Journal of Financial Economics 108 (2013) 659–674

Contents lists available at SciVerse ScienceDirect

Journal of Financial Economics journal homepage: www.elsevier.com/locate/jfec

CEO contract design: How do strong principals do it?$ ¨ Henrik Cronqvist a, Rudiger Fahlenbrach b,n a b

Claremont McKenna College, Robert Day School of Economics and Finance, 500 E. 9th St., Claremont, CA 91711, USA ´rale de Lausanne and Swiss Finance Institute, Quartier UNIL-Dorigny, 1015 Lausanne, Switzerland Ecole Polytechnique Fe´de

a r t i c l e i n f o

abstract

Article history: Received 22 February 2012 Received in revised form 21 August 2012 Accepted 28 August 2012 Available online 4 February 2013

We study changes in chief executive officer (CEO) contracts when firms transition from public ownership with dispersed owners to private ownership with strong principals in the form of private equity sponsors. The most significant changes are that a significant portion of equity grants performance-vests based on prespecified measures and that unvested equity is forfeited by fired CEOs. Private equity sponsors do not reduce base salaries, bonuses, and perks, but redesign contracts away from qualitative measures. They use some subjective performance evaluation, do not use indexed or premium options, and do not condition vesting on relative industry performance. We compare the contracts to predictions from contracting theories, and relate our results to discussions of executive compensation reform. & 2013 Elsevier B.V. All rights reserved.

JEL classification: G32 G34 J33 Keywords: LBOs Employment contracts Contracting theory Executive compensation

$ We thank Steve Kaplan (the referee), Bo Becker, Ola Bengtsson, Harry DeAngelo, Alex Edmans, Gerry Garvey, Rajna Gibson, Stu Gillan, Eric Helland, Jean Helwege, Eric Hughson, Han Kim, Mike Lemmon, Ron Masulis, Ernst Maug, Lisa Meulbroek, Eric Novak, Ludo Phalippou, Josh ¨ Rosett, Morten Sorensen, Pascal St. Amour, Neal Stoughton, Gunter ¨ Strobl, Per Stromberg, Rene´ Stulz, Steen Thomsen, Alexander Wagner, Mike Weisbach, Karen Wruck, senior executives in the U.S. private equity industry, and seminar and conference participants at Claremont McKenna College, Copenhagen Business School, European Finance Association, Geneva Finance Research Institute, Ohio State University, 7th ¨ Annual Meeting of the Swiss Finance Institute, Studieforbundet ¨ ¨ Naringsliv och Samhalle (SNS), Universita Cattolica del Sacro Cuore, and University of Lausanne (Department of Economics) for helpful discussions and comments. Jennifer Volk provided excellent research assistance. Financial support from the Financial Economics Institute at Claremont McKenna College (Cronqvist) and the Swiss Finance Institute and the Swiss National Science Foundation (Fahlenbrach) is gratefully acknowledged. Part of this research was carried out while Cronqvist was visiting researcher at the Finance and Accounting Department, China Europe International Business School (CEIBS) and while Fahlenbrach was visiting researcher at the Center for Corporate Governance, Department of International Economics and Management, Copenhagen Business School. n Corresponding author. Tel.: þ41 21 693 0098. E-mail addresses: [email protected] (H. Cronqvist), ruediger.fahlenbrach@epfl.ch (R. Fahlenbrach).

0304-405X/$ - see front matter & 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jfineco.2013.01.013

1. Introduction We contribute a new perspective on executive compensation research by studying changes to CEO employment contracts implemented by some of the most sophisticated and financially savvy principals in U.S. capital markets: private equity sponsors. If the changes in a firm’s governance structure following a leveraged buyout (LBO) allow for arm’s-length bargaining between private equity (PE) sponsors, as ‘‘strong principals,’’ and the CEOs of the portfolio companies as their agents, we may observe changes to contract features of importance to the private equity sponsors (e.g., Jensen, 1989). Our objective in this paper is to answer three questions. First, do the strong principals redesign CEO contracts? If they do, which contract features do they change? We examine a comprehensive set of features of CEO contracts in addition to cash pay, such as perquisites, equity incentives, vesting conditions, and severance pay. Second, how do the CEO contracts designed by PE sponsors square with contracting theories? Finally, do the CEO

