Journal of Financial Economics

Transcription

1 Journal of Financial Economics 102 (2011) Contents lists available at ScienceDirect Journal of Financial Economics journal homepage: Equity grants to target CEOs during deal negotiations $ Shane Heitzman n Simon School of Business, University of Rochester, Rochester, NY 14627, USA article info Article history: Received 6 November 2007 Received in revised form 4 January 2011 Accepted 3 February 2011 Available online 13 July 2011 JEL classification: G34 J33 K22 M52 abstract I investigate the determinants and consequences of granting equity to the target s Chief Executive Officer (CEO) during deal negotiations. These negotiation grants likely reflect information about the acquisition, benefit from the deal premium, and provide more timely bargaining incentives. I find that CEOs are more likely to receive equity during negotiations when they negotiate for the target, particularly when the target has more bargaining power. This suggests that boards use equity to enhance bargaining incentives for CEOs with the most influence over deal price. I find limited evidence that negotiation grants are used as compensation and no evidence that they have a material adverse effect on shareholders. & 2011 Elsevier B.V. All rights reserved. Keywords: Mergers Acquisitions Bargaining Negotiation Compensation Governance 1. Introduction Corporate directors and executives have considerable freedom over the decision to sell the firm. Once they decide to sell, the target s board has a duty to secure a deal that is in the best interests of the owners. The board is expected to monitor the sale process and often, but not always, delegates the job of negotiating the sale to the CEO $ This paper is based on my dissertation at the University of Arizona. I am particularly indebted to an anonymous referee for suggestions that substantially improved the paper. I appreciate the helpful comments of Daniel Bens, James Brickley, Dan Dhaliwal (committee chair), Merle Erickson, Sandy Klasa, Michael Raith, Clifford Smith, Mark Trombley, Jerold Warner, Joanna Wu, Jerold Zimmerman, and workshop participants at Arizona, Arizona State, Chicago, Dartmouth, Duke, Michigan, MIT, Northwestern, Notre Dame, Ohio State, Penn State, Rochester, Stanford, Texas at Dallas, Utah, and Washington University. n Tel.: þ address: shane.heitzman@simon.rochester.edu who usually has a substantial amount of personal wealth tied to the deal s outcome. Because negotiations usually begin months before the deal is announced, the target s board has an option to increase the equity held by the CEO. There are at least two reasons for doing this. First, the acquisition premium accrues to equity granted before a deal is announced and so provides compensation to the target CEO. Second, shares and options granted during negotiations increase the target CEO s incentives to push for a higher purchase price. Such grants are controversial and have been the target of shareholder litigation, securities regulation, and press commentary. In this paper, I analyze the motives for and consequences of granting equity to the target CEO during deal negotiations. In this paper, three key elements in the analysis are the CEO s role in the negotiations, the bargaining power of the target firm, and monitoring by the board of directors. First, a basic conjecture in the related literature is that the target CEO is in a position to negotiate a lower offer price X/$ - see front matter & 2011 Elsevier B.V. All rights reserved. doi: /j.jfineco

2 252 S. Heitzman / Journal of Financial Economics 102 (2011) in exchange for private benefits (e.g., Wulf, 2004; Hartzell, Ofek, and Yermack, 2004; Möeller, 2005; Bargeron, Schlingemann, Stulz, and Zutter, 2010). However, dealrelated disclosures offer a more complete picture: the CEO is mentioned explicitly as having authority to negotiate on behalf of target shareholders in just more than half of the deals in my sample. Second, bargaining power is relevant to the question if the gain from negotiating effort (or the potential loss from self-dealing) is greater when the target is in a stronger negotiating position. Finally, monitoring is relevant because directors have a fiduciary duty to maximize share value, judicial precedent emphasizes the importance of independent directors in a control transaction, and prior research finds that a board with a greater proportion of outside directors generates more value for target shareholders. I analyze a sample of 471 completed acquisitions of publicly traded U.S. firms announced between 1996 and From background disclosures, I identify the target CEO s involvement in negotiations, the existence of competing bidders before a deal is publicly announced, and when and by which side the deal is initiated. I use Thomson s insider filing database to identify grants of stock options and restricted stock to the target CEO. Negotiation grants, the focus of this study, occur between the start of private negotiations and the announcement of a deal. On average, 33% of CEOs receive stock or options during these negotiations. The compensation hypothesis for negotiation grants is rooted in the literatures on golden parachutes (e.g., Jensen, 1988) and managerial influence over compensation (e.g., Bebchuk, Fried, and Walker, 2002). If equity is used as compensation, the likelihood of a negotiation grant should increase in the expected premium. This prediction assumes that the board is less likely to use equity-based compensation when the premium and the per-share value of the grant is small. Across the full sample, the decision to grant equity has no correlation with the expected premium. However, among firms with a smaller fraction of outsiders on the board, the CEO is more likely to get a negotiation grant when the expected premium is high. This result is more consistent with managers having influence over compensation. But there is little evidence that negotiation grants provide parachute-like compensation: a CEO who is further from retirement, has no golden parachute, and earns past excess cash compensation is no more likely to receive a negotiation grant, and does not receive grants of higher intrinsic value measured at the close of the deal. The bargaining incentives