1.1.1 Moral Haz s in Many Guises ways in which ers)be h ts n ot
16 1. Corporate Governance France) and institutional investors (such as CalPERS in the United States) started enunciating codes of best practice for boards of directors. More recently, various laws and reports1 came in reaction to the many corporate scandals of the late 1990s and early 2000s (e.g., Seat, Banesto, Metallgesellschaft, Suez, ABB, Swissair, Vivendi in Europe, Dynergy, Qwest, Enron, WorldCom, Global Crossing, and Tyco in the United States). But what is corporate governance?2 The dominant view in economics, articulated, for example, in Shleifer and Vishny’s (1997) and Becht et al.’s (2002) surveys on the topic, is that corporate governance relates to the “ways in which the suppliers of finance to corporations assure themselves of getting a return on their investment.” Relatedly, it is preoccupied with the ways in which a corporation’s insiders can credibly commit to return funds to outside investors and can thereby attract external financing. This definition is, of course, narrow. Many politicians, managers, consultants, and academics object to the economists’ narrow view of corporate governance as being preoccupied solely with investor returns; they argue that other “stakeholders,” such as employees, communities, suppliers, or customers, also have a vested interest in how the firm is run, and that these stakeholders’ concerns should somehow be internalized as well.3 Section 1.8 will return to the debate about the stakeholder society, but we should indicate right away that the content of this book reflects the agenda of the narrow and orthodox view described in the above citation. The rest of Section 1.1 is therefore written from the perspective of shareholder value. 1. In the United States, for example, the 2002 Sarbanes–Oxley Act, and the U.S. Securities and Exchange Commission’s and the Financial Accounting Standards Board’s reports. 2. We focus here on corporations. Separate governance issues arise in associations (see Hansmann 1996; Glaeser and Shleifer 2001; Hart and Moore 1989, 1996; Kremer 1997; Levin and Tadelis 2005) and government agencies (see Wilson 1989; Tirole 1994; Dewatripont et al. 1999a,b). 3. A prominent exponent of this view in France is Albert (1991). To some extent, the German legislation mandating codetermination (in particular, the Codetermination Act of 1976, which requires that supervisory boards of firms with over 2,000 employees be made up of an equal number of representatives of employees and shareholders, with the chairperson—a representative of the shareholders—deciding in the case of a stalemate) reflects this desire that firms internalize the welfare of their employees. 1.1.1 Moral Hazard Comes in Many Guises There are various ways in which management may not act in the firm’s (understand: its owners’) best interest. For convenience, we divide these into four categories, but the reader should keep in mind that all are fundamentally part of the same problem, generically labeled by economists as “moral hazard.” (a) Insufficient effort. By “insufficient effort,” we refer not so much to the number of hours spent in the office (indeed, most top executives work very long hours), but rather to the allocation of work time to various tasks. Managers may find it unpleasant or inconvenient to cut costs by switching to a less costly supplier, by reallocating the workforce, or by taking a tougher stance in wage negotiations (Bertrand and Mullainathan 1999).4 They may devote insuffi- cient effort to the oversight of their subordinates; scandals in the 1990s involving large losses inflicted by traders or derivative specialists subject to insuf- ficient internal control (Metallgesellschaft, Procter & Gamble, Barings) are good cases in point. Lastly, managers may allocate too little time to the task they have been hired for because they overcommit themselves with competing activities (boards of directors, political involvement, investments in other ventures, and more generally activities not or little related to managing the firm). (b) Extravagant investments. There is ample evidence, both direct and indirect, that some managers engage in pet projects and build empires to the detriment of shareholders. A standard illustration, provided by Jensen (1988), is the heavy exploration spending of oil industry managers in the late 1970s during a period of high real rates of interest, increased exploration costs, and reduction in expected future oil price increases, and in which buying oil on Wall Street was much cheaper than obtaining it by drilling holes in the ground. Oil industry managers also invested some of their large amount of cash into noncore industries. Relatedly, economists have long conducted event studies to analyze the reaction of stock prices to the announcement 4. Using antitakeover laws passed in a number of states in the United States in the 1980s and firm-level data, Bertrand and Mullainathan find evidence that the enactment of such a law raises wages by 1–2%
