence on their boards.
In order to offset any undue influence of management and to represent
the interests of shareholders, it is argued that a minimum number of outside
directors
2
 should be on the board. Outside directors are generally considered to
be more independent of management than inside directors who are usually
senior officers of the firm. This argument is reflected in both federal and
provincial corporate legislation that requires a minimum number of outside
directors on the boards of public corporations.
3
 Furthermore, a major recom-
mendation of the draft report of the Toronto Stock Exchange Committee on
Corporate Governance in Canada (1994) was that a majority of directors on
the boards of listed companies should be unrelated.
4
Another benefit to the firm of outside directors is the expertise they can
bring to the board which may not be present among officers of the firm. For
example, directors from the financial sector may bring special knowledge with
respect to underwriting of securities, bank financing, takeovers and/or corporate
restructuring; they are also expected to have extensive knowledge of other
industries and the national economy.
401
AMOAKO-ADU & SMITH
The degree of independence and net benefit to the firm of outside directors
has been disputed. First, it is argued that outside directors tend to act in
management's interests because the selection of directors is influenced by
senior executives and these outside directors generally hold few shares.
Second, outside directors are often investment dealers, pension fund managers,
bankers and insurers who may face a potential conflict of interest given an
existing or prospective business relationship between the financial institution
and the firm. Even if they have no current business dealings with the firm,
outside financial directors may not wish to challenge senior management for
fear of losing future business. In addition, since their companies take invest-
ment positions in stocks as well as provide financial advice to clients, there is
a potential conflict of interest that inside information from the corporate
boardroom may be used for the financial institution's direct gain or to counsel
their investment clients. Mace (1986) argues that financial directors do not
have a monopoly on financial expertise and, hence, their presence on a board
should not necessarily be value-enhancing.
Shleifer & Vishny (1986) contend that large shareholders may be more
effective at monitoring management and, if necessary, they can initiate
takeovers. Such large shareholders may be family trusts, institutional or corporate
investors who are represented on the board. The implication of this argument
is that insofar as the marginal benefit of concentrated ownership outweighs
the marginal cost of entrenchment, there should be an economic incentive for
large shareholders to increase their ownership. This would, therefore, lead to a
concentrated corporate ownership structure as found in Canada. However,
whether concentrated ownership is beneficial or injurious to the Canadian
economy is an unresolved empirical question.
The primary objective of this study is to evaluate empirically the
effectiveness of outside financial directors of a large sample of non-financial
Canadian public corporations.
5
 After controlling for other determinants, three
different measures of corporate performance are related to the proportion of
outside financial directors to ascertain whether the presence of such directors on
the board enhances corporate value. Outside financial directors have been
classified in previous research as "grey area" directors.
6
 As discussed in Byrd &
Hickman (1992a), grey area directors have some type of affiliation with the
firm that may limit their ability to challenge management. In addition to out-
side financial directors, lawyers, consultants, customers, suppliers and those
having transactional or familial ties to management have also been labelled
grey area directors.
A secondary goal of this study is to examine the effect of directors in the
same line of business. We expect such directors to add value because of their
special knowledge of the industry in which they work. As in the case of out-
side financial directors, there are factors that offset the benefits related to
directors in the same line of business. For example, they may face a conflict of
interest where the firms for which they act as managers and directors are
402
OUTSIDE FINANCIAL DIRECTORS AND CORPORATE GOVERNANCE
competing for the same customers. In addition, the expertise that directors
from the same line of business can offer may be available from within the firm
or from industry consultants.
INDEPENDENT AND AFFILIATED OUTSIDE DIRECTORS
T
HE NEED FOR REQUIRING CORPORATE BOARD composition to include out-
side directors is based on the assumption that outside directors will be
independent of management or the corporation and, hence, will bring
independent thinking and objectivity into corporate boardroom discussions
or debates. In line with this thinking, the Canada Business Corporations Act
requires that public corporations appoint at least two directors who are not
employees of the corporation or its affiliates. On the other hand, the Ontario
Business Corporations Act requires that a minimum of one-third of the directors
of a public corporation be outside directors.
For outside directors to be effective in monitoring and advising manage-
ment and (at times) voting against management on issues deemed not to be in
the interest of stakeholders, they must be independent of management or the
company. This means affiliated or related directors appointed from outside the
corporation may be ineffective in monitoring management performance
because of their links with management or the corporation. Such affiliated or
captive directors, also referred to as grey area directors, include individuals
who are linked to the corporation or management through either transactional
or familial ties. Thus, the corporation's consultants, creditors, underwriters and
directors related to management through blood or marriage are considered
affiliated directors. Any outside director who faces a potential conflict of
interest may be categorized as an affiliated or grey area director. Such potential
conflict of interest will impair the effectiveness of such outside directors. Mace
(1986) goes to the extent of suggesting that securities underwriters should not
be allowed to serve as outside directors because of the potential conflict of
interest. The importance of this issue underscores the recent TSE Committee
on Corporate Governance in Canada's (1994) guideline that a majority of the
board members should be unrelated or independent directors and also that the
board's nominating committee should be comprised exclusively of unrelated
outside directors. At present, however, corporate disclosure of the background
of outside directors is not sufficient to provide us with the data needed for any
analysis of the effectiveness of affiliated or related directors.
METHODOLOGY
I
F OUTSIDE FINANCIAL DIRECTORS AND DIRECTORS in the same line of business
are effective in their role as representatives of stakeholders and provide
unique advice and monitor the performance of management, then their
presence on corporate boards should be value-enhancing. On the other hand,
403
AMOAKO-ADU & SMITH
if these outside directors do not hold a monopoly on the information they
provide to the board or if they face a conflict of interest in their role as
bankers, underwriters and competitors, then corporate performance may not
be enhanced with more of these outsiders on the board.
Two methodological research approaches are used in assessing the effec-
tiveness of outside directors in adding value to the corporation. One approach
is to make indirect inference from stock market reaction to events surrounding
board of directors' decisions, such as the appointment of outside directors
(Rosenstein & Wyatt, 1990; and Byrd & Hickman, 1992). The second method
is to use a multiple regression to relate directors' attributes to corporate perfor-
mance measures after controlling for other relevant variables which may affect
performance (Baysinger & Butler, 1985; Morck, Shleifer & Vishny, 1988 and
Hermalin & Weisbach, 1991). Because of an inadequate number of common
board-related events we chose the regression methodology for this study.
The null hypothesis being tested is that the presence of outside financial
directors and directors in the same line of business on a board should not be
associated with incremental corporate value different from that of other outside
directors. Since Morck et ai. (1988) show that there is an empirical relation-
ship between ownership and corporate performance, the test will also control
for ownership.
The directors of the firms in the sample were classified into four groups:
i) inside directors, ii) outside directors from financial institutions, iii) directors
in the same line of business, and iv) other outside directors. Inside directors
are employees and officers of the firm or related companies or shareholders
who control at least 10 percent of the votes of the firm. Financial institutions
include banks, trust companies, investment dealers, pension funds and
insurance companies. Personal investment companies are not classified as
financial institutions because these firms are usually a tax-incentive alternative
to owning shares directly. Identification of such firms was somewhat subjective
because of a lack of detailed disclosure.
The financial institution directors are subclassified into a narrow definition
(N) which includes only employees of financial institutions and a broad
definition (B) which includes both employees and directors of financial institu-
tions. Directors in the same line of business are those who are either directors or
officers of firms in the same industry. Other outside directors are all directors who
are not insider, financial institution or same-line-of-business directors.
Cross-sectional regressions are used to examine the relationship between
corporate value and outside financial directors and directors from the same
line of business, respectively.
7
 The data for the regression are averaged over a
ten-year period, from 1984 to 1993. In order to capture the expected long-
term effects of many board decisions, a ten-year period of analysis is required.
Three different performance measures are used in this study. The first is
the ratio of the year-end market value of common equity to the book value of
common equity. This performance measure serves as a proxy for Tobin's Q
404
OUTSIDE FINANCIAL DIRECTORS AND CORPORATE GOVERNANCE
measure of performance which has been used in U.S. studies. The denominator
of Tobin's Q is the replacement cost of assets, but in Canada data are not
available on replacement cost of corporate assets. The market value of common
equity equals the number of common shares outstanding multiplied by the
market price at year end.
The other performance measures - return on assets and asset turnover -
rely exclusively on accounting figures. The return on assets is calculated as
earnings before interest and taxes divided by total assets. Asset turnover,
defined as total sales divided by total assets, is considered to be devoid of
accounting deductions, such as depreciation, and research and development
expenses, which tend to distort net profit positions of corporations.
The annual variables are averaged over the number of years for which
data are available. The cross-sectional regression used for the subsequent tests
follows this general functional form.
8
Performance = f (Control Variables, Outside Financial
Directors, Same Line of Business,
Other Outside Directors) (1)
The control variables include a measure of size (log of market value of
equity), systematic risk (beta) and ownership measured as percentage of stock
holdings of all directors.
9
 In the financial literature, insider ownership of shares
is suggested as one solution to the principal-agent problem. Executive ownership
in the company is expected to enhance the alignment of the interest of the
executives and shareholders and hence reduce agency costs. Morck, Shleifer &
Vishny (1988) show that the relationship between corporate value and owner-
ship is non-monotonic so ownership is delineated into three parts: zero to 5 per-
cent, 5 percent to 20 percent, and more than 20 percent holdings. We used 20
percent as the threshold because the Ontario Securities Commission allows
cumulative share acquisition up to 20 percent of votes (bright line) before a
takeover offer must be extended to all shareholders. It should be noted that for
the U.S. study, Morck et al. (1988) used 5 percent and 25 percent as the
thresholds, after using a piecewise regression to determine the slopes. Following
the argument put forward by Morck et al., we expect ownership of zero to
5 percent (Own 5) to be positively related to stock value, ownership between
5 percent and 20 percent (Own 20) to be negatively related to value, and own-
ership in excess of 20 percent (Own 20+) to be positively related to value.
