Faculty Bibliography

The correlation between governance indices and abnormal returns documented for 1990-1999 subsequently disappeared. The correlation and its disappearance are both due to market participants' gradually learning to appreciate the difference between good-governance and poor-governance firms. Consistent with learning, the correlation's disappearance was associated with increases in market participants' attention to governance; market participants and security analysts were, until the beginning of the 2000s but not subsequently, more positively surprised by the earning announcements of good-governance firms; and, although governance indices no longer generated abnormal returns during the 2000s, their negative association with firm value and operating performance persisted.

The Securities and Exchange Commission (SEC) is currently considering a rulemaking petition requesting that the Commission shorten the ten-day window, established by Section 13(d) of the Williams Act, within which investors must publicly disclose purchases of a five percent or greater stake in public companies. In this Article, we provide the first systematic empirical evidence on these disclosures and find that several of the petition’s factual premises are not consistent with the evidence. Our analysis is based on about 2,000 filings by activist hedge funds during the period of 1994–2007. We find that the data are inconsistent with the petition’s key claim that changes in market practices and technologies have operated over time to increase the magnitude of pre-disclosure accumulations, making existing rules “obsolete” and therefore requiring the petition’s proposed “modernization.” The median stake that these investors disclose in their 13(d) filings has remained stable throughout the 17-year period that we study, and regression analysis does not identify changes over time in the stake disclosed by investors. We also find that: * A substantial majority of 13(d) filings are actually made by investors other than activist hedge funds, and these investors often use a substantial part of the ten-day window before disclosing their stake. * A significant proportion of poison pills have low thresholds of 15% or less, so that management can use 13(d) disclosures to adopt low-trigger pills to prevent any further stock accumulations by activists — a fact that any tightening of the SEC’s rules in this area should take into account. * Even when activists wait the full ten days to disclose their stakes, their purchases seem to be disproportionately concentrated on the day they cross the threshold and the next day; thus, the practical difference in pre-disclosure accumulations between the existing regime and the rules in jurisdictions with shorter disclosure windows is likely much smaller than the petition assumes. * About ten percent of 13(d) filings seem to be made after the ten-day window has expired; the SEC may therefore want to consider tightening the enforcement of existing rules before examining the proposed acceleration of the deadline. Our analysis provides new empirical evidence that should inform the SEC’s consideration of this subject — and a foundation on which subsequent empirical and policy analysis can build.

This report provides an overview and analysis of the work that the Shareholder Rights Project (SRP) undertook on behalf of a number of institutional investors during 2012 and 2013, the SRP’s first two years of operations. During 2012 and 2013, the SRP worked on behalf of eight SRP-represented investors on board declassification proposals submitted for a vote at the 2012 and/or 2013 annual meetings of 122 S&P 500 and Fortune 500 companies, and this work has produced substantial results: 100 Negotiated Outcomes: Negotiated outcomes involving a commitment to board declassification were reached with 100 S&P 500 and Fortune 500 companies, about three-quarters of the companies receiving proposals in 2012 and/or 2013. 58 Successful Precatory Proposals: During 2012 and 2013, declassification proposals brought by SRP-represented investors received majority support at 58 annual meetings of 53 S&P 500 and Fortune 500 companies (all but three of the annual meetings in which such proposals went to a vote), with average support of about 80% of votes cast. 81 Board declassifications: A total of 81 S&P 500 and Fortune 500 companies already declassified their boards during 2012 and 2013 as a result of the work of the SRP and SRP-represented investors. These 81 companies, which have an aggregate market capitalization exceeding one trillion dollars (as of Dec, 31, 2013), represent about 65% of the companies with which engagements took place and about 60% of the S&P 500 companies that had classified boards as of the beginning of 2012. Expected Impact by End of 2014: The work of the SRP and SRP-represented investors is expected to produce a significant number of additional board declassifications during 2014 as a result of (i) management declassification proposals that will go to a vote pursuant to 2012 and 2013 agreements, (ii) companies agreeing to follow the preferences of shareholders expressed in 58 successful precatory declassification proposals, and (iii) ongoing engagement by the SRP and SRP-represented investors. We estimate that, by the end of 2014, this work will have contributed to movements towards board declassification by about 100 S&P 500 and Fortune 500 companies; this large-scale change can be expected to increase board accountability and thereby to enhance shareholder value and company performance in the affected companies. Beyond Board Declassification: The SRP’s 2012 and 2013 work also facilitated a substantial increase in successful engagement by public pension funds, and in their ability to obtain governance changes favored by shareholders. The proposals that the SRP worked in 2012 and 2013 on represented over 50% of the shareholder proposals by public pension funds that received majority support in 2012 and 2013, and over 20% of all precatory shareholder proposals (by all proponents) that received majority support in 2012 and 2013. The Shareholder Rights Project (SRP) is a clinical program operating at Harvard Law School and directed by Professor Lucian Bebchuk. The SRP works on behalf of public pension funds and charitable organizations seeking to improve corporate governance at publicly traded companies, as well as on research and policy projects related to corporate governance. Any views expressed and positions taken by the SRP and its representatives should be attributed solely to the SRP and not to Harvard Law School or Harvard University.