hypothesis predicts that CEOs who negotiate have a higher marginal product of effort in the deal and hence are more likely to get a negotiation grant (and conditional on getting a grant, receive bigger incentives). I find that a CEO who negotiates is significantly more likely to receive an equity grant in the negotiation period relative to prior periods, supporting the prediction that boards use equity grants to strengthen the CEO s bargaining incentives. These tests control for the firm s normal granting practices as well as factors related to the delegation decision. In economic terms, a CEO who negotiates has a 0.14 greater probability of receiving equity during negotiations than a CEO who does not. Given that only one-third of target CEOs receive any kind of grant during negotiations, this effect is economically significant. Conditional on receiving a negotiation grant, however, a CEO who negotiates does not appear to receive significantly stronger incentives. The impact of CEO effort on the negotiated price, and hence the demand for equity incentives, can also be conditional on the target s bargaining power. Comparing deals in which the buyer makes the first move to those in which the target puts itself up for sale is an intuitive (though not perfect) way to identify the role of bargaining power as it provides a signal about which party has a stronger preference to do a deal. On average, a CEO in a buyer-initiated deal has a 0.08 higher probability of receiving a negotiation grant. Among these buyerinitiated deals, I find that a CEO who negotiates for the target has a 0.19 higher probability of receiving a negotiation grant, while an increase in the proportion of outside directors by ten percentage points raises the probability of a negotiation grant by In contrast, when the target initiates the sale, the CEO s role in negotiations and the prevalence of outside directors have no measurable impact on the decision to grant equity during negotiations. I then analyze the level of incentives conditional on receiving a negotiation grant. The evidence indicates that the target of a buyer-initiated deal provides stronger bargaining incentives when incentives are measured as the sensitivity of CEO pay to the percentage returns (i.e., premiums). A target with a smaller fraction of outsiders on the board provides stronger incentives for the CEO to avoid wealth transfers when incentives are measured as the sensitivity of CEO pay to dollar returns. Turning to shareholder wealth implications, I find no meaningful relation between negotiation grants and observed premiums. This stands in contrast to recent claims that target CEOs actually destroy substantial shareholder wealth when they receive an unscheduled negotiation grant (Maremont, 2009; Fich, Cai, and Tran, 2011). 1 The results here imply that boards attempt to increase bargaining incentives by making grants to target CEOs who negotiate the sale, and they are more likely to do so when the target s negotiator has more bargaining power. There is some evidence that grants are used as compensation since the likelihood of a negotiation grant is increasing in the expected premium. Because this occurs only among targets with a smaller fraction of 1 Fich, Cai, and Tran (2011) conclude that unscheduled option awards during private negotiations cause CEOs to negotiate significantly lower premiums, by 4.4 percentage points, destroying about $216 million in deal value or $62 in shareholder wealth for every $1 of option value granted to the CEO. But such a conclusion seems implausible. First, it seems hard to believe that target CEOs and their boards can leave that much on the table without legal repercussions. Second, and more strikingly, the CEO also holds a substantial equity stake in the firm. Given that the average CEO holds stock and options giving them a 4% interest in dollar gains, Fich et al. essentially argue that CEOs take actions that reduce their portfolio value by $2.5 for every $1 they get through an unscheduled grant. Given the average grant gains by $3.46 million in their study, these CEOs actually lose close to $5.2 million net when the loss on the portfolio is factored in.

3 S. Heitzman / Journal of Financial Economics 102 (2011) outside directors, and because the amount of grantrelated compensation is unrelated to proxies for the CEO s expected loss, this is some evidence that grants do provide unwarranted compensation. I find no evidence that the cost to shareholders exceeds the value of the equity award for CEOs who do not negotiate and more importantly, no evidence that negotiation grants cause CEOs to accept smaller premiums. 2. Background and theoretical development 2.1. Prior research on managerial equity incentives in target firms Agency conflicts between managers and shareholders can become costly when the CEO controls the outcome of takeover negotiations. In theory, a CEO with a bigger equity stakeismorelikelytoactintheshareholders interest,but such control can also facilitate entrenchment. Prior studies find that firms with lower managerial ownership are more likely to become the target of a control attempt (Mikkelson and Partch, 1989) and are more likely to be acquired(song and Walkling, 1993; Hadlock, Houston, and Ryngaert, 1999). However, there is also evidence that managers that own more obtain larger premiums (Stulz, Walkling, and Song, 1990; Song and Walkling, 1993; Cotter and Zenner, 1994), while Möeller (2005) finds an opposite result. Recent work extends the analysis to option portfolios and provides a more comprehensive treatment of the link between managerial equity incentives and deal outcomes (Kahan and Rock, 2002; Coates and Kraakman, 2004; Cai and Vijh, 2007) Motives for granting equity to CEOs during negotiations Negotiations begin an average of six months before the deal s first public announcement in my sample; surprise public bids are rare. 3 This is likely to be a downwardbiased estimate of how early the target prepares for the 2 To explain the increase in both takeover activity and executive option awards during the 1990s, Kahan and Rock (2002) conjecture that boards have made managers more receptive to selling the firm by increasing option incentives and including change-in-control provisions that lift restrictions on unvested stock and options if the firm is acquired. However, Coates and Kraakman (2004) find only limited evidence that CEO option portfolio holdings are positively associated with the decision to sell. Coates and Kraakman do find that the association between the option portfolio and the probability of being acquired is greater for lowrisk than for high-risk firms. This is consistent with options having a greater value to the CEO from selling the firm when firm risk, and the value of the options from remaining independent, is low. Cai and Vijh (2007) argue that CEOs want to sell the firm to realize the equity discount that arises because of managers under-diversification and liquidity restrictions (Meulbroek, 2001; Hall and Murphy, 2002). 