10
 The
regression specification for the first performance measure is
MVB = a
0
 + a j Log (Equity) + a
2
 (Beta) - a
3
 (Regulated) + a
4
 (Own 5)
+ a
5
 (Own 20) + a
6
 (Own 20+) + a
7
 (Fin (N)) + a
8
 (Same)
+ a
9
 (Other Outside) + e-
t
 (2)
405
AMOAKO-ADU & SMITH
MVB =
Equity
Beta =
Regulated
Own 5 =
Own 20 =
Own 20+ =
Fin (N) =
Fin (B) =
Same =
Other Outside =
ratio of market-to-book value of equity
market value of common equity (a measure of size)
Dimson (1979) adjusted systematic risk coefficient with a
lead and a lag"
Dummy variable is 1 if regulated and zero otherwise
ownership by directors between zero and 5%, and
5% if ownership is > 5%
zero if ownership is < 5%,
ownership -5% if ownership is > 5% and < 20%,
and 15% if ownership is > 20%
zero if ownership is < 20%, and
ownership -20% if ownership is > 20%
Narrow definition of financial directors. This is the
percentage of directors who are current employees of
financial institutions.
Broader definition of financial directors. These
include Fin (N) and percentage of other directors who are
just directors of financial institutions.
Percentage of outside directors who are directors or
officers of firms operating in the same industry.
Percentage of outside directors, other than those from the
financial industry defined in Fin (B), and those defined as
same-line-of-business (Same). This includes politicians,
academics, public sector employees, consultants
and lawyers.
The cross-sectional regression uses market value of equity to book value
of equity, a proxy for Tobin's Q, as a performance measure.
Tobin's Q was the performance measure used by Morck et al (1988) and
Hermalin & Weisbach (1991). However, we modify this approach to control
for both size, regulation and risk. While the sign of the size and regulated
variable cannot be determined a priori, the relationship between value and risk
is expected to be negative. In order to control for risk, we include beta in the
cross-sectional regression and expect 112 < 0 because higher risk should be
inversely related to value. From Morck et al. (1988), we expect n
4
 > 0, n
5
 < 0
and n^ > 0. If financial and same-line-of-business directors are effective in
increasing the value of the corporation, then ny > 0 and rig > 0. If the other
outside directors also add significant value to the corporation then n
9
 > 0.
DATA
T
HE MONTHLY STOCK PRICES AND SHARES outstanding were obtained from
the Canadian Financial Markets Research Centre stock database.
Corporate financial data, such as book value of common equity, were taken
406
OUTSIDE FINANCIAL DIRECTORS AND CORPORATE GOVERNANCE
FIGURE 1
AVERAGE COMPOSITION OF BOARD OF DIRECTORS OF 150 TSE LISTED COMPANIES
(1984-1993)
Same Line of
Business 6.4%
Other Outside
Directors
3
 38.9%
Inside
Directors 40%
Directors of 7 Officers of Financial
Financial Institutions 6.3% Institutions 8.4%
Note:
 a
 Other outside directors include consultants, lawyers, former politicians, academics, and employees from
the public sector or non-profit organizations.
from the Compact Disclosure Canada CD-Rom Database as well as Microfiche
copies of the annual reports. The names, positions and professional back-
ground of the directors of 200 of the Financial Post 500 public corporations
each year were compiled and coded into a workable database from the share-
holder proxy circulars and the Compact Disclosure Canada CD-Rom database.
The names of all shareholders holding more than 10 percent of the votes were
also taken from these two sources. From an initial sample of 200 public
companies, the final sample reduced to 150 because of either missing financial,
ownership or director information, or insufficient monthly stock returns for
the estimation of the beta. Five years of monthly stock returns were used to
estimate the betas but in cases where there were not enough monthly returns,
at least 24 months of stock returns were required.
BOARD COMPOSITION AND SAMPLE CHARACTERISTICS
F
IGURE 1 SHOWS THE COMPOSITION OF CORPORATE BOARDS in the sample for
the full period from 1984 to 1993. For the sample of 150 Canadian public
corporations listed on the Toronto Stock Exchange over the period from 1984 to
1993, inside directors constitute 40 percent of the board and all outside directors
comprised 60 percent. These proportions compare with 34-4 percent and 65.6 per-
cent of insiders and outsiders, respectively, reported in Rosenstein & Wyatt (1990)
for the United States. Thus, in Canada, about 60 percent of board directors are
outsiders while in the United States about 66 percent are outside directors.
407
AMOAKO-ADU & SMITH
TABLE 1
SAMPLE CHARACTERISTICS OF 150 CANADIAN COMPANIES, 1984 TO 1993
VARIABLE
STANDARD
MEAN DEVIATION MINIMUM MAXIMUM
Equity Beta
Market-to-Book Value of Equity
Proportion of Inside Directors
Proportion of Outside Directors
Proportion of
Financial Directors (N)
a
Proportion of
Financial Directors (B)b
Proportion of Directors
from the Same Line of Business
0
Market Value of Equity ($ millions)
Total Assets ($ millions)
Operating Profit Margin^ (%)
Return on Assets
6
 (%)
Asset Turnover'
0.937
1.416
0.400
0.600
0.084
0.147
0.064
710
,700
10.08
7.81
1.30
0.379
1.042
0.175
0.175
0.083
0.128
0.091
1,380
3,889
12.54
4.75
0.90
0.066
0.240
0.068
1.00
0.00
0.00
2.132
6.184
1.00
0.932
0.350
0.627
0.00 0.500
2.7 11,200
33.6 38,998
-1.59 80.70
-4.21 24.55
0.08 5.80
Notes:
 a
 Financial Directors (N) are narrowly defined as outside financial directors who are employees of
financial institutions.
Financial Directors (B) are broadly defined to include both outside financial directors who are
either employees or directors of financial institutions.
c
 Directors from the same line of business is defined as outside directors who are current CEOs or
employees, retired employees or directors of the same line of business.
Earnings before interest and taxes divided by sales.
e
 Earnings before interest and taxes divided by total assets.
Asset turnover is defined as total sales divided by total assets.
Using the broader definition of outside financial directors, about 15 per-
cent of the directors are either employees or directors of financial institutions.
These proportions are reported in Table 1. Thus, one-seventh of all directors
are associated either directly or indirectly with financial institutions. This is a
significant proportion considering the limited number of financial institutions
in Canada. For a sample of U.S. firms, Rosenstein & Wyatt (1990) find that
financial directors represent 10.7 percent of all directors. For our study, only
8.4 percent of the directors were current employees of financial institutions.
Directors from the same line of business or industry comprised 6.4 percent of
the board.
Table 1 indicates that the mean equity beta for the sample of companies
is 0.937. This low average equity beta corresponds to the fact that the sample
408
OUTSIDE FINANCIAL DIRECTORS AND CORPORATE GOVERNANCE
TABLE 2
CHARACTERISTICS OF BOARDS OF DIRECTORS OF 150 CANADIAN PUBLIC COMPANIES,
1984 TO 1993
STANDARD
VARIABLE MEAN DEVIATION MINIMUM MAXIMUM
Number of Directors
per Company 11.25 3.88 5.0 25.0
Percentage of Votes Held
by Inside Directors 40.75
3
 30.84 0.01 100.00
b
Percentage of Companies
where CEO is Chair
c
 26.70 n/a n/a n/a
Tenure of Directors in Years 8.60 4.15 0.20 27-60
Notes:
 a
 The percentage of votes held by inside directors is high for our sample of Canadian firms in relation
to that reported in studies of U.S. companies. We attribute the higher percentage to the closer
ownership of Canadian firms.
" The company which had directors holding 100.00 percent of votes between 1984 and 1993 had
dual class equity. The voting shares were held by insiders whereas the non-voting shares were held
mainly by outsiders.
c
 The percentage includes all companies that had a chief executive officer who also served as chair
of the board of directors,
n/a Not applicable.
represents the largest public corporations in Canada. These firms tend to have
relatively low systematic risk. The mean market value of equity and total
assets are $710 million and $1.7 billion, respectively.
The three performance measures have a wide range of values for the
sample of companies even though they represent an average across a number
of years. The market-to-book value of equity varies from 0.240 to 6.184.
Return on asset and total asset turnover range from —4-21 percent to 24.55
percent and 0.08 to 5.80, respectively. The firms with the highest operating
profitability tend to be utilities which have a high level of offsetting interest
expense. These utilities also have the lowest asset turnover.
As shown in Table 2, the number of directors per company ranges from
five to twenty five. On average, over the period from 1984 to 1993, there were
11.25 directors per company, of which about 6.9 were outside directors. The
Ontario Business Corporations Act (section 115), and the corresponding Federal
Act (section 97) requires that at least one-third of the board or two directors,
respectively, should be outside directors. Thus, large Canadian public corpora-
tions tend to have more outside board members than is legally required.
Table 2 shows the percentages of votes held by all directors. The mean
total votes held by directors for each company is 40.75 percent with the
minimum and maximum held being 0.01 percent and 100.0 percent, respectively.
Byrd & Hickman (1992) report that the average vote ownership by directors
409
AMOAKO-ADU & SMITH
of a sample of U.S. firms was 15 percent. The 100 percent voting control was
for Xerox Canada which issued dual-class shares. The directors of Xerox
Canada held all voting shares and outside investors owned the non-voting
shares. A further analysis of the data in the sample suggests that most of the
votes controlled by directors are owned by a few inside directors. These data
indicate that although the proportion of outside directors on the board is larger,
relative to insiders, only a small fraction of the company's votes is held by out-
side directors.