With contributions from some of the leading scholars in law and economics, this comprehensive book summarizes the state of economic research on litigation, procedure and evidence.

The Securities and Exchange Commission is currently considering a rulemaking petition that advocates tightening the rules under the Williams Act, which regulates the disclosure of large blocks of stock in public companies. In this Article, we explain why the Commission should not view the proposed tightening as a merely “technical” change needed to meet the objectives of the Williams Act, provide market transparency, or modernize its regulations. The drafters of the Williams Act made a conscious choice not to impose an inflexible 5% cap on pre-disclosure accumulations of shares to avoid deterring investors from accumulating large blocks of shares. We argue that the proposed changes to the SEC’s rules should similarly be examined in the larger context of the optimal balance of power between incumbent directors and these blockholders. We discuss the beneficial and documented role that outside blockholders play in corporate governance and the adverse effect that any tightening of the Williams Act’s disclosure thresholds can be expected to have on such blockholders. We explain that there is currently no evidence that trading patterns and technologies have changed in ways that would make it desirable to tighten these disclosure thresholds. Furthermore, since the passage of the Williams Act, the rules governing the balance of power between incumbents and outside blockholders have already moved significantly in favor of the former — both in absolute terms and in comparison to other jurisdictions — rather than the latter. Our analysis provides a framework for the comprehensive examination of the rules governing outside blockholders that the Commission should pursue. In the meantime, we argue, the Commission should not adopt new rules that would tighten the disclosure thresholds that apply to blockholders. Existing research and available empirical evidence provide no basis for concluding that such tightening would protect investors and promote efficiency. Indeed, there is a good basis for concern that such tightening would harm investors and undermine efficiency.

While staggered boards have been documented to be negatively correlated with firm valuation, such association might be due to staggered boards either bringing about lower firm value or merely reflecting the tendency of low-value firms to have staggered boards. In this paper, we use two natural experiments to shed light on the causality question. In particular, we focus on two recent court rulings, separated by several weeks, that affected in opposite directions the antitakeover force of staggered boards: (i) a ruling by the Delaware Chancery Court approving the legality of shareholder-adopted bylaws that weaken the antitakeover force of a staggered board by moving the company's annual meeting up from later parts of the calendar year to January, and (ii) the subsequent decision by the Delaware Supreme Court to overturn the Chancery Court ruling and invalidate such bylaws. We find evidence consistent with the hypothesis that the Chancery Court ruling increased the value of affected companies - namely, companies with a staggered board and an annual meeting in later parts of the calendar year - and that the Supreme Court ruling produced a reduction in the affected companies' value. The identified effects were most pronounced for firms for which control contests are especially relevant due to relative underperformance, small firm size, high asset pledgibility, or high takeover intensity in their industry. Our findings have implications for the long-standing debate on staggered boards. The findings are consistent with the market's viewing staggered boards as bringing about a reduction in firm value. Our findings are thus consistent with leading institutional investors' policies in favor of board de-staggering, and with the view that the ongoing process of board de-staggering in public firms can be expected to enhance shareholder value.

In some cases, the law permits a party that unilaterally provides a benefit to another party to recover the estimated value of this benefit. Despite calls for expanding the set of cases to which such a restitution rule applies, the law commonly applies a mutual consent rule under which a party providing another with a benefit cannot obtain any recovery without securing the advance consent of the beneficiary to the transaction. We provide an efficiency rationale for the undesirability of broad use of the restitution rule by identifying significant adverse ex ante effects of the rule that are avoided by the consent requirement. Even assuming that courts' errors in estimating buyer benefits would be unbiased, a restitution rule would strengthen sellers' hand by providing them with a put option that they may but do not have to use. As a result, the restitution rule would encourage inefficient market entry by low-quality sellers that would not contribute to any efficient transactions but would be able to extract payments from buyers seeking to avoid an exchange with them. Furthermore, the restitution rule would discourage efficient market entry by some or all potential buyers of a good or service. Beyond the restitution rule, we extend our analysis to show that similar adverse effects can also arise from other "pricing" rules that provide buyers or sellers with call or put options to force an exchange at a judicially-determined price.

