It University

V O LU M E * * N U M B E R * W I NTER * ***

Journal of

APPLIED CORPORATE FINANCE

A MO RG A N S TA N L E Y P U B L I C AT I O N

In This Issue: International Corporate Governance

Ren M. Stulz, The Ohio State University

The Limits of Financial Globalization 8

Art Durnev, McGill University, and E. Han Kim,

Explaining Differences in the Quality of Governance Among Companies: 16

University of Michigan

Evidence from Emerging Markets

David Haarmeyer

The Revolution in Active Investing: Creating Wealth and Better Governance 25

Alan D. Morrison, Oxford University, and

Investment Banking: Past, Present, and Future 42

William J. Wilhelm, Jr., University of Virginia

Stuart Gillan, Texas Tech University, and

The Evolution of Shareholder Activism in the United States 55

Laura Starks, University of Texas at Austin

Charles Elson, University of Delaware

The State of U.S. Corporate Governance: An Interview with Charles Elson 74

Lee Pinkowitz and Rohan Williamson, Georgetown

Cash Holdings, Dividend Policy, and Corporate Governance: 81

University, and Ren Stulz, The Ohio State University

A Cross-country Analysis

P ter Harbula, Deloitte

The Ownership Structure, Governance, and Performance of French Companies 88

Bartley J. Madden

For Better Corporate Governance, A Shareholder Value Review 102

Ian Callaghan, Henry Gonzalez, Diane Maurice, and

Micro nance On the Road to Capital Markets 115

Christian Novak, Morgan Stanley

The Limits of Financial Globalization

by Ren M. Stulz, The Ohio State University*

he last 60 years have seen a dramatic change in may have experienced greater consumption volatility as a

T result. 3 A nd contrary to the country irrelevance proposi-

world nancial markets. At the end of World

War II, the capital markets of most countries tion espoused by economists, the bulk of the evidence also

were separated from one another by barriers of suggests that asset prices, investor portfolios, and corporate

various kinds. For instance, many countries prohibited their nancial policies all continue to have signi cant country-

investors from owning foreign securities or obtaining the speci c components.

foreign currency to purchase them. But since 1945, many In this article, I argue that despite a dramatic increase

countries have sharply reduced such barriers to cross-border in cross-border trade in nancial assets, the positive impact

trade in nancial assets, giving rise to a development that is of nancial globalization has been surprisingly limited. My

often called nancial globalization. explanation for this puzzle is that the expected effects of

According to most economists, the globalization of increasingly global capital markets have been limited by

nancial markets is expected to have major economic what I call the twin agency problems. These problems

bene ts. In theory, it should lead to better sharing of risks stem from the reality that, in pursuing their own interests,

among investors worldwide, allow capital to ow where its both the rulers of sovereign states and corporate insiders

productivity is highest, and provide individual countries often take actions that reduce the value of companies to

with greater opportunity to reap the bene ts of their respec- outside investors.

tive comparative advantages.1 W hy twin problems rather than two distinct problems?

In the process, moreover, signi cant cross-country Because the problems tend to occur together and to feed on

differences in asset prices, investment portfolios, and corpo- one another. When one agency problem is well managed,

rate nancial policies are eventually expected to disappear, the other tends to be as well.

resulting in a condition that economists refer to as country The severity of these problems varies across countries,

irrelevance. According to this theoretical proposition, depending on the soundness of a country s institutions. At

investors in all countries will end up holding essentially one extreme are countries such as the U.K., Canada, and

the same world market portfolio, one that contains repre- the U.S., where the twin problems are limited by well-

sentative proportions of all global economies. And as a established legal, regulatory, and market institutions. At the

consequence, comparably pro table companies operating other extreme, countries like Russia provide minority share-

in the same industry but domiciled in different countries holders with little protection against expropriation by either

ought to have similar capital structures and roughly the controlling shareholders or the state. In such countries, I

same cost of capital and market valuations. show why corporations have a tough time attracting invest-

But the evidence in support of these predictions is at best ment capital unless corporate insiders co-invest heavily

mixed. While some studies have reported a positive effect with outside suppliers of equity.

of nancial globalization on growth,2 most of the evidence One result of such co-investment is a highly concentrated

to date suggests limited bene ts. Consider, for example, ownership structure. While such ownership can certainly

the following conclusion of a 2003 study of developing have bene ts for example, it has clearly played a critical

countries by the International Monetary Fund: [W]hile role in the successes of LBOs, venture capital, and other

there is no proof in the data that nancial globalization forms of private equity it also has two negative effects that

has increased growth, there is evidence that some countries can limit economic growth and the development of nancial