In 26.7 percent of the firms, the chief executive officer (CEO) is the
chair of the board. The potential for undue influence on board decisions of
this dual role has been raised as a concern in the literature on corporate gover-
nance. The mean and maximum tenures of directors in our study are 8.6 years
and 27.6 years, respectively. This indicates substantial board experience
among directors of Canada's largest public companies.
ANALYSIS OF REGRESSION RESULTS
T
HE REGRESSION RESULTS OF THE TEST using three different performance
measures as the dependent variable are presented in Tables 3 through 5. In
column one of each table, the results indicate that the proportion of total out-
side directors is not significantly related to corporate performance. The
absence of a significant relationship between outside board composition and
corporate performance is consistent with the findings of Hermalin &
Weisbach (1991). When the composition of outside directors is analyzed,
neither the proportion of directors who are employees from the financial sector
nor the proportion of directors broadly defined to include financial sector
employees and directors is consistently related to the performance measures at
the 5 percent significance level. Only the proportion of financial directors
defined broadly (B) and market-to-book value of equity are related at the 10
percent significance level. In addition, other outside directors do not have a
significant relationship with any of the performance measures. With one
exception, this is also the case for the same-line-of-business directors. Asset
turnover has a significant negative relationship with the proportion of directors
from the same line of business. This implies that companies with more directors
from the same industry generate less sales per dollar of assets.
The overall results indicate that the experience of outside directors
classified by involvement in the financial sector and the same line of business
does not consistently add more value than that provided by other directors.
This contrasts with Rosenstein & Wyatt (1990) who find that the appointment
of outside directors from the financial sector increases shareholder wealth. In
addition, there is no incremental value to having a higher proportion of other
outside directors on the boards.
The results of the control variables indicate that the systematic risk
variable, beta, is positively related to market-to-book value ratio. This implies
410
OUTSIDE FINANCIAL DIRECTORS AND CORPORATE GOVERNANCE
TABLE 3
REGRESSION RESULTS OF PERFORMANCE AND ITS DETERMINANTS
DEPENDENT VARIABLE: MARKET VALUE OF EQUITY-TO-BOOK RATIO
3
CONSTANT -0.403
(0.36)
-0.130
(0.12)
-0.188
(0.17)
0.043
(0.04)
Log (Equity)
Beta
Regulated
Ownership" 5
Ownership 20
Ownership 20+
Outside Directors
Financial Directors (N)
c
Financial Directors (B)^
Same Business Directors
Other Outside Directors
Adjusted R
2
Number of Companies
0.124
(1.02)
0.467
(1.99)**
0.341
(1.21)
7.462
(0.65)
-3.125
(0.86)
0.459
(0.95)
0.277
(0.673)
0.128
(1.05)
0.444
(1.87)*
0.381
(1.33)
7.870
(0.68)
-3.617
(1.01)
0.440
(0.90)
0.516
(0.45)
0.091
(0.74)
0.467
(2.00)*
0.382
(1.35)
7.212
(0.17)
-3.031
(0.85)
0.505
(1.05)
0.02
138
-0.139
(0.25)
0.01
138
1.324
(1.66)*
-0.220
(0.38)
0.03
138
0.125
(1.02)
0.453
(1.92)*
0.381
(1.33)
9.002
(0.77)
-4.058
(1-11)
0.368
(0.76)
-0.485
(0.45)
-0.299
(0.54)
0.01
138
Notes: Numbers in parentheses are t-statistics.
* Coefficient is statistically significant at the 10% level.
** Coefficient is statistically significant at the 5% level.
a
 The average annual ratio of the market capitalization of common equity to the book value of
common equity.
Ownership is defined as the percentage of corporate votes held by inside and outside directors.
c
 Financial Directors are narrowly defined as outside financial directors who are employees of
financial institutions,
d Financial Directors are broadly defined as outside financial directors who are either
employees or directors of financial institutions.
411
AMOAKO-ADU & SMITH
TABLE 4
REGRESSION RESULTS OF PERFORMANCE AND ITS DETERMINANTS
DEPENDENT VARIABLE: RETURN ON ASSETS
3
CONSTANT
Log (Equity)
Beta
Regulated
Ownership" 5
Ownership 20
Ownership 20+
Outside Directors
Financial Directors (N)
c
Financial Directors (B)^
Same Business Directors
Other Outside Directors
Adjusted R
2
Number of Companies
0.106
(2.04)**
-0.003
(0.49)
-0.011
(0.93)
0.026
(2.08)**
-0.078
(0.15)
-0.017
(0.10)
0.032
(1.41)
-0.000
(0.01)
—
—
—
—
0.02
130
0.090
(1.77)*
-0.003
(0.53)
-0.012
(1.08)
0.026
(2.02)**
-0.123
(0.24)
-0.005
(0.033)
0.038
(1.67)*
—
0.067
(1-25)
—
—
0.032
(1.27)
0.03
130
0.094
(1.86)*
-0.003
(0.48)
-0.01 1
(0.97)
0.025
(1.92)*
-0.065
(0.12)
-0.013
(0.08)
0.034
(1.49)
—
—
0.005
(0.13)
—
0.024
(0.87)
0.02
130
Notes: Numbers in parentheses are t-statistics.
* Coefficient is statistically significant at the 1 0% level.
** Coefficient is statistically significant at the 5% level.
a
 The average annual ratio of the market capitalization of common equity to the book
0.111
(2.11)**
-0.003
(0.57)
-0.010
(0.93)
0.026
(2.01)**
0.011
(0.02)
-0.057
(0.34)
0.029
(1.27)
—
—
—
-0.060
(1.13)
0.011
(0.42)
0.03
130
value of
common equity.
Ownership is defined as the percentage of corporate votes held by inside and outside directors.
c
 Financial Directors are narrowly defined as outside financial directors who are employees of
financial institutions.
Financial Directors are broadly defined as outside financial directors who are either
employees or directors of financial institutions.
412
OUTSIDE FINANCIAL DIRECTORS AND CORPORATE GOVERNANCE
that growth companies tend to be more risky. Tables 4 and 5 indicate that
regulated firms behave differently from non-regulated firms when considering
corporate governance issues. In Table 4, the results show that regulated firms
tend to perform better than non-regulated firms if return on assets is used as a
performance criterion, while Table 5 indicates that regulated companies have
lower asset turnover.
The size of the company, measured by the log of market value of equity,
is statistically significant as a control variable in the case where the dependent
variable was measured as asset turnover. Larger companies tend to have lower
asset turnover. This is not surprising, given that many of the larger companies
in the sample were utility companies, such as Bell Canada Enterprises.
In contrast to Morck et al. (1988), ownership up to 5 percent and owner-
ship between 5 percent and 20 percent is not significantly related to perfor-
mance. However, ownership in excess of 20 percent is found to be significantly
and positively related to corporate performance, measured as asset turnover or
return on assets. This provides weak support for the argument that a high
concentration of ownership more closely aligns the interest of shareholders
and management and, hence, concentrated ownership makes companies
perform better.
Generally, the results indicate that there is no convincing evidence of
significant incremental value associated with the presence on the board of out-
side directors from the financial industry and outside directors in the same line
of business. This is consistent with the findings of Hermalin & Weisbach (1991)
for outsiders in general. The results also indicate that the non-monotonic
relationship between ownership and corporate value as postulated and observed
by Morck et al. (1988) holds only for ownership levels above 20 percent.
CONCLUSIONS AND POLICY IMPLICATIONS
I
N THIS STUDY, CANADIAN CORPORATE DATA over the period from 1984 to
1993 were used to test whether the presence on corporate boards of outside
financial directors and directors from the same line of business is value-
enhancing. In other words, is there a statistically positive relationship between
these outside directors and corporate performance? The study also examines
the significance of other determinants of corporate performance, such as
systematic risk and size as well as the non-monotonic relationship between
board ownership and corporate performance.
Generally, the results suggest that there is no statistical evidence of a
relationship between corporate performance and proportion of outside directors
on the board. The findings for this large sample of Canadian firms are consistent
with those reported by Hermalin & Weisbach (1991) in the United States. In
addition, when the composition of the board is analyzed, there is no convincing
evidence that outside directors from the financial industry or from the same
line of business enhance the value of the corporation. The results, however,
413
AMOAKO-ADU & SMITH
TABLE 5
REGRESSION RESULTS OF PERFORMANCE AND ITS DETERMINANTS
DEPENDENT VARIABLE: ASSET TURNOVER
3
CONSTANT 5.345
(6.79)**
5.168
(6.69)**
5.201
(6.75)**
5.786
(7.39)**
Log (Equity)
Beta
Regulated
Ownership " 5
Ownership 20
Ownership 20+
Outside Directors
Financial Directors (N)
c
Financial Directors (B)^
Same Business Directors
Other Outside Directors
-0.486
(5.56)*
-0.128
(0.76)
-0.598
(3.03)*
4.340
(0.52)
-2.052
(0.78)
1.001
(2.94)*
-0.004
(0.01)
-0.492
(5.66)**
-0.153
(0.91)
-0.600
(3.06)**
3.374
(0.41)
-1.854
(0.73)
1.080
(3.17)**
1.108
(1-35)
-0.512
(5.78)**
-0.116
(0.70)
-0.605
(3.03)**
3.737
(0.45)
-1.622
(0.63)
1.053
(3.12)**
-0.508
(5.93)**
-0.131
(0.80)
-0.586
(3.01)**
7.049
(0.87)
-3.455
(1-35)
0.884
(2.65)**
0.385
(0.98)
0.818
(1-45)
0.486
(1-17)
-2.003
(2.64)*
-0.150
(0.39)
Adjusted R
2
Number of Companies
0.315
144
0.320
144
0.322
144
0.345
144
Notes: Numbers in parentheses are t-statistics.
