NBER Reporter OnLine: Fall 2004PDF Version (includes NBER Profiles, Conferences, News, and Books)
NBER Program Reports
Subscribe to the hard copy NBER Reporter.
Program Report: Corporate Finance
The NBER's Program on Corporate Finance has a strong and dedicated core group and, in its brief existence (since 1991), has initiated some very promising avenues of research. Narrowly interpreted, corporate finance is the study of the investment and financing policies of corporations. But, since firms are at the center of economic activity, and because almost any topic of concern to economists --from microeconomic issues like incentives and risk sharing to macroeconomic issues such as currency crises -- affects corporate financing and investment, it is increasingly hard to draw precise boundaries around the field.
The range of subjects that group members have addressed in their research also reflects this difficulty. In fact, some of the most interesting work in corporate finance now is being done at its interface with other areas. Here I have chosen a set of our papers, because there are far too many for me to describe all of them, that fall into fairly coherent subject areas. The order in which I describe the subjects loosely follows from micro to macro: dividend policy to international finance.
Given that the study of dividend policy is as old as the modern field of finance (recall the Miller-Modigliani work on dividends), it might seem surprising that there is something left to say about it. Yet, although the questions remain the same, we have new hypotheses and new or better evidence on old ones.
Brav et al. (9657) survey Chief Financial Officers and Treasurers of companies to determine key factors driving dividend policy. They find the traditional behavioral patterns: managers are reluctant to cut dividends, prefer to smooth dividends over time, and tie dividend increases to long-run sustainable earnings. But they are also more willing to use stock repurchases nowadays. The authors also conclude that managers give only moderate weight to traditional tax, agency, and clientele theories of dividend payout.
Other papers, however, suggest that either managers responding to these surveys do not articulate well what they do, or they respond to cues that they do not fully understand. It seems that tax, agency, and clientele rationales are alive and well in the data. Chetty and Saez (10572) test the tax theory by asking whether dividend payments increased after the individual income tax on dividends was cut in 2003. They find that more firms initiated dividends for the first time and that many firms increased the dividends they already paid. This finding is robust to the usual controls. While others have found similar responses to tax changes in the past, the fact that the long decline in dividend payment in the United States seems to have turned around, and for a traditional reason, is particularly interesting.
Desai, Foley, and Hines (8698) examine the dividend policies of foreign affiliates of U.S. multinational firms. They find that not only are they determined by tax considerations, but also by agency considerations: foreign affiliates that are only partly owned, located far from the United States, or in areas where property rights are weak, typically pay more in dividends (presumably because they cannot be trusted to keep the cash, given the parent's weak control). DeAngelo, DeAngelo, and Stulz (10599) argue that if firms did not pay out dividends, they would sit on a mountain of cash with attendant incentives to waste it. They find that firms with large amounts of retained earnings tend to pay dividends, even after one controls for their profitability and growth.
Finally, Baker and Wurgler (9542) find that firms tend to initiate dividends when the demand for dividends is high, as measured, for example, by the difference between the market-to-book ratio of dividend paying firms and non-dividend paying firms. They suggest that there are fluctuations in investor sentiment about dividends, and that firms cater to this. Of course, what they term investor sentiment may well be time-varying concerns about agency or taxes (as would occur, for example, if firms built up cash piles during cyclical upturns and ran them down in downturns). The authors do a number of tests to rule it out. Nevertheless, one could still have questions about the findings: if indeed investors become enthused about dividends when sentiment is high, then it is surprising that firms do not raise the aggregate payout ratio. However, this is a novel explanation that deserves further investigation.
Many battles have been fought over capital structure: whether firms indeed have a target capital structure that they adhere to fairly strictly; whether firms have high costs of issuing equity which they factor into decisions about how close they should be to the target; and finally whether firms are simply buffeted by market forces and do not really bother about capital structure. Welch (8782) takes the last view, and shows that the ratio of debt to market value of assets for firms is determined strongly by past equity returns and little else. One could take issue with whether debt-to-market-value is the appropriate measure of capital structure, but Welch offers some arguments in support. Kayhan and Titman (10526) soften Welch's basic finding by arguing that even though history (for example, through past movements in the stock price) tends to influence capital structure changes, the effects eventually are revers ed, and firms do tend to make financing choices that move them towards target debt ratios.