* This is an edited transcript of the author s presidential address to the American Fi- 2. For instance, Geert Bekaert, Campbell R. Harvey and Christian Lundblad, Does

nance Association in Philadelphia on January 6, 2005. Financial Liberalization Spur Growth?, Journal of Financial Economics, Vol. 77 No. 1

1. For a review of such advantages, see my article Globalization, Corporate Finance, (2005), conclude that a country s equity market liberalization leads to an increase in

and the Cost of Capital, Journal of Applied Corporate Finance, Vol. 12 No. 3 (Fall economic growth of 1%.

1999), 8-25. 3. See Eswar Prasad, Kenneth Rogoff, Shang-Jin Wei, and M. Ayhan Kose, Effects of

Financial Globalization on Developing Countries: Some Empirical Evidence, Interna-

tional Monetary Fund, (2003).

8 Journal of Applied Corporate Finance Volume 19 Number 1 A Morgan Stanley Publication Winter 2007

Figure 1 Gross Cross-Border Securities Transactions for the U.S. (Sum of

Purchases and Sales) to U.S. GDP (2006 Number Estimated from January-October Data)

Some Evidence on Financial Globalization and

markets. First, it limits risk-sharing by forcing insiders to

Country Relevance

bear far more rm-speci c risk than if they held diversi ed

portfolios. Second, it means that a company s investment is Since the end of World War II, there has been dramatic

limited by the wealth of its insiders, who can co-invest only progress in reducing formal barriers to trading nancial

to the extent they have the resources to do so. assets. In every year since 1950, the International Monetary

A s mentioned earlier, much of the expected bene t Fund has published information on restrictions on interna-

of nancial globalization is assumed to come from the tional nancial transactions. An index of openness

expanded risk-sharing that results from attracting inves- compiled and updated by Dennis Quinn shows that such

tors from different countries with different risk exposures. restrictions have almost completely disappeared in industri-

The willingness of such investors to pay a diversi cation alized countries and have decreased considerably in

developing countries.4

premium for a foreign company s shares is expected, all

other things equal, to reduce the rm s cost of capital and As can be seen in Figure 1, the volume of transactions

thus stimulate investment. Unfortunately, however, in in stocks and bonds between U.S. and foreign investors

most international settings all other things are not equal. increased dramatically from 1977 through 2006. During

In countries that pose considerable risks for investors of that period, the volume of such transactions as a percent-

expropriation by controlling investors and the state, the age of GDP increased by a factor larger than 60, from just

demand for co-investment effectively means that most of under 6% to over 365%. By comparison, the dollar volume

the risk of the rm must be borne by corporate insiders, of transactions on the NYSE grew from just over 7% of

not by well-diversi ed international investors. And this not GDP to about 130% of GDP, or by a factor of 19.

only limits the gains from nancial globalization for many With such a dramatic increase in cross-border trade

nations, but creates the possibility that, in countries where in securities and the disappearance of many formal barri-

such problems are particularly acute, the globalization of ers to international investment, we would have expected

nancial markets can at times end up reducing grow th. The

rowth. The institutional and other differences among countries to

massive capital out ows experienced by a number of such have steadily diminishing effects on both investors and

countries during periods of crisis could not have happened companies investment and nancing decisions. But it has

had they not opened up their markets. not happened. According to a number of recent studies,

4. Dennis Quinn, The Correlates of Change in International Financial Regulation,

American Political Science Review 91 (1997), 531-551.

Journal of Applied Corporate Finance Volume 19 Number 1 A Morgan Stanley Publication Winter 2007 9

Figure 2 Value-Weighted Average of the Percentage of Shares Held by

Corporate Insiders for 48 Countries in 2002

countries continue to have a major in uence on asset the amount of promising growth opportunities within a

returns, investment portfolios, and corporate nancial country rather than the availability of funds. While there

policies. are signs of a decoupling of investment from saving in some

countries, recent studies still nd a surprisingly strong

The evidence shows that countries continue to be impor-

tant for portfolio choice. Investors continue to overweight correlation. For instance, one study of emerging markets

domestic securities in their portfolios. Although such home during the 1990s the period when such markets began to

bias has decreased over time, a study of global investor take off showed that the fraction of investment nanced

portfolios showed that, as of the end of 2004, the share of by foreign investors did not increase at all during that

decade.7

foreign equities in U.S. portfolios was roughly one-fourth

of what it would have been had investors chosen to hold the Countries continue to be important for consumption.