Firms, investors, and regulators around the world are now seeking to ensure that the compensation of public company executives is tied to long-term results, in part to avoid incentives for excessive risk taking. This Article examines how best to achieve this objective. Focusing on equity-based compensation, the primary component of executive pay, we identify how such compensation should best be structured to tie pay to long-term performance. We consider the optimal design of limitations on the unwinding of equity incentives, putting forward a proposal that firms adopt both grant-based and aggregate limitations on unwinding. We also analyze how equity compensation should be designed to prevent the gaming of equity grants at the front end and the gaming of equity dispositions at the back end. Finally, we emphasize the need for widespread adoption of limitations on executives’ use of hedging and derivative transactions that weaken the tie between executive payoffs and the long-term stock price that well-designed equity compensation is intended to produce.

House Committee on Financial Services

Companies, investors, and regulators around the world are now seeking to tie executives’ payoffs to long-term results and avoid rewarding executives for short-term gains. Focusing on equity-based compensation, the primary component of top executives’ pay, the authors analyze how such compensation should best be structured to provide executives with incentives to focus on long-term value creation. To improve the link between equity compensation and long-term results, the authors recommend that executives be prevented from unwinding their equity incentives for a significant time period after vesting. At the same time, however, the authors suggest that it would be counterproductive to require that executives hold their equity incentives until retirement, as some have proposed. Instead, the authors recommend that companies adopt a combination of “grant-based” and “aggregate” limitations on the unwinding of equity incentives. Grant-based limitations would allow executives to unwind the equity incentives associated with a particular grant only gradually after vesting, according to a fixed, pre-specified schedule put in place at the time of the grant. Aggregate limitations on unwinding would prevent an executive from unloading more than a specified fraction of the executive’s freely disposable equity incentives in any given year. Finally, the authors emphasize the need for effective limitations on executives’ use of hedging and derivative transactions that would weaken the connection between executive payoffs and long-term stock values that a well-designed equity arrangement should produce.

This paper seeks to make three contributions to understanding how banks’ executive pay has produced incentives for excessive risk-taking and how such pay should be reformed. First, although there is now wide recognition that pay packages focused excessively on short-term results, we analyze a separate and critical distortion that has received little attention. Equity-based awards, coupled with the capital structure of banks, tie executives’ compensation to a highly levered bet on the value of banks’ assets. Because bank executives expect to share in any gains that might flow to common shareholders, but are insulated from losses that the realization of risks could impose on preferred shareholders, bondholders, depositors, and taxpayers, executives have incentives to give insufficient weight to the downside of risky strategies. Second, we show that corporate governance reforms aimed at aligning the design of executive pay arrangements with the interests of banks’ common shareholders - such as advisory shareholder votes on compensation arrangements, use of restricted stock awards, and increased director oversight and independence -cannot eliminate the identified problem. In fact, the interests of common shareholders could be served by more risk-taking than is socially desirable. Accordingly, while such measures could eliminate risk-taking that is excessive even from shareholders’ point of view, they cannot be expected to prevent risk-taking that serves shareholders but is socially excessive. Third, we develop a case for using regulation of banks’ executive pay as an important element of financial regulation. We provide a normative foundation for such pay regulation, analyze how regulators should monitor and regulate bankers’ pay, and show how pay regulation can complement and reinforce the traditional forms of financial regulation.

This paper contains the proceedings of the Proxy Access Roundtable that was held by the Harvard Law School Program on Corporate Governance on October 7, 2009. The Roundtable brought together prominent participants in the debate – representing a range of perspectives and experiences – for a day of discussion on the subject. The day’s first two sessions focused on the question of whether the Securities and Exchange Commission should provide an access regime, or whether it should leave the adoption of access arrangements, if any, to private ordering on a company-by-company basis. The third session focused on how a proxy access regime should be designed, assuming the Securities and Exchange Commission were to adopt such an access regime. The final session went beyond proxy access and focused on whether there are any further changes to the arrangements governing corporate elections that should be considered.