Stock Market and Investment
The recent boom and bust in the stock market, and evidence of excessive investment in certain sectors like telecommunications, has led some to ask if we should revisit the received wisdom that the stock market is a sideshow to real activity. Polk and Sapienza (10563) find that overpriced firms do tend to overinvest, and then tend to have low stock returns. Gilchrist, Himmelberg, and Gur (10537) argue that stock prices rise above fundamentals when investor beliefs are more dispersed, and short-selling constraints prevent the most pessimistic among them from registering their vote. They find that firms with more dispersed investor beliefs have higher new equity issues as well as investment. Both of these papers suggest that high stock prices push managers into investing by reducing their cost of finance. One problem with this interpretation is that high stock prices also may be signaling the valu e of future opportunities, and this may be why firms invest. Baker, Stein, and Wurgler (8750) find a clever way to tell these two explanations apart: they rank firms on whether they rely on equity for financing or not. If it is the abnormally low cost of financing that pushes managers to invest, then the investment of equity-dependent firms should be far more sensitive to stock price changes than the investment of firms that are not dependent on equity for financing. They find that this is the case.
Shleifer and Vishny (8439) present a model to explain the ludicrous prices that were paid during the merger wave of the late 1990s. Why, for instance, would America Online pay so much for Time Warner? They argue that even if both bidder and target are overvalued in some long-run fundamental sense, the bidder may still go ahead, provided the market sees synergies in the merger, and the bidder itself is sufficiently overvalued. Moeller, Schlingemann, and Stulz (10200) find that the acquirers in the mergers from 1998 to 2001 lost a total of $240 billion on announcement, while the targets gained only $134 billion. Therefore, they argue, there was massive loss in these acquisitions, in part driven by a reassessment of the bidder's value. If this indeed were the case, one has to ask whether acquisitions truly were the most effective way for those acquiring managers sitting on paper wealth to convert it to real wealth, as the Shleifer Vishny model suggests. Could they not simply have issued shares and put the proceeds in the bank? Probably not, but this suggests that we need to understand better the pressures imposed by the market on managers.
Financial Market Frictions
The difficulty of raising external finance because markets do not know enough about the borrower, or cannot control it, is one of the most investigated topics in recent years. Typically, financing frictions can be identified by asking whether a firm's investment is related to its cash flow. A positive correlation between the two is taken as evidence that the firm cannot raise enough from the capital markets and thus is forced to invest only when it has cash. An alternative explanation, however, is that cash flow serves as a proxy for the quality of investment opportunities. So, it may be no surprise that there is a correlation. Hovakimian and Titman (9432) address this issue by looking at firms that conduct asset sales. These asset sales should provide cash for investment but should not necessarily be related to investment opportunities. They find that cash from asset sales is strongly related to investment, especially when a firm has the characteristics of f irms we typically think are liquidity constrained.
Taking a related but different tack, Almeida, Campello, and Weisbach (9253) argue that firms that are likely to be liquidity constrained should save a larger fraction of cash inflows, especially in times of economic adversity. They find this to be the case. Pinkowitz, Stulz, and Williamson (10188) point out that cash holdings may serve a precautionary need, but are also likely to be misused by management. They find that a dollar of cash translates to a value of a dollar of value for minority shareholders in countries with good investor protection but only 65 cents of value in countries with poor protection.
Although some firms may be constrained by markets, they may have access to special sources of financing. Fisman and Love (8960) argue that industries dependent on trade-credit financing rely less on formal markets and thus should grow faster in countries with weak financial systems. Desai, Foley, and Forbes (10545) point out that affiliates of multinationals still may have access to financing when a country undergoes a currency crisis, and thus should be able to invest significantly more than comparable firms during and after the crisis. Both papers find evidence consistent with their predictions.