global market portfolio, with holdings in all the world s In a fully integrated world, consumption within a given

national economies in proportion to the size of their equity country would not be tightly linked to, and would be

markets.5 more stable than, the country s income. To the extent they

could, investors would effectively buy insurance against

Countries continue to be important for saving and invest-

ment. In a study produced more than 20 years ago, Martin country-speci c risks by holding internationally diversi ed

Feldstein and Charles Horioka showed an extremely high portfolios and using hedging and risk-transfer mechanisms.

correlation between domestic saving and investment.6 This Economists describe this phenomenon as consumption

gave rise to what became known among economists as the risk-sharing. Recent studies, however, nd at most a limited

Feldstein-Horioka puzzle. The puzzle is this: As investors increase in consumption risk-sharing over time.

diversify internationally, saving (which depends on domes- Countries continue to be important for corporate owner-

tic income and wealth) and investment (which depends ship. The composition of corporate ownership varies

on local growth opportunities) should become less closely systematically across countries. In a widely cited 1999 study,

related to each other. Investment should be determined by R afael La Porta, Florencio Lopez-de-Silanes, and Andrei

5. Bong-Chan Kho, Ren M. Stulz, and Frank Warnock, Financial Globalization, Gov- 7. See Joshua Aizenman, Brian Pinto, and Artur Radziwill, (2004), Sources for Fi-

ernance, and the Home Bias, Unpublished working paper, The Ohio State University. nancing Domestic Capital Is Foreign Saving a Viable Option for Developing Countries?,

6. Martin Feldstein and Charles Horioka, Domestic Saving and International Capital National Bureau of Economic Research, Working Paper 10624.

Flows, Economic Journal, Vol. 90 (1980), p. 314-329.

10 Journal of Applied Corporate Finance Volume 19 Number 1 A Morgan Stanley Publication Winter 2007

Shleifer showed that, in most countries other than the U.S. delphia almost 3,000 miles away. But distance in this case

and U.K., the representative company has a concentrated is clearly at most a second-order effect. That San Diego and

ownership structure.8 Tijuana are separated by a border is much more important

Figure 2 reports the median fraction of market capital- than their proximity to one another. And if distance has

ization held by blockholders in large companies representing limited explanatory power in accounting for differences

48 countries at the end of 2002.9 The median ownership in investor portfolios, it has even less to say about cross-

fraction for the 48 countries was just over 50%. The U.S. country differences in, say, corporate capital structures. In

was at the left tail of the gure, with a median percent- this case, the differences have nothing to do with distance

age of market capitalization held by blockholders of just and everything to do with borders.

over 15.5%. And the median percentage of blockholdings

Why Countries Might Matter:

in U.K. companies, at 10.31%, shows a similarly dispersed

The Twin Agency Problems

ownership structure. By contrast, in a country like Italy,

that fraction was over 40%. The key difference between explanations that rely on

distance and those that rely on borders is the role of the

Countries continue to be important for capital structure.

Studies have found that corporate capital structures vary sovereign state. There are two main reasons why the twin

widely among countries. For instance, after studying the agency problems depend critically on the state: First, the

capital structures of public companies in 39 developed and sovereign state affects the extent of the agency problem

developing countries, Joseph Fan, Sheridan Titman, and between corporate insiders and other investors because it is

Gary Twite concluded that a corporation s capital structure a critical determinant of the extent of investor protection.

is determined more by the country in which it is located Second, although the rulers of the sovereign state can expro-

than by its industry af liation. 10 More speci cally, they priate investors, their powers to do so stop at the border.

found that companies in countries with weak governance To illustrate the implications of the twin agency

institutions have signi cantly higher leverage than compa- problems for corporate ownership and risk-bearing, let s

rable rms in countries that provide greater protection for examine a simple example in which there are assumed to be

investors a nding I will discuss later. t wo kinds of investors: entrepreneurs and portfolio inves-

tors. An entrepreneur has a unique investment opportunity.

Countries continue to be important for governance.