nd h ng. e risk at ce of moral hazard is lar nd am de ts.Yet, ues be ke it hard fe d)e
1.1. Introduction: The Separation of Ownership and Control 17 of acquisitions and have often unveiled substantial shareholder concerns with such moves (see Shleifer and Vishny 1997; see also Andrade et al. (2001) for a more recent assessment of the long-term acquisition performance of the acquirer–target pair). And Blanchard et al. (1994) show how firms that earn windfall cash awards in court do not return the cash to investors and spend it inefficiently. (c) Entrenchment strategies. Top executives often take actions that hurt shareholders in order to keep or secure their position. There are many entrenchment strategies. First, managers sometimes invest in lines of activities that make them indispensable (Shleifer and Vishny 1989); for example, they invest in a declining industry or old-fashioned technology that they are good at running. Second, they manipulate performance measures so as to “look good” when their position might be threatened. For example, they may use “creative” accounting techniques to mask their company’s deteriorating condition. Relatedly, they may engage in excessive or insufficient risk taking. They may be excessively conservative when their performance is satisfactory, as they do not want to run the risk of their performance falling below the level that would trigger a board reaction, a takeover, or a proxy fight. Conversely, it is a common attitude of managers “in trouble,” that is, managers whose current performance is unsatisfactory and are desperate to offer good news to the firm’s owners, to take excessive risk and thus “gamble for resurrection.” Third, managers routinely resist hostile takeovers, as these threaten their long-term positions. In some cases, they succeed in defeating tender offers that would have been very attractive to shareholders, or they go out of their way to find a “white knight” or conclude a sweet nonaggression pact with the raider. Managers also lobby for a legal environment that limits shareholder activism and, in Europe as well as in some Asian countries such as Japan, design complex cross-ownership and holding structures with double voting rights for a few privileged shares that make it hard for outsiders to gain control. (d) Self-dealing. Lastly, managers may increase their private benefits from running the firm by engaging in a wide variety of self-dealing behaviors, ranging from benign to outright illegal activities. Managers may consume perks5 (costly private jets,6 plush offices, private boxes at sports events, country club memberships, celebrities on payroll, hunting and fishing lodges, extravagant entertainment expenses, expensive art); pick their successor among their friends or at least like-minded individuals who will not criticize or cast a shadow on their past management; select a costly supplier on friendship or kinship grounds; or finance political parties of their liking. Self-dealing can also reach illegality as in the case of thievery (Robert Maxwell stealing from the employees’ pension fund, managers engaging in transactions such as below-market-price asset sales with affiliated firms owned by themselves, their families, or their friends),7 or of insider trading or information leakages to Wall Street analysts or other investors. Needless to say, recent corporate scandals have focused more on self-dealing, which is somewhat easier to discover and especially demonstrate than insufficient effort, extravagant investments, or entrenchment strategies. 1.1.2 Dysfunctional Corporate Governance The overall significance of moral hazard is largely understated by the mere observation of managerial misbehavior, which forms the “tip of the iceberg.” The submerged part of the iceberg is the institutional response in terms of corporate governance, finance, and managerial incentive contracts. Yet, it is worth reviewing some of the recent controversies regarding dysfunctional governance; we take the United States as our primary illustration, but the universality of the issues bears emphasizing. Several forms of dysfunctional governance have been pointed out: Lack of transparency. Investors and other stakeholders are sometimes imperfectly informed about 5. Perks figure prominently among sources of agency costs in Jensen and Meckling’s (1976) early contribution. 6. Personal aircraft use is one of the most often described perks in the business literature. A famous example is RJR Nabisco’s fleet of 10 aircraft with 36 company pilots, to which the chief executive officer (CEO) Ross Johnson’s friends and dog had access (Burrough and Helyar 1990). 7. Another case in point is the Tyco scandal (2002). The CEO and close collaborators are assessed to have stolen over $100 million
4 The t nd200.n2 19821 ed pay:top n 04a