The standard narrative of the meltdown of Bear Stearns and Lehman Brothers assumes that the wealth of the top executives at these firms was largely wiped out along with their firms. In the ongoing debate about regulatory responses to the financial crisis, commentators have used this assumed fact as a basis for dismissing both the role of compensation structures in inducing risk-taking and the potential value of reforming such structures. This Article provides a case study of compensation at Bear Stearns and Lehman Brothers during 2000-2008 and concludes that this assumed fact is incorrect. We find that the top-five-executive teams at these firms cashed out large amounts of performance-based compensation during this period. From 2000-2008, they were able to cash out large amounts of bonus compensation that were not clawed back when the firms collapsed, and to pocket large amounts from selling shares. Overall, we estimate that the top executive teams of Bear Stearns and Lehman Brothers derived cash flows of about $1.4 billion and $1 billion, respectively, from cash bonuses and equity sales during 2000-2008. These cash flows substantially exceeded the value of the executives' initial holdings at the beginning of the period, and the executives' net payoffs for the period were thus decidedly positive. The divergence between how the top executives and their shareholders fared implies that it is not possible to rule out, as standard narratives suggest, that the executives' pay arrangements provided them with excessive risk-taking incentives. We discuss the implications of our analysis for understanding the possible role that pay arrangements have played in the run-up to the financial crisis and how they should be reformed going forward.

Treasury Secretary Geithner announced a plan, which the Treasury is willing to finance with up to $1 trillion of public funds, to partner with private capital to buy banks’ “troubled assets.” The Treasury has not yet settled on the plan’s design, and its announcement has encountered substantial skepticism as to whether an effective plan for a public-private partnership in buying troubled assets can be worked out. This paper argues that, yes, it can. The paper also analyzes how the plan should be designed to contribute most to restarting the market for troubled assets at the least cost to taxpayers. The government’s plan should focus on establishing a significant number of competing funds that will be privately managed and dedicated to buying troubled assets – not on creating one, large public-private aggregator bank. Establishing competing funds, I show, is necessary both to securing a well-functioning market for troubled assets and to keeping costs to taxpayers at a minimum. Each new fund will be partly financed with private capital, with the rest coming (say, in the form of non-recourse debt financing) from the government’s Investment Fund planned by the Treasury. One important element of the proposed design is a competitive process in which private managers seeking to establish a fund participating in the program will submit bids as to what fraction of the fund’s capital will be funded privately. The government will set the fraction of each participating fund’s capital that must be financed with private money at the highest level that, given the received bids, will still enable establishing new funds with aggregate capital equal to the program’s target level. Overall, I show that the proposed design will leverage private capital to the fullest extent possible and will provide the most effective and least costly mechanism for restarting the market for troubled assets.

We investigate the relative importance of the twenty-four provisions followed by the Investor Responsibility Research Center (IRRC) and included in the Gompers, Ishii, and Metrick governance index (Gompers, Ishii, and Metrick 2003). We put forward an entrenchment index based on six provisions: staggered boards, limits to shareholder bylaw amendments, poison pills, golden parachutes, and supermajority requirements for mergers and charter amendments. We find that increases in the index level are monotonically associated with economically significant reductions in firm valuation as well as large negative abnormal returns during the 1990-2003 period. The other eighteen IRRC provisions not in our entrenchment index were uncorrelated with either reduced firm valuation or negative abnormal returns.

This Essay analyzes how government intervention in the market for banks' troubled assets is best designed, and also uses this analysis to evaluate the public-private investment program announced by the U.S. government in March 2009. I begin by presenting the case for using government funds to restart the market for troubled assets. I then discuss the advantages of providing government capital to competing privately managed funds--a strategy I have advocated in past work--and I outline the key elements that such a plan should include.

Despite the large infusion of capital into the financial sector and low interest rates, the flow of financing to operating firms has failed to return to normal levels. One explanation suggested is that banks still lack confidence because they need time to adjust to the new environment. Another is that the reduced flow of credit reflects banks' rational assessment of borrowers' bleak prospects in the current economic conditions. This paper puts forward a third explanation. When the prospects of operating firms are interdependent, banks' decisions on whether to lend to a given operating firm depend on their expectations about whether other operating firms will receive financing. As a result, an inefficient credit market freeze may arise in which banks avoid lending to operating firms due to their self-fulfilling expectations that other banks will not be lending. Such an inefficient freeze may persist even when banks have ample capital and interest rates are brought down to a barely positive level. Facing an inefficient and persistent credit freeze, I argue, the government should consider getting the economy out of it by taking upon itself the credit risks involved in providing a substantial amount of new lending to operating firms. This can be accomplished by (i) providing banks with non-recourse financing for portfolios of new loans to operating firms, or otherwise agreeing to bear part of the risks generated by such portfolios, in return for a share of the upside, or (ii) setting up government-funded, privately managed funds dedicated to making such loans. These mechanisms funds can contribute substantially, both directly and indirectly, to producing a credit thaw.