3 With respect to the duties to disclose merger negotiations to shareholders, Gilson and Black (2000, p. 1192) note, an issuer does not have an affirmative duty to disclose preliminary merger negotiations, simply because these negotiations are materialythe issuer may delay the disclosure of any information about material merger negotiations for a valid business purpose, such as the risk that premature disclosure would jeopardize the deal, and respond to queries with silence or a statement of no comment. In practice, a deal is not usually announced to the public until the terms of the agreement are approved by the board. sale of the firm as top insiders often know the firm is in play long before a specific acquirer is identified, especially when the target puts itself up for sale. The target s board thus has an opportunity to modify the CEO s stock and option awards during or even before private negotiations begin. There are at least two reasons for doing this. First, negotiation grants provide compensation. Second, negotiation grants provide more powerful bargaining incentives. I elaborate on these explanations next Compensation hypothesis Although private deal negotiations may take place over several months, the information is not reflected in the target s stock price until a few weeks before the deal is announced, and then only partially (Jarrell and Poulsen, 1989; Schwert, 1996). Given a positive expected premium, granting stock and options early in negotiations effectively provides a builtin gain to the CEO. Vesting restrictions are normally lifted upon a change in control and options are often cashed out. Even if the target CEO receives acquirer stock or options in the deal rather than cash, they can usually monetize these positions after the deal closes and diversify. 5 This makes equity awards a valuable source of compensation. One reason for compensating the CEO has to do with their incentive to resist a transaction that makes them worse off. For example, the CEO invests significant human capital in the firm and realizes it in the form of salary and bonus, equity appreciation, and other pecuniary and nonpecuniary benefits. The value of these benefits depends on the CEO s horizon with the firm, but only about half of target CEOs stay on the job following an acquisition (Agrawal and Walkling, 1994; Harford, 2003; Hartzell, Ofek, and Yermack, 2004). Because target shareholders usually gain from selling, Jensen (1988) and Knoeber (1986) argue that shareholders have an incentive to hold the CEO indifferent to the transaction. Change-in-control provisions address this by, for example, requiring payment of a fixed multiple of prior salary, bonus, or other incentive compensation if the executive is dismissed or resigns within a certain period of time after a change in control (i.e., a golden parachute). 6 While these provisions 4 The maintained assumption throughout the paper is that the board and CEO know about the firm s sale far enough in advance to affect compensation decisions. For deals with short negotiation periods, the ability of firms to backdate grants before 2003 opens the door for boards wanting to grant equity late in the process. In the case of Fore Systems, described in Appendix A, directors backdated option grants to target executives as the firm was in discussions with buyers. 5 This is true unless the executive is constrained by vesting provisions or contractual arrangements preventing the sale of acquirer shares. To the best of my knowledge, large sample evidence on such agreements, and the speed at which target CEOs divest acquirer equity, is not yet available. Such an agreement was entered into by Jeffery Kilts, CEO of Gillette, which was acquired by Procter and Gamble (P&G). According to Securities and Exchange Commission (SEC) filings, Kilts stock and option holdings in Gillette were converted into stock and options of P&G. As a condition of the sale, Kilts agreed not to sell any P&G stock or exercise any option for 18 months following the completion of the acquisition. 6 A change-in-control is defined by the terms of the agreement and can be triggered by the acquisition of a certain percentage of voting stock by an acquirer or the replacement of a minimum fraction of the board within a single year, among other things. While these provisions are common (75% of CEOs in my sample are covered by one), there is

4 254 S. Heitzman / Journal of Financial Economics 102 (2011) are often written long before negotiations start, it is not uncommon for the board to wait until actual deal negotiations to modify the payout (Hartzell, Ofek, and Yermack, 2004). Important for this study, the acquisition premium accruing to the equity also compensates the target CEO for selling the firm and may be preferred when expected premiums are high; when expected premiums are low, a cash payment may be preferred. 