Countries explain a very large part of the variation in the If she takes advantage of that opportunity, she starts a rm,

quality of governance across companies. In a recent study, sells equity to outside investors, and becomes a corporate

Craig Doidge, Andrew Karolyi, and I found that country insider. But if she decides to pass on the investment oppor-

characteristics explained more than 70% of the variation tunity, she becomes a portfolio investor.

in companies S&P corporate governance scores.11 Thus, in Let s consider now how the twin agency problems affect

countries with signi cant agency problems, even companies the entrepreneur and investors. Since the corporate insid-

with highly effective governance systems are likely to nd ers observe the cash ow produced by the company before

their capital choices limited by where they are domiciled anybody else, they have the ability to expropriate some of it

(although, as I will argue later, volunteering to list on an before it comes to the attention of outsiders or representa-

overseas exchange can help address this problem). tives of the state. The rulers of the sovereign state are next

Economists have worked hard to explain why countries in line; they can con scate cash ows outright, tax them,

matter as much as they do. There are more than 100 studies redirect them to their favorite causes or allies, and so on.

that attempt to explain the home bias of portfolios, with The cash ow left after expropriation by corporate insiders

explanations centering on variables such as transportation and the rulers of the sovereign state can then be (and let s

costs, transaction costs, consumption preferences, and assume for our example that it is) distributed to sharehold-

differences in information among investors. But this litera- ers as a liquidating dividend.

ture fails to be convincing. Much of it focuses, for example, Corporate insiders can transfer value from outside

on differences that are attributable to distance. A s such investors to themselves by consuming private bene ts in

reasoning goes, investors who are far away may be less well relatively benign ways, such as buying fancy airplanes, or

informed about securities than investors who are on the spot. through more extreme ways, including outright theft. But

But there is an obvious problem with this type of argument: it s important to keep in mind here that the corporate insid-

According to the distance school of thought, portfolios of ers, precisely because of their ability to expropriate wealth from

investors in San Diego should be more like the portfolios of minority holders, are limited in the amount of equity they

investors in Tijuana than those held by investors in Phila- can raise from outside investors. When pricing the equity

8. Rafael La Porta, Florencio Lopez-de-Silanes, and Andrei Shleifer, Corporate Own- 10. Joseph P.H. Fan, Sheridan Titman, and Garry Twite, An International Comparison

ership Around the World, Journal of Finance, Vol. 54 (1999). of Capital Structure and Debt Maturity Choices, Unpublished working paper, University

9. Using data reported in Worldscope. of Texas, (2006).

Journal of Applied Corporate Finance Volume 19 Number 1 A Morgan Stanley Publication Winter 2007 11

offering, such investors assume that expropriation will take the rm s investment and growth stemming from the limits

place. And given that the shares are priced to re ect such of insiders wealth and ability to co-invest.

expected expropriation, insiders may have little choice but W hat role does the state play in this corporate gover-

to expropriate; indeed, they can be viewed as having already nance problem? A signi cant threat of expropriation by state

paid for the right to expropriate in the discount at which

or the right to expropriate in the discount at which rulers also encourages co-investment by corporate insiders.

the shares are sold. By expropriation I mean the (presumably self-interested)

The problem with such an equilibrium, however, is actions by state rulers that have the effect of reducing the

that when they routinely expropriate minority shareholders, returns of corporate investors. Although the case of Yukos

insiders also incur deadweight costs in the form of risk comes to mind, the term expropriation covers a wide

of embarrassment and possible exposure to legal sanctions. range of activities. Besides con scating assets, state rulers

The size of such deadweight costs varies widely among can impose redistributive taxes, prohibit or limit value-

countries. In countries such as the U.S. and U.K., expro- adding activities (such as laying off redundant workers),

priation is costly for insiders and so outside investors expect favor friends and allies through permits and regulation, or

little of it to take place. But in countries where expropria- demand bribes.

tion is expected, insiders can sell shares at a higher price (or, Such activities are, of course, by no means unknown

a lternatively, reduce the deadweight costs associated with in developed countries. In the U.S., we have evidence that

the possibility of expropriation) only by co-investing more the stock prices of S&P 500 companies with Republi-

with outside investors. As insiders increase their investment can-connected board members reacted positively to the

in a company, they have less to gain from expropriating Republican win in the election of 2000 and S&P 500 rms

with Democratic board connections responded negatively.12

minority shareholders.