* Coefficient is statistically significant at the 10% level.
** Coefficient is statistically significant at the 5% level.
a
 The average annual ratio of the market capitalization of common equity to the book value of
common equity.
Ownership is defined as the percentage of corporate votes held by inside and outside directors.
c
 Financial Directors are narrowly defined as outside financial directors who are employees of
financial institutions.
Financial Directors are broadly defined as outside financial directors who are either
employees or directors of financial institutions.
4H
OUTSIDE FINANCIAL DIRECTORS AND CORPORATE GOVERNANCE
indicate that corporate size is positively related to performance, that corporate
value increases when ownership exceeds 20 percent, and that the governance
structure of regulated firms has a different effect on their performance compared
to that of non-regulated companies.
The two public-policy implications of this research are related to the
effectiveness of outside directors and the disclosure of the background of out-
side directors which will enable analysts, investors and securities regulators to
identify and therefore separate affiliated and independent directors. First, the
effectiveness of outside directors may not be due to knowledge per se, but may
rather be dependent on other personal attributes. This stems from the empiri-
cal finding that the presence of highly informed outside financial directors on
the board has no significant causal relationship to corporate performance.
Thus, legislation that sets a minimum proportion of independent outside
directors is not likely to have a significant positive or negative effect on corpo-
rate performance. Second, improvements are needed regarding the disclosure
of background, affiliation, interlocking directorships, and tenure of outside
directors. Especially weak is the requirement to disclose whether firms with
which a director is associated — eithe r as an employee or a director — are
related by common ownership. At present it is difficult to identify who is an
unrelated or independent outside director. Without better disclosure, legislation
based on a minimum proportion of independent outside directors will not be
effectively monitored.
ENDNOTES
1 It appears that the issue as to whose interests are represented by directors is
debatable. Wainberg &. Wainberg (1987) state that directors, acting as a
board, are agents of the corporation. They are agents in the sense that they
have the responsibility of managing the assets of the corporation to make
profit and they are also trustees in the sense that they have the responsibility
of preserving the assets of the corporation. Hatton (1990) also argues that
legally, directors are required to represent the best interests of the corporation
and not the interests of shareholders and other stakeholders because the
powers of directors to manage the operations of the company come from
legal statutes and not the shareholders. Gillies (1992) and Mace (1986)
take the broader approach and argue that directors have a fiduciary
responsibility to all stakeholders including shareholders, creditors, employees,
consumers and society as a whole.
2 With some exceptions, most empirical studies have found either a positive
or neutral relationship between the proportion of outside directors on the
board and corporate performance. Rosenstein & Wyatt (1990) in analyzing
415
AMOAKO-ADU & SMITH
the effectiveness of outside directors find that positive wealth effects result
from the appointment of outside directors. Kaplan & Minton (1994) find
that the appointment of outside directors leads to higher executive
turnover and a modest improvement in performance. Brickley, Coles &
Terry (1994) report that the stock market reaction to the announcement
of poison pills is positive when the board has a majority of outside directors
and negative when it does not. Brickley, Lease & Smith (1988) also provide
evidence indicating that institutional investors and blockholders are more
likely to vote against management when they think that a corporate proposal
would harm shareholders. However, Murphy (1985), Hermalin &
Weisbach (1988, 1991) and Kaplan & Reishus (1990) find no evidence
that outside directors add value to the company. In addition, Pound
(1988) provides evidence which indicates that due to conflict of interest
pressures, institutional investors and outside board directors are more likely
to vote in support of management but against their fiduciary responsibilities
to shareholders and other stakeholders.
3 The Canada Business Corporations Act (CBCA) requires that public corpora-
tions have at least two directors who are not employees of the corporation
or its affiliates. On the other hand, the Ontario Business Corporations Act
(OBCA) states that a minimum of one-third of the directors of a public
corporation must be outside directors. In this case, the minimum proportion
of outside directors required may be at conflict. Two outside directors on a
board of ten will meet the CBCA but not the OBCA requirements.
4 The draft report of the Toronto Stock Exchange Committee on Corporate
Governance in Canada defines an unrelated director as "a director who is
free from any interest and any business or other relationship which could,
or could reasonably be perceived to materially interfere with the director's
ability to act with a view to the best interests of the corporation".
Unrelated directors are those outside directors who would be considered
independent of management. It is unclear whether firms would classify
their outside financial directors as unrelated.
5 Empirical studies from the United States and Japan on the role of directors
from the financial sector do not provide consistent results. In a study of
U.S. firms, Van Nuys (1993) shows that even directors from financial
institutions that do not have existing business with the firm tend to
support managers on anti-takeover announcements more frequently than
other groups. However, Rosenstein & Wyatt (1990) find that in the
United States appointments of outside directors from any particular industry
have no different impact from any other industry. Kaplan & Minton
(1994) report that Japanese directors who are employees of banks are
appointed in companies that are financially distressed and have large bank
borrowings. Directors from non-financial firms are appointed to companies
that have temporary problems. In both cases, the outside directors play a
monitoring role.
416
OUTSIDE FINANCIAL DIRECTORS AND CORPORATE GOVERNANCE
6 For a discussion of "grey area" directors, see Baysinger & Butler (1985),
Hermalin & Weisbach (1988), and Byrd & Hickman (1992a).
7 Byrd &. Hickman (1992) argue that the cross-sectional analysis relating
corporate performance to board composition used in this study and in U.S.
studies such as Hermalin & Weisbach (1991) are unlikely to find significant
results because the presence of outside directors has an impact only at the
time of extraordinary events such as takeover attempts or the adoption of
poison pills.
8 The literature is unclear as to the direction of causality between corporate
performance and board composition. The reverse causality where the
board composition can be determined by corporate performance is analyzed
in St-Pierre, Gagnon & Saint-Pierre (1992).
9 A number of Canadian companies have foreign parent companies. In
almost all cases, executives from the foreign parent company were directors
on the board of the Canadian subsidiary. The percentage of votes controlled
by directors includes those votes controlled by the foreign parent company.
10 Morck et al. (1988) argue that the relationship between corporate value
and ownership can be positive reflecting convergence-of-interests hypothesis
or negative indicating that the entrenchment hypothesis outweighs the
convergence-of-interests hypothesis. Their piece-wise linear regression of
Tobin's Q on ownership indicated that the slope was positive to ownership
between zero and 5 percent, negative for the 5 percent and 25 percent
range, and positive but weaker for ownership in excess of 25 percent.
11 The systematic risk was estimated with the adjusted beta of Dimson
(1979). This measure adjusts for possible thin trading. The Dimson beta is
the sum of the BjS from the following regression:
+ 1
Rjt = OQ + £ Pi Rmt+i
 + e
jt
i = -l
where R,-
t
 and R
mt
 are the monthly holding period returns for the stock
and the TSE 300 Composite Index. The regression was run over the five-
year period prior to the cross-sectional analysis in equation (2).
12 Brickley, Lease &. Smith (1988) distinguish between outside directors from
"pressure-sensitive" financial institutions (banks, trust companies, invest-
ment dealers and insurance companies) and those from "pressure-resistant"
financial institutions (public pension funds and mutual funds). No distinction
is made in this study because insufficient information was available to
separate underwriters from other institutional investors.
417
AMOAKO-ADU & SMITH
ACKNOWLEDGEMENT
T
HIS RESEARCH WAS FUNDED WITH A Social Sciences and Humanities
Research Council Grant and a Wilfrid Laurier University Short-Term
Research Grant. We would like to thank our research assistants, Yuxing Yan,
Yuhui Liu and Raman Krishnaprasad, for the data collection and programming
work. This study was presented at the Conference on Corporate Decision
Making in Canada, organized by Industry Canada and the Financial Research
Foundation of Canada, March 20-21, 1995.
BIBLIOGRAPHY
Ang, J. S., and J. H. Chua. "Corporate Bankruptcy and Job Losses Among Top Level
Managers." Financial Management. (Winter 1981):70-74.
Arnold, Hugh J., and Patrick F. O'Callaghan. "The New Board of Directors: A Survey of
Chief Executive Officers." Business Quarterly. (Summer 1988).
Barr, Joseph W. "The Role of the Professional Director." Harvard Business Review. (May-
June 1976).
Baysinger, Barry D., and Henry N. Butler. "Corporate Governance and the Board of
Directors: Performance Effects of Changes in Board Composition." Journal of Law,
Economics, and Organization, 1, 1 (Fall 1985):101-124.
Brancato, C. K. and P. A. Gaughan. "Institutional Investors and Their Role in Capital
Markets." Monograph, Columbia University Center for Law and Economics, 1988.
Brickley, J. A., J. L. Coles and R. L. Terry. "Outside Directors and the Adoption of Poison
Pills." Journal of Financial Economics, 35 (1994):371-390.
Brickley, J. A., R. C. Lease and C. W. Smith, Jr. "Ownership Structure and Voting on
Antitakeover Amendments." Journal of Financial Economics, 20 (1988):267-91.
Byrd, J. W. and K. A. Hickman "Do Outside Directors Monitor Managers? Evidence From
Tender Offer Bids." Journal of Financial Economics, 32 (1992a):195-221.
. "The Case for Independent Outside Directors." Journal of Applied Corporate
Finance, 5 (1992b):78-82.
Coughlan, A. T. and R. M. Schmidt. "Executive Compensation, Managerial Turnover, and
Firm Performance: An Empirical Investigation." Journal of Accounting and Economics.
(1985):43-66.
DeAngelo, L. E. "Managerial Competition, Information Costs, and Corporate
Governance." Journal of Accounting and Economics. (January 1989):3-36.
Demsetz, H. "The Structure of Ownership and the Theory of the Firm." Journal of Law and
Economics, 26 (1983):375-90.