7 Moreover, the grant s intrinsic value is conditional on the takeover actually occurring, as equity values for target firms tend to fall back to pre-offer prices if an offer is cancelled (Bradley, Desai, and Kim, 1983) and unvested stock and options usually remain unvested. A second explanation by Bebchuk, Fried, and Walker (2002) is that top managers have influence over their compensation and can dictate the timing and amount of equity granted by the board to maximize their personal wealth at the expense of shareholders. Prior work finds that managers influence the value of their option awards by timing stock option awards and cash exercises on days with low stock prices (Yermack, 1997; Aboody and Kasznik, 2000; Lie, 2005; Heron and Lie, 2007; Cicero, 2009; Dhaliwal, Erickson, and Heitzman, 2009). In a setting more similar to this study, Lowry and Murphy (2007) find no evidence that CEOs of firms doing initial public offerings are more likely to get options whose strike price is tied to the offer price when the expected underpricing and gain to the option holder is greater. The compensation hypothesis leads to a number of empirical predictions. First, the decision to grant equity should be positively correlated with the expected premium. Second, if equity is used to compensate the CEO for agreeing to the sale, the value of the negotiation grant should be increasing in the CEO s expected loss. Third, if negotiation grants reflect self-dealing by influential managers, they should be related to monitoring provided by outside directors. (footnote continued) little consistent evidence that golden parachutes have a material impact on acquisition decisions or shareholder wealth. Investors react positively when the firm adopts a golden parachute, which suggests that these contracts reduce agency costs in many firms (Lambert and Larcker, 1985). However, the adoption of a golden parachute contract does not necessarily signal an impending takeover (Machlin, Choe, and Miles, 1993). 7 There is anecdotal evidence on the existence of such tradeoffs in a general setting. For example, Case Corporation permits participants to exchange part of or their entire annual cash bonus for shares of common stock equal in value to 133 1/3% of the foregone cash award. (Case Corporation Proxy Statement, May 12, 1999.) In a separate example, Rubbermaid s compensation committee approved a program that allows an executive to elect to receive stock options in lieu of cash compensation, performance shares, and the accrual to the executive s supplemental retirement plan. The plan was approved in 1997 prior to the acquisition of Rubbermaid by Newell Co. announced in In 1997, all Rubbermaid executives, including the CEO, exercised this option. In another instance, approximately eight months before the acquisition of Santa Fe Snyder Corp by Devon Energy was announced in 2000, the CEO of the target agreed to receive restricted stock in exchange for terminating his employment agreement that provided severance pay under certain conditions if the firm was acquired. The value of the restricted stock received based on the acquisition price was over $4.5 million, nearly 4.5 times the CEO s highest salary and bonus over the prior 3 years Bargaining incentives hypothesis It is commonly held that the target CEO controls the sale of the firm and has the power to negotiate a lower price for shareholders in exchange for private benefits (e.g., Wulf, 2004; Hartzell, Ofek, and Yermack, 2004; Möeller, 2005; Bargeron, Schlingemann, Stulz, and Zutter, 2010). While the CEO s cooperation in the sale is considered important, the board is not required to have the CEO negotiate key terms like deal price and structure. Many boards delegate negotiations to independent directors or financial advisors. Others give the CEO substantial freedom to negotiate. The board s decision to delegate negotiations to the CEO should depend on many factors such as the CEO s explicit and implicit incentives, firm-specific knowledge, bargaining skills, and conflicts of interest. A critical point here is that reliance on the simple assumption that all CEOs negotiate can result in misleading inferences. With that in mind, the bargaining incentives hypothesis relies on two intuitive assumptions rooted in agency theory. First, a manager with a higher marginal product of effort in deal negotiations ought to be given greater incentives to exert that effort. Second, equity provides direct incentives for the negotiator to secure a better deal for shareholders by rewarding bargaining effort and penalizing self-dealing. 8 This leads to a number of predictions. First, a CEO who negotiates is more likely to receive a negotiation grant than a CEO who does not. Second, a CEO with more bargaining power is more likely to receive a negotiation grant. Third, among CEOs that receive negotiation grants, those who negotiate or have more bargaining power should be given stronger incentives. If CEOs who get more equity have stronger incentives to push for a higher price, then negotiation grants should also have a positive association with realized acquisition premiums. 9 However, because negotiation grants are endogenously determined along with other important dimensions of the deal, identifying a causal relation between equity grants and shareholder wealth is somewhat difficult. At the extreme, if all boards make the optimal grant decision, there should be no correlation between observed negotiation grants and deal premiums 8 Two points are worth making here. First, many theoretical models of delegated bargaining do not incorporate negotiator effort; bids are cheap-talk. But anecdotal evidence and intuition suggests that effort matters. Negotiations take many months and distract the CEO from other duties at the firm to deal with complex issues relating to valuation, antitrust issues, tax and accounting complexities, and leadership, among other things. Second, models of incentive contracts invariably provide a link between performance and pay, since effort is unobservable. In practice, we do not usually observe compensation contracts tied solely to stock price, as stock price movements outside of the control of the manager can impose undue risk. However, during deal negotiations, it seems reasonable to believe that stock price provides a relatively cleaner signal of CEO effort precisely because the CEO is negotiating the price. This provides a basis for the prediction that boards are more likely to increase the equity holdings of CEOs actually negotiating the acquisition price. 9 An alternative argument suggests that larger grants lead to lower premiums when CEOs are risk-averse. But even if equity incentives have perverse effects on managers concerned about diversification and liquidity as claimed by Cai and Vijh (2007), active monitoring by an independent board of directors or special committee can presumably address this.