Corporate Governance in the United States
Turning to corporate governance, Kaplan and Holmstrom (8220, 9613) take a broad look at U.S. corporate governance in the last two decades. They argue that the primary instrument of governance in the 1980s was hostile mergers and buyouts, while internal corporate governance mechanisms have played a much bigger role in the 1990s. Of course, recent corporate scandals do raise questions about the effectiveness of corporate governance in the United States. The authors do not see the problems as symptomatic of systemic failure -- they see U.S. corporations as performing favorably relative to corporations in other countries -- and argue that the regulatory, legislative, and market responses in all likelihood would deal quickly with the remaining problems. Of course, the entire credit for the performance of U.S. corporations over this period should not be attributed only to governance -- the favorable ma croeconomic environment in the United States over this period undoubtedly helped. Nevertheless, they offer a provocative argument to those who believe that managerial compensation has become unconscionable, and that U.S. corporate governance is broke.
Bebchuk, Fried, and Walker (8661) are in the latter camp. They feel that managerial compensation has become excessive, and much of it is rents extracted by powerful managers. The lack of any indexing of option grants to market indexes (so that manager are not simply rewarded for market-wide movements) is just one example of the practices they find egregious. Bertrand and Mullainathan (7604) in fact try to estimate how much managerial pay is for factors under managers' control and how much for luck. They find that executive pay in the oil industry increases substantially with oil prices, even though higher oil prices are, for all practical purposes, outside the control of the executive. Presumably, managerial compensation cannot be all good or bad. Rajan and Wulf (10494) examine the canonical symbol of managerial excess, the company plane. They find evidence that com pany planes are used where they have the most effect in enhancing the productivity of executives - for example, when the company is located far from a major airport. By contrast, they find little evidence that better governance diminishes perks in firms where they might be most egregious. They conclude that a blanket indictment of perks is unwarranted.
International Corporate Governance
How important is corporate governance across the world? Dyck and Zingales (8711) construct a measure of the private benefits of control (crudely, a measure of what the market thinks owners can skim from minority holders) in 39 countries. This ranges between 4 percent and 65 percent of the value of the firm. Capital markets are less developed, ownership is more concentrated, and fewer privatizations take place in countries where these private benefits are large. Interestingly, the authors find that measures like a stronger press, a high rate of tax compliance, and a high degree of product market competition have at least as much explanatory power for the level of private benefits as factors like the statutory protection of minority rights. The more general point seems to be that a range of institutions (and, more generally, popular awareness and support for them) seem to be important for good governance.
Bertrand, Mehta, and Mullainathan (7952) offer a nice way to get at the extent of misgovernance in Indian business groups. They argue that one way profits are siphoned out of firms is through pyramid structures. The owner of the firm at the top of the pyramid gets a large share of its dividends but only a small share of the dividends of the firm at the bottom of the pyramid, even though he may control it via the pyramid structure. Therefore, he has an incentive to divert profits from the firm at the bottom to the firm at the top via mechanisms like transfer pricing and possibly fraud. If this is so, then reported earnings in bottom firms should respond far less to positive changes in industry conditions (because a significant fraction of the additional profits are skimmed off to the top) than reported earnings of the firm at the top. Also, earnings for firms at the top should respond to increases in earnings for firms at the bottom but not vice versa (after t aking out the effect of any dividends going from the bottom firm to the top firm). The authors find these patterns in the data.
Finally, Caprio, Laeven and Levine (10158) examine the effects of governance structures on bank valuation around the world. They find that: 1) larger cash flow rights by the controlling owner boosts valuations; 2) stronger shareholder protection laws increase valuations; and 3) greater cash flow rights mitigate the adverse effects of weak shareholder protection laws on bank valuations.