Demsetz, H. and K. Lehn. "The Structure of Corporate Ownership." Journal of Political
Economy, 93 (1985):! 155-77.
Dimson, E. "Risk Measurement When Shares are Subject to Infrequent Trading." Journal of
Financial Economics, 7 (1979):197-226.
Easterbrook, F .H. and D. R. Fischel "Voting in Corporate Law." Journal of Law and
418
OUTSIDE FINANCIAL DIRECTORS AND CORPORATE GOVERNANCE
Economics, 23 (1983):395-427.
Epstein, Edward M. "The Corporate Social Policy Process and the Process of Corporate
Governance." American Business Law Journal, 25 (1987).
Fama, E. F. and M. Jensen. "Separation of Ownership and Control." Journal of Law and
Economics. (1983):301-25.
Fredrickson, J. W., D. C. Hambrick, and S. Baumrin. "A Model of CEO Dismissal."
Academy o/Management Review. (April 1988):255-70.
Furtado, E. P. H. and V. Karan. "Causes, Consequences, and Shareholder Wealth Effects of
Management Turnover: A Review of the Empirical Evidence." Financial Management.
(1990):60-75.
Gaughan, P. A. Mergers and Acquisitions. New York: Harper Collins Publishers Inc., 1991.
Gillies, James. Boardroom Renaissance: Power, Morality and Performance in the Modern
Corporation. Toronto: McGraw-Hill Ryerson Limited, 1992.
Hatton, Michael J. Corporations and Directors: Comparing the Profit and Not'/or-Pro/it Sectors.
Toronto: Thompson Educational Publishing, Inc., 1990.
Hermalin, Benjamin E. and Michael S. Weisbach. "The Effects of Board Composition and
Direct Incentives on Firm Performance." Financial Management, 20, 4 (Winter
1991):101-12.
. "The Determinants of Board Composition." Rand Journal of Economics, 19, 4
(Winter 1988), 589-606.
Holderness, C. G. and D. P. Sheenan "The Role of Majority Shareholders in Publicly Held
Corporations: An Exploratory Analysis." Journal of Financial Economics, 20 (1988):317-46.
Jensen, M. C. "Eclipse of the Public Corporation." Harvard Business Review. (1989):61-74.
Jensen, M. C. and W. H. Meckling. "Theory of the Firm: Managerial Behaviour, Agency
Costs and Ownership Structure." Journal of Financial Economics. (1976):305-60.
Jensen, M. C. and J. B. Warner. The Modern Corporation and Private Property. New York:
Macmillan Inc., 1988.
Kaplan, S. N., and B. A. Minton. "Appointments of Outsiders to Japanese Boards:
Determinants and Implications for Managers." Journal of Financial Economics, 36
(1994):225-58.
Kaplan, Steven N., and David Reishus. "Outside Directorships and Corporate
Performance." Journal o/Financial Economics, 27 (1990):389-410.
Kovacheff, J. D. "Managing the Board: A Survey of Chairmen of Canadian Corporations."
London, ON: National Centre for Management Research and Development, University
of Western Ontario, 1991.
Leighton, David S. R. and Donald H. Thain. "The Role of the Corporate Director."
Business Quarterly. (Autumn 1990).
Mace, Myles L. Directors: Myth and Reality. Boston, Harvard Business School Press, Division
of Research, 1986.
. "The Board and the New CEO." Harvard Business Review. (March-April 1977).
. "The President and the Board of Directors." Harvard Business Review. (March-
April 1972):37-49.
Morck, R., A. Shleifer, and R. W. Vishny. "Management Ownership and Market Valuation:
An Empirical Analysis." Journal of Financial Economics. (March 1988):842-52.
."Alternative Mechanisms for Corporate Control." American Economic Review.
(September 1989):293-315.
Murphy, K. J. "Corporate Performance and Managerial Remuneration." Journal of
Accounting and Economics. 7 (1985).
419
AMOAKO-ADU & SMITH
Osier, Hoskin and Harcourt. "Directors' Duties: A Guide to the Responsibilities of
Corporate Directors in Canada", Toronto, 1993.
Patton, Arch and John C. Baker. "From The Boardroom." Harvard Business Review.
(November-December 1987):10-18.
Pawling, John D. "The Crisis of Corporate Boards: Accountability vs. Misplaced Loyalty."
Business Quarterly. (Spring 1988):26-8.
Pound, J. "Proxy Contests and the Efficiency of Shareholder Oversight." Journal of Financial
Economics, 20 (1988):237-65.
Rosenstein, S. and Jeffrey G. Wyatt. "Outside Directors, Board Independence and
Shareholder Wealth." Journal of Financial Economics, 26 (1990):175-91.
St-Pierre, Josee, Jean-Marie Gagnon and Jacques Saint-Pierre. "Takeover Bids, Structure of
Ownership and Board Composition." Working Paper, Laval University, 1992.
Shleifer, A. and R. Vishny. "Large Shareholders and Corporate Control." Journal of Political
Economy, 94 (1986):461-88.
Thain, D. H. and D. R. S. Leighton. "Ownership Structure and the Board." Canadian
Investment Review, IV (Fall 1991):61-5.
Toronto Stock Exchange Committee on Corporate Governance in Canada. '"Where Were
the Directors?': Guidelines for Improved Corporate Governance in Canada." Toronto:
Toronto Stock Exchange, 1994.
Vance, Stanley C. Corporate Leadership: Boards, Directors and Strategy. New York: McGraw-
Hill Book Company, 1983.
Vancil, R. F. Passing the Baton: Managing the Process of CEO Succession. Boston, MA:
Harvard Business School Press, 1987.
Van Nuys, K. "Corporate Governance Through the Proxy Process: Evidence from the 1989
Honeywell Proxy Solicitation." Journal of Financial Economics, 34 (1993):101-32.
Wainberg, J. M. and Mark I. Wainberg. Duties and Responsibilities of Directors in Canada.
Toronto: CCH Canadian Limited, 6th Ed., 1987.
Waitzer, E. J. "Are Institutional Investors Really Impacting Corporate Governance?"
Canadian Investment Review, IV (1991):9-14.
Warner, J. B., R. L. Watts and K. H. Wruck. "Stock Prices and Top Management
Changes." Journal of Financial Economics. (March 1988):461-92.
Weisbach, M. S. "Outside Directors and CEO Turnover." Journal of Financial Economics, 20
(1988):431-60.
420
DISCUSSANTS' COMMENTS ON PART IV
Mark R. Huson
Faculty of Business
University of Alberta
Commentary on Part IV
M
UCH OF THE EXISTING LITERATURE on the agency problems in firms is
devoted to understanding the conflict of conflicting interests between
shareholders and bondholders, and between shareholders and managers.
Solutions to these conflicts include debt covenants, incentive compensation
schemes and monitoring by shareholders. Increasingly, institutional shareholders
are playing an important role in the process of monitoring managers in the
United States. The studies presented in this session investigate the extent to
which Canadian institutional investors are active in creating firm value.
Because of the unique ownership structure that exists in Canada, the
extent to which Canadian institutional investors are active in corporate
governance and the effect of their presence on firm value should not be
compared too directly to the activities of U.S. institutional investors. The
presence of controlling shareholders provides an additional agency conflict:
the shareholder/shareholder conflict. The ability of the controlling shareholder
to affect corporate direction reduces incentives for institutional investors to
engage in governance activities.
In the realm of governance, institutions are faced with a number of
decisions. First, they have to decide whether or not to monitor. Then, based
on the information set they choose, they must decide whether to exit, exercise
voice, or remain loyal. Patry and Poitevin consider the institutional incentives
to monitor as well as incentives to exit, exercise voice or remain loyal.
Foerster looks at a potential model for monitoring and discusses the types of
governance activism that are currently being used in Canada. The effective-
ness of these activities are examined by Macintosh and Schwartz and
Amoako-Adu and Smith. Macintosh and Schwartz examine the effect of the
presence of institutional investors on firm value. Amoako-Adu and Smith
investigate the effect of the presence of outside financial directors on firm per-
formance. It should be noted that failure to reject the null hypotheses of these
empirical studies means that one cannot distinguish between two alternatives
- i.e., that institutions do not try to influence corporate performance, or, that
they try, but are ineffective.
Patry and Poitevin look specifically at the governance and organizational
structures of institutional investors to determine if these have any effect on
their incentives to monitor. In order to do so, they first discuss the nature of the
agency problem in firms, the exit/voice mechanisms, and then they highlight the
421
HUSON/WEISBACH
increasing importance of institutions in financial markets. They go on to discuss
the governance structure of institutional investors and how it affects monitoring
incentives. They consider two types of institutions: mutual funds and pension
funds.
Their analysis leads them to conclude that the nature of mutual funds
leads fund managers to behave in a manner that results in maximizing return
to investors. However, the liquidity concerns of mutual fund managers do not
allow for long'term investments. Pension fund managers, on the other hand,
do not have the same liquidity constraints and therefore can accommodate a
longer-term investment horizon. The problem is that since most pension funds
have captive investors, there is no incentive to maximize investor return and,
therefore, there is low incentive to monitor.
The analysis assumes that mutual fund managers act to maximize share-
holder return. Failure to maximize returns will result in a removal of funds
from a manager's discretion. If returns can be augmented through monitoring,
then mutual fund managers who monitor should outperform other funds. In
outperforming other funds, a fund manager will draw more funds to his
management, and liquidity concerns will be less pronounced. This suggests
that if monitoring is a value additive, mutual fund managers will engage in
monitoring. The lesson to be learned from the lack of monitoring by mutual
funds is not altogether clear. Is it that the very structure of mutual funds makes
it inconceivable, or is it that monitoring is not a value additive?
I would like to raise three additional points with regard to this analysis.