We investigate the relationship between CEO centrality -- the relative importance of the CEO within the top executive team in terms of ability, contribution, or power -- and the value and behavior of public firms. Our proxy for CEO centrality is the fraction of the top-five compensation captured by the CEO. We find that CEO centrality is negatively associated with firm value (as measured by industry-adjusted Tobin's Q). Greater CEO centrality is also correlated with (i) lower (industry-adjusted) accounting profitability, (ii) lower stock returns accompanying acquisitions announced by the firm and higher likelihood of a negative stock return accompanying such announcements, (iii) higher odds of the CEO's receiving a "lucky" option grant at the lowest price of the month, (iv) greater tendency to reward the CEO for luck in the form of positive industry-wide shocks, (v) lower likelihood of CEO turnover controlling for performance, and (vi) lower firm-specific variability of stock returns over time. Overall, our results indicate that differences in CEO centrality are an aspect of firm management and governance that deserves the attention of researchers.

While prior empirical work and much public attention have focused on the opportunistic timing of executives' grants, we provide in this paper evidence that outside directors' option grants have also been favorably timed to an extent that cannot be fully explained by sheer luck. Examining events in which public firms granted options to outside directors during 1996-2005, we find that 9% were "lucky grant events" falling on days with a stock price equal to a monthly low. We estimate that about 800 lucky grant events owed their status to opportunistic timing, and that about 460 firms and 1400 outside directors were associated with grant events produced by such timing. There is evidence that the opportunistic timing of director grant events has been to a substantial extent the product of backdating and not merely spring-loading based on private information. We find that directors' luck has been correlated with executives' luck. Furthermore, grant events were more likely to be lucky when the firm had more entrenching provisions protecting insiders from the risk of removal, as well as when the board did not have a majority of independent directors.

We investigate the distribution of pay in the top executive team in public companies. In particular, we study the CEO's pay slice (CPS), defined as the fraction of the aggregate top-five total compensation paid to the CEO. A firm's CPS might reflect the relative significance of the CEO – in terms of ability, contribution to the firm, or power – relative to other members of the top executive team. We find that CPS has been going up over the past decade. During this period, CEOs have increased their fraction of both equity-based compensation and non-equity compensation. The level of CPS is associated with various characteristics of the top team and the firm’s governance arrangements. Among other things, CPS is high when the CEO has long tenure; when the CEO chairs the board; when few other executives are members of the board; and when the firm has more entrenching provisions. High CPS is associated with lower firm value as measured by Tobin's Q. Using a simultaneous equations approach yields findings consistent with the possibility that this negative correlation is at least partly due to high CPS, or the factors that it reflects, bringing about a lower Tobin's Q. High CPS is also associated with a reduction in the sensitivity of CEO turnover to performance. This is the case especially in firms with high entrenchment levels. Overall, our results indicate that the distribution of compensation in the top executive team is an aspect of pay arrangements and corporate governance that is worthy of financial economists' attention.

This piece provides the brief submitted to the Delaware Chancery Court by plaintiff in the case of Bebchuk vs. CA, Inc. The case concerns the attempt by CA to exclude from the corporate ballot a stockholder proposal to adopt a proposed bylaw concerning the use of poison pills on grounds that it would be invalid under Delaware law. Under the proposed bylaw, a board would not be able to maintain a pill indefinitely without periodic determination than doing so would serve shareholders. The brief explains why, in contrast to the claims made by CA, the proposed bylaw would neither violate any of the provisions of the Delaware corporate code nor impedes directors' fulfillment of their fiduciary duties. Although the court decided that the case was not sufficiently ripe for a ruling on the validity of the proposed bylaw, the court's decision rejected CA's arguments that the bylaw is clearly invalid.

This paper shows that "one-sided" terms in standard contracts, which deny consumers a contractual benefit that seems efficient on average, may arise in competitive markets without informational problems (other than those of courts). A one-sided term might be an efficient response to situations in which courts cannot perfectly observe all the contingencies needed for an accurate implementation of a "balanced" contractual term when firms are more concerned about their reputation, and thus less inclined to behave opportunistically, than consumers are. We develop this explanation, discuss its positive and normative implications, and compare them to those of information-based explanations for one-sided terms.