Contracting and Organizational Structure
One important area of emerging study is the nature of organizations and the contracts that define them. Kaplan and Stromberg (7660,8202, 8764) study the contracts that venture capitalists write with entrepreneurs in the United States. They note how these contracts allocate cash flow rights and a variety of control rights separately. Typically, if the company performs poorly, the VC gets full control; otherwise he retains cash flow rights but gives up control rights. The nature of contingencies built into the contracts relate to the perceived risks associated with the venture, with greater risk generally leading to more rights for the venture capitalist. Lerner and Schoar (10348) analyze private equity transactions outside the United States. While transactions in common law countries seem similar to those in the United S tates, with greater use of contingencies and contingent instruments like preferred stock, investors in other countries have fewer contractual protections and tend to use uncontingent ownership, like common stock. These contractual differences have real consequences with larger, higher value transactions in the common law countries. These detailed empirical studies of contracting represent a major new advance in corporate finance, and verify as well as inform the theories.
Our researchers are also studying organizations. Rajan and Wulf (9633) find that large firms in the United States are adopting flatter organizational structures, with fewer levels between the CEO and divisions, and more direct reports to the CEO. These changes also are being reflected in pay, with steeper pay differentials in the flatter firms. They conjecture that these changes have to do with the changing nature and importance of human capital, and they find some consistent evidence.
Entrepreneurship and Ownership
How do firms start? What are the constraints on their growth? Rajan and Zingales (7546) argue that one fundamental concern for entrepreneurs is how to bring in employees and financiers to help generate rents while at the same time preventing them from expropriating those rents. For instance, employees can walk away with trade secrets. They develop a theory of the origins and growth of firm hierarchies which can explain stylized facts, such as why firms typically are started with family management (family members are more trusted to not expropriate, and are especially important when the firm is young and at its most vulnerable); why human-capital-intensive firms have flatter hierarchies with more ownership rights granted to successful employees; and why firms remain small in countries with weak property right protection. Burkart, Panunzi, and Shleifer (8776) develop a model of the evolution of the entrepreneurial firm in different legal environments and conclude that widely held professional corporations are most likely where there is strong legal protection of minority investors, while family succession is most likely when legal protection is weak.
Gompers, Lerner, and Scharfstein (9816) examine the factors that lead to venture capital start-ups. They examine two alternative views of this process: employees of established firms are trained to become entrepreneurs by coming into contact with other entrepreneurs and venture capitalists, or individuals become entrepreneurs because the firms they work for do not fund their ideas. They find the data to be more consistent with the first view.
Finally, Franks, Mayer, and Rossi (10628) and Khanna and Palepu (10613) examine the evolution of family ownership in the United Kingdom and India respectively. These are fascinating and careful studies that challenge the perceived wisdom that families in both countries were effete rent-seekers.
Stein (7705) offers an intriguing theory of hierarchies, in which large hierarchical firms are at a comparative disadvantage in processing soft information: in large firms, decisions have to be made by managers who are organizationally or geographically distant from the site where the information is gathered; and, soft information (such as whether a customer is trustworthy) does not travel well. Berger et al. (8752) test this theory with bank lending data and find that, as predicted, large banks tend to be less willing than small banks to lend to informationally "difficult" credits, including those who do not keep financial records, even after correcting for factors like the endogeneity of matching.
Durnev, Morck, and Yeung (8093) distinguish between industries that have greater firm-specific stock price variation and industries in which prices tend to move with the market. The former tend to use more external financing and allocate capital more precisely, suggesting that the market is able to better understand these firms, and perhaps guide their investment.