The first and second have to do with the assumption that mutual fund
managers act to maximize shareholder return. In motivating this assumption,
the authors claim that mutual fund shareholders do not have a comparative
advantage in selecting stocks and should therefore leave the management of
the portfolio in the hands of the manager. They go on to suggest that because
there is sufficient information to enable investors to judge the manager's
performance and also enough suitable alternatives available to investors, fund
managers are forced to maximize return.
My first point deals with the idea that mutual fund shareholders should
leave the management of the fund to the fund managers. This should be kept
in mind since it deals directly with the concept of "relational investing"
(discussed elsewhere) as a means of corporate governance. The second point
has to do with the presence of readily available substitutes for specific mutual
funds as a way of inducing individual fund managers to maximize value. This is
related to the foreign investment constraints placed on Canadian institutional
investors.
My third point is aimed at the notion that pension fund managers lack
the incentive to monitor. I suggest this could be handled via incentive com-
pensation contracts. In addition, the high level of activism shown by CalPERS
and other pension funds in the United States demonstrates that pension funds
do indeed have an incentive to monitor.
422
DISCUSSANTS' COMMENTS ON PART IV
The CalPERS monitoring and activism model is analyzed by Foerster,
who also looks at monitoring and intervention techniques of Canadian public
pension funds. Foerster asks why Canadian institutional investors have not
been as "visibly" active as their American counterparts. He offers four possible
reasons. The first is that there are regulatory differences that limit the activity
of Canadian institutions. The second is that Fairvest and PIAC assume the
activist role on the part of institutions. The third is that Canadian institutions
have only recently become large enough to act effectively, but more activity
should be expected now that they are large. The final possibility is that the
"Canadian style" of activism is not as visible as the American style.
Foerster writes:
During the 1990s, numerous large pension funds have become more actively
involved in corporate governance issues, largely because of their growing size
and presence in equity markets - which makes it more difficult for them simply
to sell their shares if they are dissatisfied with management. In addition to the
Caisse, funds such as the Ontario Municipal Employee's Retirement System
(OMERS), the Ontario Teachers' Pension Plan Board (Ontario Teachers),
CN Investments and other large funds have gained reputations for their
involvement in corporate governance issues... .'
If these statements are accurate, then Canadian institutional investors
do indeed engage in activist pursuits. This statement also lends credence to
the third explanation offered by Foerster. It would be interesting to know what
were the issues involved in these institutional interventions, as well as some-
thing about their effects.
Foerster describes the screening method used by CalPERS to identify
problem companies and the methods employed to induce change. He goes
through an example of the screening process, using a sample of Canadian
firms, to show that it can be done with Canadian data. The question Foerster
does not answer, however, is what percentage of the firms identified as requir-
ing intervention have controlling shareholders. He states that CalPERS
systematically avoids firms with high levels of insider ownership. Considering
that approximately 80 percent of the TSE 300 have insider ownership over 20
percent, it would be interesting to see the correlation between relative perfor-
mance and the presence of a controlling shareholder. This is necessary since
the presence of a controlling shareholder influences the effectiveness of
CalPERS' intervention style. In fact, CalPERS' style of intervention may not
even be relevant, given the ownership structure of Canadian firms.
Additionally, Foerster's use of return series that exclude dividends makes
it likely that high-dividend-paying stocks will under-perform relative to the
index and low-dividend-paying stocks will over-perform relative to the index.
To illustrate, however, consider two stocks that have the same total return to
an investor; one pays a dividend and one does not. Calculate the return
423
HUSON/WEISBACH
excluding dividends and the performance of these firms relative to a sans
dividend index. Even though both stocks provide the same total return, the
dividend-paying stock will appear to under-perform the index. This should not
be used as an example of how to rank companies.
In describing the "Canadian style", Foerster reports a reactive style that
emphasizes non-public communication with management to air grievances
about performance. He makes an effort to contrast this with the confronta-
tional CalPERS style. I think the difference in style is more perceived than real.
CalPERS' system is also reactive. It targets firms that have performed poorly. It
then tries to alleviate the problems through private discourse with manage-
ment. Only if it is rebuffed does CalPERS launch a public campaign to change
managerial practices. I consider this to be very similar to the "Canadian style".
Whereas Foerster's analysis suggests that Canadian institutional investors
are involved in governance issues, Macintosh and Schwartz examine the effect
of the presence of institutional shareholders on the performance of the firm.
They discuss the nature of the shareholder/shareholder conflict that exists in
has an incentive to monitor managers in order to maximize firm value. He also
has the further incentive to redirect the flow of value to himself and away
from minority shareholders. With this in mind, the presence of controlling
shareholders should be positively related to accounting measures of perfor-
mance, but inversely related to market measures of performance.
Their evidence is consistent with the theory that controlling shareholders
help to control the manager/shareholder conflict of interest but exacerbate the
shareholder/shareholder conflict of interest. The evidence on whether institu-
tional investors affect firm value is a little more difficult to interpret. They
find that institutions tend to have more ownership in firms with high account-
ing numbers (earnings, ROA) but an inverse relation between institutional
holdings and a market measure of performance. While this is consistent with
institutions monitoring managers (since institutions cannot redirect resources
to themselves), both measures of performance should be positively related to
the presence of institutions. Given the timing mismatch of the measures, I
think the evidence is more supportive of the theory that institutions follow
value-based strategies of buying stocks with high earnings-price ratios and selling
stocks with low earnings-price ratios. They do, however, provide some evidence
that institutions tend to control the shareholder/shareholder conflict.
Nonetheless, the results of this study, as it stands, do not allow for conclusions
to be drawn or for policy statements to be made.
Amoako-Adu and Smith look at the effect of the presence of outside
financial directors on firm value. This research is important because it bears
directly on the question of the efficacy of putting representatives of institu-
tional investors on the board. However, in its current state the research does
not separate captive from non-captive financial directors. Also, by including
any holder of a block greater than 10 percent in the insider group, the authors
424
firms with controlling shareholders. Specifically, the controlling shareholder
DISCUSSANTS' COMMENTS ON PART IV
potentially include representatives of institutional investors as insiders.
2
 They
conclude that their results indicate no value associated with the presence of
outside financial directors or outside directors in the same line of business.
These results are interesting in light of the recommendations of the Dey
Committee and those contained in other studies presented in this session -
that more outsiders be placed on boards. However, I cannot overemphasize the
need for more compelling evidence about the effects of board composition
than that presented in Amoako-Adu and Smith before policy recommendations
are made.
One interesting result in the Amoako-Adu and Smith study that the
authors do not make much of is the differential effect that a large control
block has on ROA and Asset Turnover. Define the following.
ROA< = ^II
Assets j
Asset Turnover,- = 1
Assets ,-
^ Sales; — EDIT. Sates,-a,
Costs j =
 l
- L = ! '
Assets j Assets j
If there were no correlation between ownership structure and cost structure, the
coefficient in the ROA regression would be 0.10 multiplied by the coef-
ficient in the asset turnover regression, since that is the average profit margin.
The fact that the coefficient in the ROA regression is less than 10 percent of
the coefficient in the asset turnover regression indicates a positive correlation
between ownership structure and cost structures. This is consistent with
controlling shareholders removing value that other shareholders cannot share.
The discussion in Macintosh and Schwartz as well as the information
contained in Foerster indicate that, despite the problems enumerated in Patry
and Poitevin, institutions do become involved in corporate governance.
Additionally, this involvement appears to be beneficial. The results in
Amoako-Adu and Smith suggest that the reform advocated by the Dey
Committee is, at best, ineffective. This begs the question, what other reforms,
if any, are needed to improve corporate governance in Canada?
Given the evidence of increased activism by public pension funds, it is
not clear that reform is needed. Additionally, reforms that reduce the ability of
controlling shareholders to siphon off the gains from their monitoring of
managers may not benefit minority shareholders. It is true that minority share-
holders would now get a bigger slice of the pie; however, the pie itself may well
get smaller and, if the pie does not get smaller (or larger), then the most
reform(s) will do is legislate a wealth transfer. Nevertheless, since institutional
425
Sales
HUSON/WEISBACH
investors can potentially monitor the manager/shareholder conflict, it may be
beneficial to reduce the inter-shareholder conflicts. One way to achieve this is
to reduce the costs of institutional involvement along with the insulation from
punitive measures currently enjoyed by controlling shareholders. I will now
discuss some of the reforms mentioned in the papers as well as offer some of
my own thoughts on how to improve corporate governance in Canada.
Patry and Poitevin suggest mandatory fund indexation and flip taxes on
capital gains. Both of these reforms would increase the cost of funds exiting,
thereby increasing their incentives to use voice. Exit is using voice; it tells
management in no uncertain terms that their practices are not acceptable.
Exit strategies are already costly for Canadian institutions, given the size of
their holdings and the thinness of Canadian equity markets. Besides, increasing
exit costs would prove beneficial to controlling shareholders because it would
force the institutions to stand still longer while their wealth is being expropriated.
Another suggestion that runs through the papers is the use of relational
investing. The general idea behind relationship investing seems to be that
institutions should take large stakes and then participate in "active manage-
ment". However, the meaning of the term "active management" is not clear.
Does it mean direct involvement in strategic decisions? Does it mean involve-
ment in day-to-day business decisions? Recall from an earlier reference here
that Patry and Poitevin indicated that investors in mutual funds would not
think of telling the managers which stocks to buy. Why then should we expect
a fund manager to have the ability to tell an operating manager the best way
to make widgets?
In my opinion the way to achieve increased institutional activism is to
reduce the costs of using voice and exit. To this end I suggest the following.
• Change the proxy solicitation rules as they apply to institutional
investors to allow communication on governance issues without
the cost of formal solicitation.