18 1. Corporate Governance the levels of compensation granted to top management. A case in point is the retirement package of Jack Welch, chief executive officer (CEO) of General Electric.8 Unbeknownst to outsiders, this retirement package included continued access to private jets, a luxurious apartment in Manhattan, memberships of exclusive clubs, access to restaurants, and so forth.9 The limited transparency of managerial stock options (in the United States their cost for the company can legally be assessed at zero) is also a topic of intense controversy.10 To build investor trust, some companies (starting with, for example, Boeing, Amazon.com, and Coca-Cola) but not all have recently chosen to voluntarily report stock options as expenses. Perks11 are also often outside the reach of investor control. Interestingly, Yermack (2004a) finds that a firm’s stock price falls by an abnormal 2% when firms first disclose that their CEO has been awarded the aircraft perk.12 Furthermore, firms that allow personal aircraft use by the CEO underperform the market by about 4%. Another common form of perks comes from recruiting practices; in many European countries, CEOs hire family and friends for important positions; this practice is also common in the United States.13 Level. The total compensation packages (salary plus bonus plus long-term compensation) of top executives has risen substantially over the years and reached levels that are hardly fathomable to the 8. Jack Welch was CEO of General Electric from 1981 to 2001. The package was discovered only during divorce proceedings in 2002. 9. Similarly, Bernie Ebbers, WorldCom’s CEO borrowed over $1 billion from banks such as Citigroup and Bank of America against his shares of WorldCom (which went bankrupt in 2001) and used it to buy a ranch in British Columbia, 460,000 acres of U.S. forest, two luxury yachts, and so forth. 10. In the United States grants of stock options are disclosed in footnotes to the financial statements. By the mid 1990s, the U.S. Congress had already prevented the Financial Accounting Standards Board from forcing firms to expense managerial stock options. 11. Such as Steve Jobs’s purchase of a $90 million private jet. 12. As Yermack stresses, this may be due to learning either that corporate governance is weak or that management has undesirable characteristics (lack of integrity, taste for not working hard, etc.). See Rajan and Wulf (2005) for a somewhat different view of perks as enhancing managerial productivity. 13. Retail store Dillard’s CEO succeeded in getting four of his children onto the board of directors; Gap’s CEO hired his brother to redesign shops and his wife as consultant. Contrast this with Apria Healthcare: in 2002, less than 24 hours after learning that the CEO had hired his wife, the board of directors fired both. public.14 The trend toward higher managerial compensation in Europe, which started with lower levels of compensation, has been even more dramatic. Evidence for this “runaway compensation” is provided by Hall and Liebman (1998), who report a tripling (in real terms) of average CEO compensation between 1980 and 1994 for large U.S. corporations,15 and by Hall and Murphy (2002), who point at a further doubling between 1994 and 2001. In 2000, the annual income of the average CEO of a large U.S. firm was 531 times the average wage of workers in the company (as opposed to 42 times in 1982).16 The proponents of high levels of compensation point out that some of this increase comes in the form of performance-related pay: top managers receive more and more bonuses and especially stock options,17 which, with some caveats that we discuss later, have incentive benefits. Tenuous link between performance and compensation. High levels of compensation are particularly distressing when they are not related to performance, that is, when top managers receive large amounts of money for a lackluster or even disastrous outcome (Bebchuk and Fried 2003, 2004). While executive compensation will be studied in more detail in Section 1.2, let us here list the reasons why the link between performance and compensation may be tenuous. First, the compensation package may be poorly structured. For example, the performance of an oil company is substantially affected by the world price of oil, a variable over which it has little control. Suppose that managerial bonuses and stock options are not indexed to the price of oil. Then the managers can make enormous amounts of money when the price of oil increases. By contrast, they lose little from the lack of indexation when the price of oil 14. For example, in 1997, twenty U.S. CEOs had yearly compensation packages over $25 million. The CEO of Traveler’s group received $230 million and that of Coca-Cola $111 million. James Crowe, who was not even CEO of WorldCom, received $69 million (Business Week, April 20, 1998). 15. Equity-based compensation rose from 20 to 50% of total compensation during that period. 16. A New Era in Governance, McKinsey Quarterly, 2004. 17. For example, in 1979, only 8% of British firms gave bonuses to managers; more that three-quarters did in 1994. The share of performance-based rewards for British senior managers jumped from 10 to 40% from 1989 to 1994 (The Economist, January 29, 1994, p. 69)