A number of papers examine the effect of physical distance on information. Garmaise and Moskowitz (8877) study the effect of information problems in the real estate market. They find that these problems are resolved by participants buying properties that are nearby, trading properties with long histories, and avoiding informed professional brokers. Petersen and Rajan (7685) find that the distance between banks and their borrowers has been increasing over time and suggest that this is consistent with greater and better use of information technology by banks. Finally, Guiso, Sapienza, and Zingales (8923) examine the effects of differences in local access to finance in Italy on the propensity to start businesses and grow them. They find that even local financial development matters for growth, suggesting that physical distance is still an important barrier for finance. P>
Liquidity has become an area of renewed focus in the banking literature. Diamond and Rajan (8937) argue that liquidity shortages can create a contagion of failures because bank failures themselves subtract liquidity from the market. Gorton and Huang (9158) argue that, while liquid assets are useful because they allow transactions to take place, private agents may supply too few of these assets. They argue that there is a role for the government in providing such assets, one example of which is government bailouts of banking systems. In a similar vein, Caballero and Krishnamurthy (7792) argue that companies in emerging markets have an incentive to underinsure against the shortage of foreign currency, which is why these companies are so willing to issue foreign currency debt despite the risks.
Empirical work confirms the importance of liquidity. Gatev and Strahan (9956) test the proposition that banks, being able to hedge liquidity demands well, are best able to offer liquidity support. In particular, they find that when the commercial paper market dries up, and spreads increase, banks experience inflows allowing them to offer back-up lines of credit to commercial paper issuers. Lerner and Schoar (9146) argue that private equity funds making long-run investments with high information asymmetries are likely to prefer deep-pocket investors who have little need for liquidity. Consistent with this hypothesis, they find that later funds organized by a firm (where information problems will be lower because of the firm's past record) have fewer transfer restrictions on investors. Similarly, funds investing in industries with longer investment cycles, such as pharmaceuticals, have more transfer constraints. Fin ally, investors who have long horizons, such as endowments, are less likely to have transfer constraints imposed on them.
La Porta, Lopez de Silanes, and Shleifer (7620) examine government ownership of banks around the world and find it associated with low levels of income, financial development, and productivity growth. While this is an indictment of government ownership of banks in developing countries, it is not clear that private ownership would be better. La Porta, Lopez de Silanes, and Zamarripa (8848) find that privatized banks in Mexico indulged in significant amounts of related lending, and that the default rates in such loans were significantly higher than in unrelated loans.
Carow, Kane, and Narayanan (10623) find that in megamergers, the large customers of the target are relatively unaffected, while the small customers of target firms fare especially badly on announcement of the merger. The effects are particularly pronounced for customers who show signs of being credit constrained. While this evidence is also consistent with the Stein (7705) hypothesis, the authors attribute it to changes in bargaining and monopoly power as a result of the merger. By contrast, Morgan and Strahan (9710) focus on some virtues of bank integration in the United States, finding that bank integration across U.S. states dampened economic volatility within those states. However, they do not find similar effects for international bank integration.
Desai, Foley, and Hines have written a number of papers exploiting the fact that when a multinational has affiliates in a number of countries, local conditions will affect the behavior of the affiliates differently. This work can be used to test theories. In 10337, for example, they examine the effects of local capital controls. Clearly, these will cause firms to shift profits towards the parent via transfer pricing: the reported profits for affiliates located in countries with capital controls indeed are significantly lower than for affiliates in other countries. Also, the local cost of capital is higher: affiliates in countries with capital controls face a 5.4 percent higher interest rate than the norm. Finally, multinationals invest less in countries with capital controls, and affiliates there are approximately 15 percent smaller.
Arslanalp and Henry (9369) examine the effects on the stock market of debt relief agreements under the Brady plan. They find an average appreciation of 60 percent in dollar terms, which is not explained by IMF agreements or liberalization. Instead, it appears that the stock market forecasts higher future net resource transfers and GDP growth, as would be suggested by debt-overhang theories. Chari and Henry (10318) find that capital account liberalizations do not draw in unthinking investors as some suggest, but rather investors who seem to allocate funds based on a firm's prospective cash flow and on the fact that the cost of capital in the country has fallen. However, investors do not seem to be drawn to firms that have benefited the most from a fall in the firm-specific risk premium.