5 S. Heitzman / Journal of Financial Economics 102 (2011) (Demsetz and Lehn, 1985; Core and Guay, 1999; Himmelberg, Hubbard, and Palia, 1999). Not surprisingly, prior evidence on the link between executive equity incentives and future firm performance has produced mixed results (Core, Guay, and Larcker, 2003) The role of monitoring It is well known that the board of directors plays a fundamental part in the acquisition process. During negotiations, directors meet frequently and consult with legal and financial advisors before making decisions. The board, or a special committee formed by the board, can direct the negotiator to push for a higher price, different terms, or competing bids and usually approves the acquisition agreement before it is announced to shareholders. Allegations that management or the board unfairly restricted bid competition (for example, to deal with a favored acquirer) or received unwarranted compensation are not uncommon. But boards with a majority of independent directors are presumed to make decisions in good faith, so litigation costs can in theory be minimized by having independent and informed directors monitor negotiations and limit opportunistic behavior by the CEO (American Law Institute, 1994; Gilson and Black, 2000). 10 Consistent with the conjecture that board structure affects deal negotiations, targets with a greater fraction of outsiders on the board appear to receive higher takeover premiums (Byrd and Hickman, 1992; Cotter, Shivdasani, and Zenner, 1997; Möeller, 2005). The point here is that a board with a greater proportion of outside directors arguably provides better monitoring of negotiations and is less likely to approve unwarranted compensation to the manager. Consistent with this, board independence appears relevant for understanding the motives behind control-related arrangements. For example, Brickley, Coles, and Terry (1994) show that shareholders reaction to the adoption of a poison pill is positive when outside directors are a majority, negative when they are not. Thus, I use the fraction of outside directors on the board as a measure of board independence in the empirical tests, both as a determinant of negotiation grants and as a way to understand the link between negotiation grants and shareholder wealth. 3. Deal negotiations and negotiation grants 3.1. Data and sample I obtain a sample of completed acquisitions of publicly traded U.S. companies announced between 1996 and 2006 from the Securities Data Corporation (SDC) Mergers and Acquisitions database. In order for a transaction to be 10 See the landmark case of Unocal Corp. v. Mesa Petroleum Co. 493 A.2d 946 (Del. 1985), in which the court held that when defensive actions are taken by a board with a majority of outside independent directors, proof of good faith is materially enhanced. This was reinforced in Polk v. Good 507 A.2d 531, 537 (Del. 1986) in which the court held that the presence of a majority of independent directors who rely on advice by legal and financial advisors, constitute[s] a prima facie showing of good faith and reasonable investigation. included in the sample, the acquirer must not own a majority interest prior to the acquisition, the target is covered by Center for Research in Security Prices (CRSP) and Compustat and is not bankrupt, the target CEO is listed on the insider filings database and has at least 3 years of executive compensation and holdings data available from ExecuComp, and the deal background disclosure is available from mergerrelated documents filed with the Securities and Exchange Commission (SEC) Deal characteristics In the first column of Table 1, Panel A, I summarize the deal characteristics for the 471 deals in my sample. Based on the entire sample, the realized premium, calculated as the total return starting 20 days prior to the announcement date and ending 1 year after announcement or delisting (whichever comes first), averages 35.4%. I focus on total returns as the buyer and seller are negotiating the acquisition price which determines the level of compensation and the total gains to shareholders. The results are similar if I use market-adjusted returns following Schwert (1996).The initial offer premium, calculated using the initial offer price reported by SDC, averages 36.1%. Target shareholders are paid solely in cash in 44% of deals, the remainder being all stock (28%) or a combination of cash and stock. According to SDC data, about 6% of targets have a competing offer after the announcement date. Using the background data, I find that 42% of targets have competing offers before the announcement, suggesting that acquisition contests are more competitive than what the SDC data imply and is consistent with the evidence in Boone and Mulherin (2007). In 65% of the deals, the buyer initiates negotiations while in the remaining deals, the target puts itself up for sale. Private negotiations last an average of 196 days measured from initiation to announcement, supporting the assumption that the target s executives and directors possess material information about an upcoming takeover well in advance of any public disclosure Firm and CEO characteristics For fiscal year data (e.g., total assets, earnings, etc.), I use the last fiscal year with available data ending no later than 3 months following the first announcement of the deal. The target firms in this sample are large, with an average market capitalization of around $3.3 billion. Earnings before interest and taxes (EBIT) appear stable, averaging 8.8% of total assets in the negotiation year and 8.7% the year before. However, abnormal stock returns are 1.3% in the negotiation year and 6.2% the year before, suggesting that target firms have weaker stock performance leading up to the deal. About two-thirds of directors are considered independent. The CEO has been in the position an average of 8.6 years, chairs the board about two-thirds of the time, and is covered by a golden parachute in almost three out of four deals. 11 Based on the 11 Payouts for golden parachutes are based on provisions such as three times base bonus and salary. Many of these agreements also