• Change the definition of "control person" as it now applies to
institutional investors. Doing so would allow such investors to
act in concert against controlling shareholders. It would also
remove the need to have secondary distributions to exit, and free
them from concerns of insider trading violations if they do exit.
3
• Change the valuation technique when shareholder appraisal
rights are used. There are certain actions taken by firms that are
known to reduce value. Poison pills and greenmail come to
mind. In cases like these, the change in value attributable to
managerial action is fairly clear. In these cases, an investor
should be able to put the stock back to the issuer at the pre-event
price level.
426
DISCUSSANTS' COMMENTS ON PART IV
• Link the pay of outside board members to the value of the non-
voting shares of stock.
• Remove the foreign investment limits from the portfolios of
institutional investors.
I would like to elaborate more on the last point. The size of the Canadian
equity market and the inter-relatedness of the ownership of Canadian firms
makes exit a questionable strategy if an investor is dissatisfied with manage-
ment. The dissatisfied investor utilizing an exit strategy merely goes from
Bronfman I to Bronfman II.
Additionally, one of the costs borne by a controlling shareholder for
expropriating wealth from minority shareholders is a higher cost of capital.
The higher cost of capital is a result of funds being diverted to uses where
investors are not held up. By limiting the ability of investors to take their
money out of Canada, a captive investor pool is maintained. This keeps the
supply of investment capital high and prevents the cost of capital from rising
to a level that reflects the amount of self-dealing engaged in by controlling
shareholders. Relaxing the foreign investment criteria is similar to providing a
vast pool of mutual funds to retail investors. It provides a larger market where
entrepreneurs compete for funds. This, in turn, gives entrepreneurs incentives
to maximize shareholder value if they want a low cost of capital.
ENDNOTES
1 Similar claims are made by Macintosh and Schwartz. They state that
Canadian institutional investors have voted against management proposals
(sometimes in conjunction with other institutional investors), and have
also threatened to use dissent rights and sue managements. This is also
evidence that Canadian institutions are actively and visibly involved in
governance issues.
2 I would also like to add that they should consider utilities separately. For
one thing, utilities have an additional layer of regulatory oversight that
should control agency problems. Additionally, utilities have low asset
turnover and large asset bases. The authors document an association
between board composition and asset turnover. Are they just picking up
utilities?
3 The last point should take effect only if the institutional investor sunshine
trades.
427
HUSON/WEISBACH
Michael S. Weisbach
College of Business and Public Administration
University of Arizona
Commentary on Part IV
Institutional Investors and the
Governance of Canadian Corporations:
An American Perspective
I
T WAS A PLEASURE TO READ AND TO THINK ABOUT the four interesting studies
on the topic of the influence of institutional investors on Canadian corporate
governance, by Macintosh and Schwartz, Patry and Poitevin, Foerster, and
Amoako-Adu and Smith. In doing so, I feel that I have learned a great deal
about the way corporate finance is practiced here in Canada. Since it has
been a long day I will try to keep my remarks brief.
My first reaction to the studies is that their tone is overly deferential to
the United States. The general theme tends to be: "Such and such is done this
way in the United States and we do it differently here in Canada, so what can
we change?" I find this kind of amusing, because it seems that every time I go
to a conference in the United States, the theme is that we are doing every-
thing wrong and we should be copying the Japanese or the Germans.
1
 When
you think about what we do south of the border, you should be careful to
remember that we are not quite sure if American corporate governance practices
and the associated regulations make sense for us, let alone whether they
should be copied by other countries.
A question that summarizes the agenda of the studies presented in this
session is: "How can Canadian institutional investors be encouraged to monitor
Canadian Companies better?" One important issue here is the extent to which
Canadian institutions utilize the blocks of Canadian stock they currently hold.
Canadian institutional investors should probably be encouraged to hold onto
those blocks and to gain as much influence as they can on the governance of
the corporations affected - including the possibility of obtaining a seat on the
board. I would also caution against underestimating the amount of influence
being exerted by institutions at this time. At least in the United States, most
influence activity goes on quietly behind the scenes so it is easy to underestimate
the extent to which this type of activity already occurs by concentrating on
what one reads in the press.
When we think about the issue of institutional ownership, one issue that
has come up repeatedly throughout this conference is the restrictions that are
placed on Canadian financial institutions with respect to ownership of foreign
equity. In finance theory, it is quite clear that the benefits of diversification are
428
DISCUSSANTS' COMMENTS ON PART IV
maximized when the set of risky assets held is the value-weighted portfolio of
all the risky assets in the world. This means that since the Canadian equity
market is probably (at most) one or two percent of the world's equity market,
Canadian institutions should have at most one or two percent of their invest-
ments in Canadian equities. In addition, one ought to consider the human
capital of the beneficiaries. Since the beneficiaries are Canadians, the value of
their human capital will vary according to how well the Canadian economy is
doing. Because of the positive correlation between Canadian human capital
and the Canadian stock market, the optimal fraction of pension wealth that
ought to be held in Canadian securities is actually lower than this one or two
percent. In fact, if one were to be persistent about having some sort of foreign
equity restrictions, perhaps the most sensible rule would be to prohibit
Canadian institutions from owning any Canadian securities at all, rather than
the current rule which forces them to own Canadian securities. Such a restriction
would encourage financial institutions to act in the interest of their beneficiaries,
providing them with the diversification they desire.
Let me briefly discuss the studies individually. The one that I found to be
the most inflammatory was the paper by Foerster on CalPERS. He puts the
question: "Why haven't pension funds in Canada been as visibly active as
CalPERS?" I would phrase the question somewhat differently: "Why haven't
pension funds anywhere been as visibly active as CalPERS?" I think the answer
has something to do with the fact that highly visible activism as practiced by
CalPERS is costly and does not help the beneficiaries of the fund. Foerster
relies on studies by Nesbitt, which claim a 40 percent stock price improve-
ment following CalPERS involvement. I have not seen these studies, nor have
they been published in an academic journal, so it is difficult to know what to
make of them. One issue that seems relevant is a growing literature documenting
that stock prices tend to exhibit mean reversion over certain horizons. Therefore,
if one picks any random group of stocks that has done extremely poorly and
follows them over time, it is likely that they will have positive abnormal
performance. Since CalPERS targets stocks based on their poor perfor-
mance, this may be one alternative explanation for Nesbitt's findings of high
subsequent performance.
A recent study by Wahal is relevant here.
2
 He finds that there are small
(about 1 percent) increases in the prices of a stock on the announcement of
CalPERS' involvement in a company, and no long-term effects on either stock
prices or accounting measures of performance. Especially with Wahal's results
in mind, it seems to me that Foerster drastically overstates the effectiveness of
CalPERS' monitoring on American corporations.
The Patry and Poitevin study surveys a large literature and raises a number
of different issues. One of their proposals that intrigues me calls for mandatory
indexing of pension fund assets. Pension fund investment policies are a puzzle
to many economists for the following reason: Indexing appears to be the best
way to maximize risk-adjusted returns net of expenses, yet few funds are
429
HUSON/WEISBACH
indexed. The market apparently has not found the optimal solution. The reason
why the market has not found the optimal solution is not clear, but Patry and
Poitevin's explanation of moral hazard on the part of fund managers is as
plausible as any I have heard.
I tend to think that this sort of regulation isn't such a bad idea. However,
a potential cost associated with regulating portfolio strategies is that such
regulation reduces flexibility on the part of the pension funds. Perhaps a better
solution would be to allow Canadian funds to follow the example of TIAA-
CREF, whose trustees and managers voluntarily decided to index a large portion
of their stock portfolio.
The two remaining studies, by Macintosh and Schwartz, and Amoako-
Adu and Smith, both use the methods developed by Morck, Shleifer & Vishny
(1988) to measure the effect of corporate governance on firm performance.
The technique is to regress an estimate of Q, the ratio of the market value of
the firm's assets to the replacement cost, on variables representing the firm's
governance structure. Presumably, higher Qs are representative of better
operating performance, so that the empirical relations uncovered would capture
the effect of alternative ownership structures on firm performance. In the case
of the Macintosh and Schwartz study, the governance variables represent
institutional ownership of the firm, and in the Amoako-Adu and Smith study
they represent the composition of the firm's board of directors.
I think an important issue for the authors of these studies is to consider
what economists call endogeneity problems. These problems arise if corporate
governance structures are themselves determined by the firm's performance
and not vice versa, as the authors (of both studies) claim. For example, institu-
tional investors could be attracted to firms that have performed well historically
and could purchase shares in these firms after a run-up in the firm's stock price.
If this were true, it would imply that there would be a relation between firms
with high Qs and institutional ownership, but for reasons different from those
put forward by Macintosh and Schwartz. Similarly, if firms adjust their board
composition in response to their performance, we would find a relation
between composition and performance, but the causality would be different
from that discussed by Amoako-Adu and Smith.
ENDNOTES
1 This practice did slow a bit with the collapse of the Japanese stock market
several years ago.
2 This study came to my attention following the conference in Toronto at
which these remarks were made; so it was not mentioned in the comments
I made there. See Wahal, S., "Pension Fund Activism and Firm
Performance," Unpublished Ph.D. Dissertation, Kenan-Flagler Business
School, University of North Carolina, Chapel Hill, NC,1995.
430
Part V International Aspects of
Corporate Governance
This page intentionally left blank 
Randall Morck
Faculty of Business
University of Alberta
Bernard Yeung
School of Business Administration
University of Michigan
12
The Corporate Governance of Multinationals
INTRODUCTION
G
LOBALIZATION IS A RESULT OF AN INCREASING NEED for companies to gain
access to larger markets in order to recoup the costs associated with the
increased pace of innovation in many industries. This study argues that global-
ization and multinational firms are likely to be even more important to
Canada's competitive position in coming years than they have been in the
past. This global competitive pressure may render moot many of the contem-
porary public policy concerns about corporate governance. In a global economy,
customers, investment capital and highly skilled employees need not tolerate
poor management; they can simply do business with better-run rivals.