To understand why, consider a simple example. Assume And the widespread interference by the state and labor unions

that the state expropriates 10% of a rm s cash ows after the in most European countries has clear effects on how corpo-

rations are owned and organized.13 But however common

controlling shareholders have already taken their cut in the

form of private bene ts. In that case, the outside sharehold- in developed nations, state intervention in business enter-

ers receive dividends of 90 cents from each dollar of cash prise is likely to be even more pronounced, and destructive,

ow net of private bene ts. And let s start by supposing that in developing countries.

insiders own just 10% of the rm. In this case, by taking one And such intervention has indirect as well as direct effects

dollar of private bene ts, insiders would reduce the liquidat- on corporate governance and behavior. One fairly direct effect

ing dividend by 90 cents, of which their share would be just is the heightened incentives of corporate insiders to reduce the

9 cents. As a consequence, the net gain to insiders of one state s proceeds from expropriation by increasing their own

dollar of private bene ts before deadweight costs would be bargaining power with the state. The corporate insiders can

91 cents in this case and insiders would be expected to accomplish this by building connections with the rulers and by

refrain from consuming this dollar of private bene ts only if making themselves more critical to the success of their compa-

the associated deadweight costs exceeded 91 cents. nies through their choice of investment, contracting, and

But now consider what happens when insiders own nancing policies. But by making themselves more indispens-

50% of the company. In this case, each dollar of private able to a rm, insiders also become less subject to discipline by

bene ts causes insiders to forgo 45 cents of dividends and the market for corporate control and their outside sharehold-

they accordingly stop consuming private bene ts when the ers and so their incentive and ability to expropriate value

marginal deadweight cost reaches 55 cents. Thus, the net increases, along with the associated deadweight costs. Thus

gain to insiders of a dollar of private bene ts is reduced by the net result of such managerial entrenchment may be that

increases in their ownership share. By co-investing more, the gains to minority shareholders from the reduced likeli-

insiders reduce their own incentives to extract private hood of expropriation by the state are completely offset by the

bene ts. And the result of higher co-ownership is less expro- losses from expropriation by insiders.

priation of value by insiders, and thus higher values for all Furthermore, when management has a small owner-

shareholders. ship stake, it is more likely to use its entrenched position to

But, as already mentioned, there are two major costs extract the best deal from the state for itself instead of using

associated with additional co-investment by insiders: the its power to protect shareholders. This in turn means that

higher cost of capital that results from corporate insiders when the expected costs of state expropriation are signi -

bearing risks they cannot diversify, and the constraint on cant, the governance problems inherent in the separation of

11. Craig Doidge, G. Andrew Karolyi, and Ren M. Stulz, Why do Countries Matter 13. See Mark Roe, Political Determinants of Corporate Governance, Oxford Univer-

so Much for Corporate Governance?, Journal of Financial Economics, forthcoming. sity Press, (2003).

12. Eitan Goldman, J rg Rocholl, and Jongil So, Does Political Connectedness Affect

Firm Value?, Unpublished Working Paper, University of North Carolina, (2006).

12 Journal of Applied Corporate Finance Volume 19 Number 1 A Morgan Stanley Publication Winter 2007

ownership and control are likely to become less manageable in rm value detected by these studies seem disappoint-

ing.17 The twin agency problems can help us understand

than when governments do not expropriate.

In sum, corporate insiders with large ownership stakes why the impact of nancial liberalization might fall short

are more likely to make decisions that limit the potential of expectations.

for state expropriation. And their large ownership stakes W hen cross-country barriers to nancial trading are

can also work to assure outside investors that the insiders lowered, part of the anticipated increase in shareholder

will make decisions that bene t all investors. Absent such

nvestors. Absent such wealth is attributable to the expected pro ts from new

a stake, they might be tempted to let the state expropri- investments that can be undertaken only if more capital is

apita l is

ate minority shareholders at will in exchange for the right invested. But because of the twin agency problems, much of

granted by the state to appropriate more private bene ts. the increase in capital investment would have to come from

insiders who may already be constrained by limits on their

How the Twin Agency Problems Blunt the wealth and by their inability to diversify their holdings. In

Effect of Financial Globalization this fashion, the requirement for concentrated ownership

I ve discussed each of the two agency problems separately, could lead to far lower corporate investment than might be

but now let s consider brie y how they feed on each other. expected under a governance system that encouraged more

First, to the extent corporate insiders are connected with dispersed ownership.

state rulers, they can use the state to transfer value from other Even if foreign investors could be persuaded to take

investors. As Raghu Rajan and Luigi Zingales show in their larger ownership stakes in overseas companies, the result-

recent book Saving Capitalism from the Capitalists,14 business ing decrease in ownership by corporate insiders could well

incumbents routinely use laws and regulations to tilt the lead, at least in the near term (and for reasons already

playing eld in their favor, thus hindering the development discussed), to greater expropriation o

Copyright © 2007 Morgan Stanley. All rights reserved. No part of this publi-

entitlement to exemption.

Contact this candidate