6 256 S. Heitzman / Journal of Financial Economics 102 (2011) Table 1 Descriptive statistics and the CEO s role in negotiations for a sample of 471 completed acquisitions announced between 1996 and Panel A provides summary deal statistics. Panel B provides summary statistics for the target firm and CEO for the negotiation year. Panel C provides evidence on the determinants of the CEO s role in negotiations. The realized premium (%) is the total stock return for the target firm starting 20 days prior to the first public announcement of the deal through deal completion or 1 year after announcement, whichever comes first. The initial offer premium is the total return based on the initial offer price disclosed by SDC, and the stock price 20 days prior to the announcement. The total dollar premium is at deal completion and is reported in millions. Indicator variables for tender offers, hostile bids, all stock, and all cash deals are based on SDC data. Competing public (private) offer is an indicator variable equal to one if there is more than one bidder after (before) the deal has been announced and is based on SDC classification (background disclosures). The CEO s role in negotiations and the initiation date and initiating party are obtained from background disclosures. MVE is the market value of equity 20 days before the announcement. Market-to-book is the market value of assets at the end of year t divided by the book value of assets. EBIT is earnings before interest and taxes divided by total assets. Return is the annual value-weighted marketadjusted return ending in the sixth month of the year. s(return) is the standard deviation of daily returns for 12 months ending in the sixth month of the current year. % Outside directors is the fraction of the board considered independent using RiskMetrics when available and hand-collected for the remaining 104 firms. Indicator variables for the existence of a poison pill amendment and a classified board are drawn from the RiskMetrics data set. Golden parachute equals one if the firm has change-in-control agreements with top executives, zero otherwise. Tenure is the number of years the CEO has been in office. Chairman equals one if the CEO is also the Chairman of the Board, zero otherwise. The CEO portfolio measures are expressed as pay-performance sensitivities (PPS) following Core and Guay (2002) and Jensen and Murphy (1990). p-value based on a two-tailed t-test of differences in means. Panel A: Deal characteristics All firms (N¼471) CEO negotiates (N¼257) CEO does not negotiate (N¼214) p-value of difference Realized premium (%) Initial offer premium n (%) Total dollar premium ($mil) , o0.01 Acquirer is public o0.01 Tender offer Hostile offer All stock o0.01 All cash o0.01 Competing public offer Competing private offer o0.01 Buyer initiates deal o0.01 Days between initiation and announcement Days between announcement and close Panel B: Target firm and CEO characteristics All firms CEO negotiates CEO does not negotiate p-value of difference Sales ($bil.) o0.01 MVE ($bil.) o0.01 Market-to-book EBIT t (%) EBIT t 1 (%) Return t (%) Return t 1 (%) s(return) % Outside directors Poison pill nn o0.01 Classified board nn CEO golden parachute nn CEO tenure CEO is chairman CEO portfolio PPS% ($thou. per 1% D in value) Shares only Options only o0.01 CEO portfolio PPS$ ($1 per $1k D in value) o0.01 Shares only o0.01 Options only Panel C: Determinants of the decision to delegate negotiations to the CEO Prob(CEO negotiates¼1) Coeff. (t-stat) Marginal effect Coeff. (t-stat) Marginal effect ln(mve) 0.24 (4.14) (3.95) 0.07 Market-to-book 0.03 (0.44) (0.73) 0.02 EBIT t 0.75 (0.95) (0.31) 0.10 Return t 0.19 (1.29) (1.68) 0.10 s(return) 0.42 (0.31) (0.22) 0.10 % Outside directors (2.01) (1.98) 0.10

7 S. Heitzman / Journal of Financial Economics 102 (2011) Table 1 (continued ) Panel C: Determinants of the decision to delegate negotiations to the CEO Prob(CEO negotiates¼1) Coeff. (t-stat) Marginal effect Coeff. (t-stat) Marginal effect Stock holdings 0.03 ( 1.96) ( 0.96) 0.01 Option holdings 0.04 (0.96) (0.65) 0.01 CEO tenure 0.00 (0.38) ( 0.04) 0.00 CEO is chairman 0.23 (1.67) (1.26) 0.07 Buyer initiates deal 0.45 (3.36) (3.19) 0.15 Expected premium 0.01 (0.06) (0.11) 0.01 Poison pill 0.35 (2.35) 0.12 Classified board 0.10 (0.69) 0.03 Golden parachute 0.16 ( 0.92) 0.05 Year effects Yes Yes Pseudo R N n 465 observations, nn 408 observations. CEO s portfolio incentives calculated as the dollar change in CEO wealth for a $1,000 change in equity value (PPS$), the average CEO holds about 2.5% of the firm s outstanding shares ($25.22/$1,000), while delta-adjusted option holdings are about 1.4% relative to outstanding shares CEOs who negotiate vs. CEOs who do not For each deal, I read the disclosure on the background of the deal contained in filings with the SEC to identify the CEO s involvement in negotiations with the buyer. I determine CEO involvement based on whether the board authorizes the CEO to negotiate and the extent to which the CEO participates in actual negotiations over price and other important terms. The classification process, described more fully in Appendix B, reveals that the CEO is mentioned explicitly as having the authority to negotiate with buyers in 54.6% of the deals in my sample. 12 In the second and third columns of Table 1, I provide summary statistics for deals that the CEO negotiates and those that the CEO does not. At the univariate level, the comparison reveals several significant differences. Since the decision to have the CEO negotiate is likely determined by factors that also determine premiums and compensation decisions, failing to control for those factors could lead to (footnote continued) provide for accelerated vesting of pension benefits and restricted stock and options, extended medical coverage, or continued use of a company vehicle, among other things. Further, these payments are often grossedup to the extent the CEO incurs a tax on excess parachute benefits. See additional evidence in Agrawal and Knoeber (1998) and Hartzell, Ofek, and Yermack (2004). 