The Effects of the Business Environment
La Porta, Lopez de Silanes, and Shleifer (9882) ask what aspects of securities law help the development of stock markets. They conclude that greater mandatory disclosure, together with a relatively low burden of proof on investors claiming improper or inadequate disclosure by issuers (that is public rules and private enforcement), tends to be associated with better stock market development. Of course, more disclosure is not always good. Gomes, Gorton, and Madureira (10567) find that the adoption of a rule intended to stop the practice of selective disclosure in the United States (where firms gave information ahead of public disclosure to a few analysts) resulted in a welfare loss for small firms because analysts stopped following them.
Djankov, La Porta, Lopez de Silanes, and Shleifer extend a very interesting literature on the connection between law and finance, begun by some of these authors and attempt to understand how the legal system (for example, common law versus civil law) actually matters. In (8890), they measure and describe the exact procedures used by litigants and courts to evict a tenant for non-payment of rent and to collect a bounced check. They use these data to construct an index of procedural formalism of dispute resolution for each country. They find that such formalism is systematically greater in civil than in common law countries. Moreover, procedural formalism is associated with higher expected duration of judicial proceedings, more corruption, less consistency, less honesty, less fairness in judicial decisions, and inferior access to justice.
Doidge, Karolyi, and Stulz (8538) ask why so few firms cross-list in the United States since it appears that those firms are valued more highly than comparable firms in domestic markets that do not cross-list. The authors conclude that firms that do not treat their minority shareholders well (and thereby trade at a discount) face costs in going to the better-policed U.S. markets. This is why much of the difference in valuation between cross-listed firms and firms that do not cross-list may simply be a matter of self-selection: the good firms tend to face fewer costs and greater benefits from cross listing. Reese and Weisbach (8164) do find that cross-listed firms seem to use the discipline of cross listing to raise more equity capital.
A number of papers study the effect of the business environment on firm creation. Desai, Gompers, and Lerner (10165) find that greater protection of property rights increases average entry rates, reduces exit rates, and reduces average firm size. Klapper, Laeven, and Rajan (10380) find that high bureaucratic barriers to entry hamper both entry and the growth in value added in naturally high-entry industries. They find that these entry barriers have little effect in corrupt countries; this suggests that an efficient and overweening bureaucracy is particularly detrimental for business. Fan and White (9340) argue that the Homestead exemption (by which individuals are allowed to shield a portion of their homes from creditors) gives entrepreneurs insurance against bad outcomes. Home-owning families are 35 percent more likely to own a business if they live in a high-exemp tion state than if they live in a low-exemption state. However, one cannot argue from this that the Homestead exemption expands access to credit. Indeed, it also should make it more difficult for any poor individual to buy a home or to raise money against it, as White indeed has shown in previous work. In (8852), Johnson, McMillan, and Woodruff ask whether stronger property rights or greater access to finance is more important. From a survey of new firms in post-communist countries, they conclude that weak property rights discourage firms from reinvesting profits even when bank loans are available, and thus have a greater adverse effect on growth.
Finally, what determines whether a country adopts proper rules regarding financial markets and competition? Countries seem to have experienced dramatic changes in their absolute and relative level of financial development over time; these are inconsistent with static explanations for the development of financial markets, such as their legal origin (for legal theories, see an excellent review by Beck and Levine (10126). In (8178), Rajan and Zingales argue that the time-varying incentives of the dominant interest groups in a country explain whether they are willing to allow finance to develop. Tracing financial development in a number of countries over the twentieth century, we provide evidence consistent with their conjectures.
It is not possible, given space limitations, to do justice to the range of issues our members are working on. I hope this sampling gives you a taste for more. You can access the full array of NBER working papers by the Corporate Finance Program at the NBER's web site.
1. Rajan directs the NBER's Program on Corporate Finance and is the Joseph L. Gidwitz Professor of Finance at the University of Chicago's Graduate School of Business. He is currently on leave and serving as the International Monetary Fund's Director of Research.