so million golde the ve d by bring We have already al may hy $ and Philip that
1.1. Introduction: The Separation of Ownership and Control 19 plummets, since their options and bonuses are then “out-of-the money” (such compensation starts when performance—stock price or yearly profit—exceeds some threshold), not to mention the fact that the options may be repriced so as to reincentivize executives. Thus, managers often benefit from poor design in their compensation schemes. Second, managers often seem to manage to maintain their compensation stable or even have it increased despite poor performance. In 2002, for example, the CEOs of AOL Time Warner, Intel, and Safeway made a lot of money despite a bad year. Similarly, Qwest’s board of directors awarded $88 million to its CEO despite an abysmal performance in 2001. Third, managers may succeed in “getting out on time” (either unbeknownst to the board, which did not see, or did not want to see, the accounting manipulations or the impending bad news, or with the cooperation of the board). Global Crossing’s managers sold shares for $735 million. Tenet Health Care’s CEO in January 2002 announced sensational earnings prospects and sold shares for an amount of $111 million; a year later, the share price had fallen by 60%. Similarly, Oracle’s CEO (Larry Ellison) made $706 million by selling his stock options in January 2001 just before announcing a fall in income forecasts. Unsurprisingly, many reform proposals have argued in favor of a higher degree of vesting of managerial shares, forcing top management to keep shares for a long time (perhaps until well after the end of their employment),18 and of an independent compensation committee at the board of directors. Finally, managers receive large golden parachutes19 for leaving the firm. These golden parachutes are often granted in the wake of poor performance (a major cause of CEO firing!). These high golden parachutes have been common for a long time in the United States, and have recently made 18. The timing of exercise of executives’ stock options is documented in, for example, Bettis et al. (2003). They find median values for the exercise date at about two years after vesting and five years prior to expiration. 19. Golden parachutes refer to benefits received by an executive in the event that the company is acquired and the executive’s employment is terminated. Golden parachutes are in principle specified in the employment contract. their way to Europe (witness the $89 million golden parachute granted to ABB’s CEO). The Sarbanes–Oxley Act (2002) in the United States, a regulatory reaction to the previously mentioned abuses, requires the CEO and chief financial officer (CFO) to reimburse any profit from bonuses or stock sales during the year following a financial report that is subsequently restated because of “misconduct.” This piece of legislation also makes the shares held by executives less liquid by bringing down the lag in the report of sales of executive shares from ten days to two days.20 Accounting manipulations. We have already alluded to the manipulations that inflate company performance. Some of those manipulations are actually legal while others are not. Also, they may require cooperation from investors, trading partners, analysts, or accountants. Among the many facets of the Enron scandal21 lie off-balance-sheet deals. For example, Citigroup and JPMorgan lent Enron billions of dollars disguised as energy trades. The accounting firm Arthur Andersen let this happen. Similarly, profits of WorldCom (which, like Enron, went bankrupt) were assessed to have been overestimated by $7.1 billion starting in 2000.22 Accounting manipulations serve multiple purposes. First, they increase the apparent earnings and/or stock price, and thereby the value of managerial compensation. Managers with options packages may therefore find it attractive to inflate earnings. Going beyond scandals such as those of Enron, Tyco, Xerox,23 and WorldCom in the United States and Parmalat in Europe, Bergstresser and Philippon (2005) find more generally that highly incentivized CEOs exercise a large number of stock options during years 20. See Holmström and Kaplan (2003) for more details and an analysis of the Sarbanes–Oxley Act, as well as of the NYSE, NASDAQ, and Conference Board corporate governance proposals. 21. For an account of the Enron saga and, in particular, of the many off-balance-sheet transactions, see, for example, Fox (2003). See also the special issue of the Journal of Economic Perspectives devoted to the Enron scandal (Volume 12, Spring 2003). 22. Interestingly, one WorldCom director chaired Moody’s investment services, and it took a long time for the rating agency to downgrade WorldCom. 23. A restatement by the Securities and Exchange Commission reduced Xerox’s reported net income by $1.4 billion over the period 1997–2001. Over that period, the company’s CEO exercised options worth over $20 million
rge fr he明 uves:An ov