12 I recognize that this discrete measure cannot and does not capture the many nuances of CEO influence on negotiations and suffers from variation in disclosure choices. For example, a heavily involved CEO may be treated as having little involvement if the background disclosure does not mention the CEO directly. And a CEO who does appear to negotiate directly may nevertheless have little freedom in bargaining if the board constrains their discretion. These shortcomings in the data add noise to the measure and should work against finding any meaningful results using the proxy for negotiating CEOs. misleading inferences. To provide evidence on which factors matter, I estimate a probit model of the decision to delegate negotiations to the CEO based on a number of factors that capture the CEO s attributes, the firm s economic characteristics, and governance mechanisms. The results are in Panel C. From the full sample reported in the first column, the CEO is more likely to negotiate when the firm is larger, has higher past stock returns, and a higher proportion of outsiders on the board. For example, when more than 60% of the board is comprised of outside directors, the CEO has a 0.09 higher probability of negotiating for the target. When the target initiates the deal, the CEO has a 0.15 lower probability of negotiating. A CEO that also chairs the board is more likely to negotiate, while a CEO that owns a larger percentage of the firm is less likely to negotiate. For a subsample of 408 firms with additional governance data, poison pills are associated with a 0.12 higher probability of delegating negotiations to the CEO. The existence of a golden parachute has a negative, but insignificant effect. To mitigate concerns about the endogeneity of CEO involvement in negotiations, I control for these determinants to the extent possible Measuring equity grants and incentives To construct the most accurate and comprehensive measure of equity awards, I focus on both stock and option grants as reported on insider filings with the SEC. I obtain CEO stock and option holdings data prior to the acquisition from ExecuComp. I merge the insider filing data with ExecuComp data by firm identifier and CEO name, and hand check all targets to ensure a correct match. Because option grant data are not available prior to 1996 from the insider filings database, I supplement the option grant history with grants inferred from Execu- Comp prior to the first disclosed grant in the insider filing database. 13 For the primary measure of equity grants, 13 In the limited cases in which I use option grants reported by ExecuComp, I infer the grant date based on the expiration date assuming

8 258 S. Heitzman / Journal of Financial Economics 102 (2011) Table 2 The frequency and size of equity grants to target CEOs during negotiations. Panel A provides summary statistics for stock and option grants to 471 target CEOs between the initiation date and the announcement date of a sample of completed deals announced between 1996 and Panel B provides summary statistics for equity grants over the 12-month period before announcement. The CEO s role in negotiations and the initiation date are obtained from background disclosures as described in Appendix B. PPS% is the dollar change in grant value given a 1% change in stock price (in thousands). PPS$ is the dollar change in grant value given a $1,000 change in stock price. The intrinsic value of the award is based on the price at the completion of the deal, or 1 year after announcement, whichever is first (in millions). p-value based on a two-tailed t-test of differences in means, and a median test for differences in medians. All CEOs CEO negotiates CEO does not negotiate p-value of difference (N¼471) (N¼257) (N¼214) Mean [Median] Mean [Median] Mean [Median] Mean [Median] Panel A: Equity grants between initiation and announcement All equity awards % Receiving equity PPS% 35.9 [15.1] 43.6 [18.0] 24.8 [14.1] 0.05 [0.15] PPS$ 3.1 [1.4] 2.7 [1.2] 3.9 [1.5] 0.27 [0.19] Intrinsic value 2.3 [0.7] 2.6 [0.8] 1.9 [0.7] 0.33 [0.72] Option awards % Receiving options PPS% 34.5 [16.5] 42.8 [20.4] 22.4 [12.3] 0.05 [0.04] PPS$ 3.1 [1.5] 2.8 [1.4] 3.6 [1.5] 0.46 [0.48] Intrinsic value 1.5 [0.6] 1.7 [0.6] 1.3 [0.5] 0.37 [0.99] Stock awards % Receiving stock PPS% 20.5 [10.0] 20.6 [10.5] 20.2 [9.5] 0.96 [0.98] PPS$ 1.5 [0.4] 0.9 [0.3] 2.4 [1.1] 0.04 [0.05] Intrinsic value 2.6 [1.2] 2.7 [1.3] 2.5 [1.2] 0.92 [0.95] Panel B: Equity grants during 12 months prior to announcement All equity awards % Receiving equity o0.01 PPS% 46.6 [22.0] 57.2 [24.7] 29.4 [15.4] o0.01 [o0.01] PPS$ 3.1 [1.6] 2.7 [1.4] 3.6 [1.8] 0.11 [0.11] Intrinsic value 3.0 [1.2] 3.5 [1.3] 2.1 [0.8] 0.05 [0.02] Option awards % Receiving options o0.01 PPS% 42.9 [21.7] 52.5 [23.5] 27.2 [13.9] 0.01 [0.01] PPS$ 2.9 [1.5] 2.6 [1.4] 3.5 [1.8] 0.13 [0.12] Intrinsic value 1.9 [0.7] 2.1 [0.8] 1.6 [0.6] 0.29 [0.19] Stock awards % Receiving stock o0.01 PPS% 23.2 [7.9] 26.1 [8.5] 16.8 [7.1] 0.33 [0.52] PPS$ 1.2 [0.4] 1.0 [0.3] 1.8 [1.0] 0.06 [0.07] Intrinsic value 3.0 [1.1] 3.4 [1.2] 2.1 [1.0] 0.27 [0.50] I focus on negotiation grants which include shares and options granted during the negotiation period, i.e., between the initiation and announcement dates. I use four measures of equity grants in this paper: an indicator variable if the CEO receives any equity during negotiations (i.e., negotiation grants), two measures of the incentives provided by the negotiation grants (i.e., the pay-performance sensitivities), and the intrinsic value of the negotiation grant measured at the close of the deal (or 1 year after announcement, whichever is earlier). (footnote continued) the grant term is in whole years (over 92% of grants to CEOs have a term based in whole years rather than partial years). As a validity check, I find that the correlation between equity grant values estimated using SEC data and ExecuComp data consistently exceed 0.9 when data are available from both sources. I hand check random cases in which the ExecuComp and insider filing-based measures disagree by a material amount and find that the differences are driven by errors in both the SEC filings and the ExecuComp data. Grants of stock and options from insider filings are identified primarily by transaction codes A and J. Reported in Panel A of Table 2, 33% of target CEOs receive a negotiation grant. Option grants are more frequent, but stock awards are not trivial. Of the CEOs who receive equity, about 27% get only stock, 58% get only options, and 15% get both. Of CEOs who negotiate, 36% receive a negotiation grant compared to 30% of CEOs who do not. While this difference is insignificant, CEOs who negotiate are significantly more likely to receive a stock grant. One concern in interpreting these frequencies is that the length of the negotiation period, hence the window over which grants are identified, varies across deals. This is important because as the window over which I look for a grant increases, so does the likelihood of observing an equity grant, independent of the deal. A second concern is that if the initiation date or the grant date is not precisely identified, there will be additional measurement error in negotiation grants. This issue could arise if there is strategic disclosure of initiation and grant dates, or if targets plan for the sale long before a specific acquirer is indentified. To address these