This article contains the edited transcript of a forum on personal liability of directors held at Harvard Law School in November 2005. Eleven panelists offer their diverse views and perspectives on this subject. The participants in the panel, which is moderated by Lucian Bebchuk, are: Joseph Bachelder, Roel Campos, Byron Georgiou, Alan Hevesi, William Lerach, Robert Mendelsohn, Robert Monks, Toby Myerson, John Olson, Leo Strine, Jr., and John Wilcox.

The analysis of this paper was subsequently combined with that of our companion paper “Lucky Directors,” http://ssrn.com/abstract=952239. The combined paper, titled “Lucky CEOs and Lucky Directors,” is available on SSRN at http://ssrn.com/abstract=1405316 and was published at Journal of Finance, Vol. 65, No. 6, pp. 2363-2401, 2010. We study the relation between corporate governance and opportunistic timing of CEO option grants. Investigating the incidence of lucky grants - defined as grants given at the lowest price of the month - we estimate that about 1,150 lucky grants resulted from opportunistic timing, and that 12% of firms provided one or more lucky grant due to opportunistic timing during the period 1996-2005. We find no evidence that opportunistically timed grants served as a substitute for other forms of compensation; indeed, total reported compensation from other sources was higher (relative to peer companies) in firms providing lucky grants. For any given CEO with two or more grants, grants were more likely to be lucky when they took place in months in which the potential payoffs from opportunistic timing were relatively high. Grants were also more likely when the company did not have a majority of independent directors on the board and/or the CEO had longer tenure, both factors that are associated with increased influence of the CEO on pay-setting and board decision-making. Luck was persistent, with a CEO's chance of getting a lucky grant increasing when a preceding grant was lucky as well. Finally, we find that opportunistic timing was present in each of the economy's 12 (Fama-French) industries, and we do not find evidence that it was significantly driven by industry norms and culture. Because our analysis suggests that the existence of lucky CEO grants is a variable that can be useful to research studying firms’ governance and decision-making, we make available on the website of the Harvard Program on Corporate Governance a dataset of CEO luck indicators based on our work.

Because public firms are not required to disclose the monetary value of pension plans in their executive pay disclosures, financial economists have generally analyzed executive pay using figures that do not include the value of such pension plans. This paper presents evidence that omitting the value of pension benefits significantly undermines the accuracy of existing estimates of executive pay, its variability, and its sensitivity to performance companies. Studying the pension arrangements of CEOs of S&P 500, we find that the CEOs' plans had a median actuarial value of $15 million; that the ratio of the executives' pension value to the executives' total compensation (including both equity and non-equity pay) during their service as CEO had a median value of 34%; and that including pension values increased the median percentage of the executives' total compensation composed of salary-like payments during and after their service as CEO from 15% to 39%.

We study the extent to which decisions to expand firm size are associated with increases in subsequent CEO compensation. Controlling for past stock performance, we find a positive correlation between CEO compensation and the CEO's past decisions to increase firm size. This correlation is economically meaningful; for example, other things being equal, CEOs who in the preceding three years were in the top quartile in terms of expanding by increasing the number of shares outstanding receive compensation that is higher by one-third than the compensation of CEOs belonging to the bottom quartile. We also find that stock returns are correlated with subsequent CEO pay only to the extent that they contribute to expanding firm size; only the component of past stock returns not distributed as dividends is correlated with subsequent CEO pay. Finally, we find an asymmetry between increases and decreases in size: while increases in firm size are followed by higher CEO pay, decreases in firm size are not followed by reduction in such pay. The association we find between CEOs' compensation and firm-expanding decisions undertaken earlier during their service could provide CEOs with incentives to expand firm size.

We study the extent to which decisions to expand firm size are associated with increases in subsequent CEO compensation. Investigating a broad universe of firm-expansion choices, we find, controlling for performance and firm characteristics, a positive and economically meaningful correlation between CEO compensation and the CEO's past decisions to increase firm size. We demonstrate that the identified correlation is not driven by large corporate acquisitions, and that it remains significant and economically meaningful when firms making large acquisitions during the relevant period are excluded. We further find an asymmetry between size increases and decreases: while size increases are positively correlated with subsequent CEO pay, size decreases are not negatively correlated with subsequent CEO pay. The identified association between expansion decisions and subsequent CEO pay is relevant for assessing CEO incentives with respect to a broad range of choices made by firms.