20 1. Corporate Governance in which discretionary accruals form a large fraction of reported earnings, and that their companies engage in higher levels of earnings management. Second, by hiding poor performance, they protect managers against dismissals or takeovers or, more generally, reduce investor interference in the managerial process. Third, accounting manipulations enable firms not to violate bank covenants, which are often couched in terms of accounting performance.24 Lastly, they enable continued financing.25 When pointing to these misbehaviors, economists do not necessarily suggest that managers’ actual behavior exhibits widespread incompetency and moral hazard. Rather, they stress both the potential extent of the problem and the endogeneity of managerial accountability. They argue that corporate governance failures are as old as the corporation, and that control mechanisms, however imperfect, have long been in place, implying that actual misbehaviors are the tip of an iceberg whose main element represents the averted ones. 1.2 Managerial Incentives: An Overview 1.2.1 A Sophisticated Mix of Incentives However large the scope for misbehavior, explicit and implicit incentives, in practice, partly align managerial incentives with the firm’s interest. Bonuses and stock options make managers sensitive to losses in profit and in shareholder value. Besides these explicit incentives, less formal, but quite powerful implicit incentives stem from the managers’ concern about their future. The threat of being fired by the board of directors or removed by the market for corporate control through a takeover or a proxy fight, the possibility of being replaced by a receiver (in the United Kingdom, say) or of being put on a tight leash (as is the case of a Chapter 11 bankruptcy in the United States) during financial distress, and the prospect of being appointed to new boards of directors or of receiving offers for executive directorships in more prestigious companies, all contribute to keeping managers on their toes. 24. See Section 2.3.3 for a discussion of covenants. 25. For example, WorldCom, just before bankruptcy, was the second-largest U.S. telecommunications company, with 70 acquisitions under its belt. Capital market monitoring and product-market competition further keep a tight rein on managerial behavior. Monitoring by a large institutional investor (pension fund, mutual fund, bank, etc.), by a venture capitalist, or by a large private owner restricts managerial control, and is generally deemed to alleviate the agency problem. And, as we will discuss, product-market competition often aligns explicit and implicit managerial incentives with those of the firm, although it may create perverse incentives in specific situations. Psychologists, consultants, and personnel officers no doubt would find the economists’ description of managerial incentives too narrow. When discussing incentives in general, they also point to the role of intrinsic motivation, fairness, horizontal equity, morale, trust, corporate culture, social responsibility and altruism, feelings of self-esteem (coming from recognition or from fellow employees’ gratitude), interest in the job, and so on. Here, we will not enter the debate as to whether the economists’ view of incentives is inappropriately restrictive.26 Some of these apparently noneconomic incentives are, at a deeper level, already incorporated in the economic paradigm.27 As for the view that economists do not account for the possibility of benevolence, it should be clear that economists are concerned with the study of the residual incentives to act in the firm’s interests over and beyond what they would contribute in the absence of rewards and monitoring. While we would all prefer not to need this sophisticated set of 26. For references to the psychology literature and for views on how such considerations affect incentives, see, for example, Bénabou and Tirole (2003, 2004, 2005), Camerer and Malmendier (2004), Fehr and Schmidt (2003), and Frey (1997). 27. For example, explicit or implicit rules mandating “fairness” and “horizontal equity” can be seen as a response to the threat of favoritism, that is, of collusion between a superior and a subordinate (as in Laffont 1990). The impact of morale can be partly apprehended through the effects of incentives on the firm’s or its management’s reputation (see, for example, Tirole 1996). And the role of trust has in the past twenty years been one of the leitmotivs of economic theory since the pioneering work of Kreps et al. (1982) (see, for example, Kreps 1990). Economists have also devoted some attention to corporate culture phenomena (see Carrillo and Gromb 1999; Crémer 1993; Kreps 1990). Economists may not yet have a fully satisfactory description of fairness, horizontal equity, morale, trust, or corporate culture, but an a priori critique of the economic paradigm of employee incentives as being too narrow is unwarranted, and more attention should be devoted to exactly what can and cannot be explained by the standard economic paradigm