9 S. Heitzman / Journal of Financial Economics 102 (2011) concerns, I provide summary statistics in Panel B for grants measured over a constant 12-month period prior to the announcement. Over the full negotiation year, 71% of CEOs who negotiate receive equity compared to 53% of those who do not, and the difference is significant. The difference arises in both stock and option grants. Following Jensen and Murphy (1990) and Yermack (1995), one measure of the pay-performance sensitivity is the dollar change in managerial wealth given a $1,000 change in firm value, or PPS$ ¼ ðd #Options grantedþþ#shares granted $1,000 Shares outstanding, ð1þ in which d, the first derivative of option value with respect to a change in stock price, is estimated using the methodology in Core and Guay (1999). PPS$ reflects the CEO s fractional interest in firm value and, in the spirit of Baker and Hall (2004), is more appropriate if the board is concerned about the CEO negotiating side agreements that benefit the manager at the expense of shareholders. 14 In other words, if a CEO negotiates an extra dollar for himself in exchange for a dollar less in deal price, he gains at most $1(1 a), in which a is the fractional interest in equity value. Increasing PPS$ increases a and lowers the incentive to give up premium for private benefits. Conditional on receiving a negotiation grant, the mean PPS$ for the full sample of target CEOs in the negotiation period is $3.1 per $1,000 change in firm value, equivalent to a 0.31% fractional interest in the premium for negotiation grants alone. The average incentives from options exceed those from stock: $3.1 vs. $1.5 per $1,000 change in firm value. CEOs who do not negotiate appear to receive larger incentives, although this is likely due to differences in firm size (CEOs of larger firms tend to have lower PPS$). The difference is statistically significant for restricted stock grants. Following Core and Guay (1999), a second measure of the pay-performance sensitivity is the dollar change in grant value given a 1% change in stock price: PPS% ¼ ðd #Options grantedþþ#shares granted Stock price ð2þ 100 with stock price measured at the grant date. PPS% captures the manager s dollars at stake and is an intuitive measure of the incentives to exert effort in negotiations as it reflects the dollar increase in CEO wealth for each additional 1% premium paid by the buyer (Baker and Hall, 2004). From Table 2, the mean pay-performance sensitivity of negotiation grants using this methodology implies that the average negotiation grant increases CEO 14 Baker and Hall (2004) note that the appropriate measure of equity incentives depends on the question asked. They suggest that for CEO actions that affect percentage returns, such as reorganizations, measuring incentives as the dollar change in value of the executive s equity-based incentives given a 1% increase in stock price is most appropriate. However, if CEO actions are related to dollar returns, such as the consumption of perquisites, the appropriate measure is based on dollar changes in stock price. Because the questions in this paper involve potentially both of these actions, I examine the effects using both constructs. wealth by $35.9 thousand for every 1% increase in deal price. Of the CEOs who receive equity, CEOs receive grants with significantly larger option incentives when they negotiate relative to when they do not ($43 thousand vs. $22 thousand per 1% change in stock price), while stock incentives are similar at roughly $20 thousand. The intrinsic value of the negotiation grant is measured at the completion of the deal. The statistics reported in Table 2 are conditional on receiving a grant. From Panel A, the average negotiation grant has an intrinsic value of $2.3 million. Option grants average $1.5 million while stock grants average $2.6 million. Conditional on receiving a negotiation grant, the intrinsic value of the stock and option awards does not differ significantly between CEOs who negotiate and those who do not. Prior research suggests that the CEO s portfolio of stock and options could influence the acquisition outcome as well as current year grants (e.g., Core and Guay, 1999; Möeller, 2005; Cai and Vijh, 2007). Moreover, the portfolio holding of stock and options provides information about the firm s past granting practices. I measure portfolio incentives using the pay-performance sensitivities of stock and options in both PPS$ and PPS% terms. I use the oneyear approximation method described in Core and Guay (2002) to calculate the sensitivity of the executive s portfolio to a change in stock price and to estimate exercise price and time-to-maturities for options held by the CEO at the beginning of the year. I report the summary statistics of the portfolio incentives in Panel B of Table The determinants and consequences of negotiation grants 4.1. The determinants of negotiation grants to target CEOs The first set of tests focuses on the decision to grant equity and whether it is driven by the board s decision to compensate the CEO, to provide bargaining incentives, or both. Specifically, I estimate probit and ordinary least squares (OLS) regressions of the following model using up to three observations for each CEO: Grant it ¼ b 0 þb 1 DNEG t þb 2 Expected premium i DNEG t þb 3 CEO negotiates i DNEG t þb 4 Buyer initiates i DNEG t þb 5 %Outside directors i DNEG t þb 6 Expected premium i þb 7 CEOnegotiates i þb 8 Buyer initiates i þb 9 %Outside directors i þy k Controls kit þf year þf ind þu it : ð3þ The left-hand side variable is either the probability of a grant or the pay-performance sensitivity of grants at target i in year t, in which t equals zero in the period of negotiations. DNEG is a dummy variable equal to one for observations during the negotiation period and zero for those during the pre-negotiation period. Expected premium is the average total premium paid in deals within the firm s industry over the 12 months prior to announcement of the deal. The industry is defined at the four-digit Standard Industrial Classification (SIC) level