Canadian firms must deal with their governance problems not because they
are legally required to do so, but because their survival will depend on it. In
this context, government's best option for improving Canadian corporate
governance may well be to foster competition and openness while providing
good legal and educational infrastructure. This entails weaning firms from
subsidies and captive markets, and providing sound basic public services like
education, health care and law. Some specific issues as to the governance of
multinational subsidiaries in Canada do arise, especially with regard to minority
Canadian shareholders. We advocate tighter disclosure requirements to
increase transparency, and also argue that the boards of foreign subsidiaries
with Canadian minority public shareholders should have conduct committees
charged with approving non-arm's-length transactions with the parent or
other related companies. Indeed, requiring this of all firms with controlling
blockholders and publicly traded minority shares might solve many of
Canada's corporate governance problems in one stroke. Before we can formulate
suggestions about corporate governance in multinationals, we must make clear
why multinational firms are important and how they are different from purely
domestic firms.
Multinationals have always been an important source of capital for
Canada. As of 1990, there was $108 billion worth of foreign direct investment
433
MORCK & YEUNG
in Canada, or 6.6 percent of the world's total stock.
1
 In 1987, foreign owned
subsidiaries accounted for 64-8 percent of total manufacturing sales, 75.4 per-
cent of manufacturing exports and, in 1985-88, 88.3 percent of manufacturing
imports (Corvari et al., 1993). The flow of new foreign direct investment into
Canada has averaged about $10 billion per year in recent years, mostly from
the United States. (Knubley et al, 1994).
Canada's public policy toward inward foreign direct investment has thus
been central to its relations with the United States. In the 1960s and early
1970s cool relations with Washington and popular fears of U.S. domination
led to the establishment of the Foreign Investment Review Agency (FIRA) in
1974. FIRA was allegedly designed to ensure that inward foreign direct invest-
ment brought significant benefits to Canada. Similar emotions brought
about the National Energy Program (NEP), a system of partial expropriation
of foreign investments in the energy sector.
In 1985 the conservative election victory brought about a sharp change
in policy. The NEP was eliminated. FIRA was recast into Investment Canada
and given the mandate of attracting new foreign direct investment to Canada.
The 1989 Canada-U.S. Free Trade Agreement (FTA) and the subsequent North
American Free Trade Agreement (NAFTA) further opened the Canadian
economy, although both provided special status for specific industries, most
notably the so-called "cultural" industries. Investment Canada was made part
of the Department of Industry in 1993, and ceased to exist as an independent
agency in 1994.
Economic policies in the twentieth century have been products of political
ideology, popular opinion and economic realism. Political ideology seems a
spent force, and popular opinion seems increasingly aligned with economic
realism. The growing realization that governments are not monopolies, but are
in competition for footloose investment by global businesses reflects this. One
effect is the on-going development of free trade and regional investment blocs
which is resulting in the multilateral trade and investment liberalization
envisioned in the Uruguay Round of the General Agreement on Tariffs and
Trade (GATT) and the new World Trade Organization (WTO). Given this so-
called "GATTization" of our economy, what policy should Canada adopt toward
foreign-owned subsidiaries in the remaining years of this century and beyond?
It is difficult to over-emphasize the importance to Canada of the global-
ization of business. Rapid advances in information and communications
technology and the economic liberalization of entire regions of the globe like
China, Eastern Europe, and (it is hoped) Latin America have been accompanied
by an explosive growth in foreign direct investment. Business success often
depends on globally integrated marketing, production, research and develop-
ment, and human resources management. The increased competition due to
globalization has brought about substantial improvements in management
skills and in the way business is conducted in many countries. This free flow of
ideas means that firms now strive to learn from the best of their rivals all over
434
THE CORPORATE GOVERNANCE OF MULTINATIONALS
the world. Constant innovation is a requirement for survival. Multinational
enterprises are a conduit of this globalization phenomenon. They are vital for
prosperity and yet they are ruthless enforcers of the "survival of the fittest".
Canada's policy toward foreign-owned subsidiaries in the new economic
order must be informed by an understanding of the economics of multi-
national firms. In the next section of this study, we explore the following
issues: Why do multinationals come to attain a competitive advantage over
uninational firms? How has the behaviour of subsidiaries changed over time?
What effect does foreign direct investment have on a host economy? We
then explore how the international environment for foreign direct invest-
ment is changing. Finally, we present our views on the policy challenge and
make some suggestions.
OUR THEORETICAL AND EMPIRICAL UNDERSTANDIN G
OF MULTINATIONAL FIRMS
WHY FIRMS ESTABLISH FOREIGN SUBSIDIARIES
INTERNATIONAL OPERATIONS ARE NOT SIMPLE TO RUN. when a company enters a
new environment, it must start from scratch to build up an understanding of
the local culture, legal system, regulatory environment, and the business
environment in general. Moreover, doing business is more than just building a
factory or a marketing outlet, it involves making local contacts, hiring correctly
in the local labour and management markets, building up a good working
relationship with suppliers, distributors, and transportation service companies,
as well as with the local government. These considerations suggest that foreign
entrants have an information disadvantage relative to indigenous firms. In the
business literature, the assumption is often referred to as the "home-court"
advantage of local firms. For example, foreign entrants may often have to pay
a premium in hiring local workers. They may make costly mistakes in building
up working relationships with local suppliers and distributors. The question is:
Why, given all these difficulties, would foreign entrants be interested in
establishing overseas operations in the first place, and what allows them to
overcome their local rivals' home-court advantage?
Practitioners offer several answers: access to inexpensive raw materials
and/or labour; access to markets, for example by jumping trade barriers; strategic
response to a competitors' presence; improved flexibility in production or
marketing; access to business intelligence, like the development of the latest
products, production techniques, marketing ideas, etc.; and reducing taxes by
shifting income to subsidiaries in countries with low tax rates.
Since all these ideas seem prima facia legitimate, researchers have sought
to understand more precisely what creates the synergies in a multinational
network of affiliates. Does a multinational structure, in itself, create enough
435
MORCK & YEUNG
value to overcome local firms' home-court advantage, or is there a deeper
economic reason for the survival of multinationals?
Recent research suggests that the intrinsic value of a multinational
structure stems from the intemalization of markets for a company's intangible
assets; and that the above reasons are all tactics to achieve this. Intangible
assets are information-based. They include innovative production techniques,
new marketing skills, brand names, company images, company-based manage-
ment skills, and new organization routines. In short, they are innovations that
bestow a competitive advantage or "edge" to the firm.
Because such intangible assets are information-based, expanding the
scale or scope of their application adds few costs and does not deplete the
intangible assets, but greatly increases the return on their up-front develop-
ment costs. The firm should thus try to take advantage of its edge to as great
an extent as possible in order to gain the most from its innovation.
It is difficult to sell another firm the rights to use an information-based
asset. For example, a buyer might reasonably demand information about a new
marketing technique before paying for it; but once the buyer has the information,
there is no further need to pay for it, since the buyer already has everything he
needs. Patents and copyrights mitigate these problems to some extent, but not
completely. Licensing a new technology to a foreign firm may create a vigorous
future competitor.
To employ its innovative edge to the greatest extent possible while
preserving exclusive control requires that the firm itself expand. In the inter-
national context, this means foreign direct investment. This is what is meant
by internalizing the market for these information-based intangible assets.
Morck & Yeung (1991) examine the relationship between firm value
and multinational structure. Their purpose is to discover whether a multi-
national structure indeed enhances firm value, and to uncover the source of
any changes in firm value.
2
 They find that a multinational structure is correlated
with enhanced firm value to the extent of about 8# per dollar of total physical
assets. Without intangible assets, a multinational structure either has no effect
on firm value, or decreases it. More importantly, Morck & Yeung find that this
increase in firm value is strongly correlated with a firm's past R&D or advertising
spending. Without intangible assets, a multinational structure either has no
effect at all or a decreased effect on firm value. This implies that information-
based intangible assets are required for a multinational structure to add value,
and that without them a multinational structure is a potential liability. Morck
& Yeung (1992) find that announcing the acquisition of a foreign firm adds an
average of 2 percent to the total value of a U.S. firm. This positive stock price
reaction is also restricted to firms with probable intangible assets.
3
 The
announcement of a foreign acquisition by a U.S. firm without intangible assets
either does not change or decreases its value. This indicates that expanded
multinational structure causes the increased firm value in the presence of
intangibles, but without intangibles expanded multinational structure destroys
436
THE CORPORATE GOVERNANCE OF MULTINATIONALS
firm value. It also shows that, in firms with intangibles, foreign acquisitions
cause the value increase and not the converse. Additional studies show that,
on average, multinational firms have a much higher propensity to invest in
intangibles than purely domestic firms.
Upon reflection, it is apparent that the reasons cited by practitioners
derive from increasing the scale of a firm's edge. A firm with such an innovative
edge can afford to enter difficult markets and source raw materials in ways that
would be unprofitable for other firms. Flexibility, strategic moves, and obtaining
intelligence are all long-term investments that a competitive edge makes
possible. Income shifting is most useful when a firm has extra income to shift,
and because the increasing sophistication of the tax authorities is making
transfer pricing more difficult, intra-firm transactions involving intangibles are
becoming the income shifting method of choice (Harris et al, 1993).
In short, from a firm's point of view, international expansion is fundamentally
a way to expand the scale and scope of application of its intangible competitive
edge. International expansion is not an end in itself; rather, it is a means to
combine firm-specific assets with local assets and thereby enhance profits.
Furthermore, both for Canadian firms going abroad and for foreign
parents es