660

H. Cronqvist, R. Fahlenbrach / Journal of Financial Economics 108 (2013) 659–674

contracts we study avoid some of the most criticized compensation practices in U.S. public firms? Regulators and shareholder interest groups should be interested in whether their proposals differ markedly from contracts where a shareholder with significant ownership and financial expertise bargains with a CEO. While existing work has established that executives own a larger percentage of the equity after an LBO (e.g., Kaplan, 1989; Kaplan and Stein, 1993; Leslie and Oyer, 2009), surprisingly little is known about whether PE sponsors change other features of CEO contracts. Baker and Wruck (1989) study the O.M. Scott & Sons Company, and Denis (1994) studies Safeway, pre- and post-LBO, but other than these case studies from the early 1980s LBO wave, we are not aware of any study of changes in CEO contracts which use data on private equity transactions. We compile a new data set of CEO employment contracts, change-of-control agreements, equity incentive plans, and equity rollover agreements. Our final sample is small and nonrandom; it involves 20 large firms and the largest PE sponsors in U.S. capital markets (e.g., Blackstone, Carlyle, and Kohlberg, Kravis & Roberts). But our approach still provides several advantages. First, while we have selected PE sponsors that have been successful in the past, this may actually strengthen our analysis because we are likely to have selected value-maximizing principals. Second, the success of a PE sponsored firm is uncertain immediately after the LBO, enabling us to study ex ante contracts rather than ex post data possibly distorted by (the anticipation of) a liquidity event. Finally, other studies of PE sponsors (e.g., Baker and Wruck, 1989; Denis, 1994) and large shareholders (e.g., Holderness and Sheehan, 1991; DeAngelo and DeAngelo, 2000; Becht, Franks, Mayer and Rossi, 2009) as strong principals have resulted in important and widely cited evidence despite a sample size of one in each of those studies.1 We find that the strong principals redesign many, but not all, CEO contract features. First, we find increases in base salary and bonuses by 25%. Salary increases are concentrated in firms who appoint new executives and are potentially required to make the CEO take on the extra risk that working in a levered firm with a strong owner and a more performance-sensitive contract entails. Increases in target bonuses are observed across both retained and newly appointed CEOs. Consistent with principal-agent models and the informativeness principle ¨ (e.g., Holmstrom, 1979; Harris and Raviv, 1979; Innes, 1990), PE sponsors contract on several signals that likely correlate with CEO effort. They redesign contracts away from qualitative, nonfinancial, and earnings-based measures. Instead, they contract on cash flow-based measures, such as earnings before interest, taxes, depreciation and amortization (EBITDA), that may allow for less accounting discretion with respect to depreciation, amortization, and taxes. For longer-term performance, the principals contract on internal rates of return (IRRs) or multiples.

While we find no evidence of a change in the multiplier for CEO severance cash pay, a significant change is that unvested options and restricted stock are forfeited to a larger extent if a CEO is dismissed. Even though the PE sponsored firm is private, which restricts resales per se, the PE sponsors further restrict the resale market for vested shares for dismissed executives. In particular, they do so by a right of first refusal and by limiting the set of parties to which executives can sell their shares, making it practically impossible for dismissed CEOs to unwind equity. This evidence is consistent with theories predicting reduced severance pay contracts when governance is stronger (e.g., Maug, 1997; Almazan and Suarez, 2003). Finally, perquisites such as personal usage of corporate aircraft and tax gross-ups, remain unchanged after the PE transaction. This result is particularly surprising given the bad reputation of these perks in the media, but is consistent with a ‘‘perks as productivity’’ argument (e.g., Rajan and Wulf, 2006). We find some evidence that private equity sponsors allow retained CEOs to carry over previously acquired perks, while the perquisites granted to newly appointed CEOs are smaller, although the economic magnitude of these changes is small. The PE transactions we as financial economists analyze are not controlled or natural experiments with exogenous shifts in corporate control for random firms. As a result, at least two caveats apply to any study of LBOs. One is that PE sponsored firms are special (e.g., they may have significantly more cash flow than valuable investment opportunities). While this may have been the case for the 1980s, our sample consists of middle-aged firms in growing industries such as IT/telecom/media, healthcare, ¨ and services, consistent with Kaplan and Stromberg’s (2009) description of LBOs since 1990. Another concern is that the redesign of the CEO contract could be driven by the change in capital structure, not by the strong principal. We examine two matched samples, but do not find that the boards of public firms that undertake voluntary leverage changes (leveraged recapitalizations or large debt-financed acquisitions) comparable in magnitude to those of LBO-sponsored firms also change CEO contracts to those we find in our LBO sample.2 Our research significantly extends existing work on LBOs, strong principals, and CEO contracts. For example, a previously undocumented change is that almost all of our sample firms use performance-vesting post-LBO. A common contract is that if the agent produces certain multiples or IRRs for the principal at a future exit event, about 50% of equity grants will performance-vest. Using vesting conditions that focus on an exit event, the PE sponsors are able to align the incentives of executives with their own objectives. Because at an exit event, the private firm now receives a precise valuation, this contract change is in line with contracting models that suggest maximizing the sensitivity of the agent’s compensation to a signal correlated with the agent’s effort. Importantly, equity grants change from a time-vesting to a performance-vesting

1 Other small-sample studies related to ours include Andrade and Kaplan (1998) and Holderness and Sheehan (1985).

2 For evidence on leveraged recapitalizations from the 1980s, see, e.g., Denis and Denis (1993), Denis (1994), and Wruck (1994).

H. Cronqvist, R. Fahlenbrach / Journal of Financial Economics 108 (2013) 659–674

schedule even when the former CEO is retained, suggesting that this part of the contract is non-negotiable and cannot be part of a ‘‘quid pro quo’’ scheme. Our paper is also related to additional strands of the corporate finance literature. We assume that the PE sponsors are value-maximizing principals, consistent with Kaplan (1989), Baker and Wruck (1989), Lerner, ¨ Sorensen and Stromberg (2011), and Guo, Hotchkiss and Song (2011). The PE sponsors we study have strong incentives to maximize value, receive few private benefits of control, and the partners controlling the sponsor generally receive at least 20% of the value-added.3 The approach of comparing observed contracts to those predicted by contracting theory is related to Kaplan and ¨ Stromberg’s (2003) work on venture capital contracts. Our study is also related to work on changes in governance pre- and post-LBO. Gertner and Kaplan (1996) examine one governance mechanism, the board of directors, and find that private equity sponsors change the board so that they have control post-LBO. Muscarella and Vetsuypens (1990) and Leslie and Oyer (2009) show differences in equity incentives between LBO firms and public firms by focusing on reverse LBOs and a set of control firms. CEO compensation in non-LBO private firms has been studied by Cole and Mehran (2008), Bengtsson and Hand (2010), and Gao, Lemmon and Li (2012). Our work is also related to studies of CEO contracts in public firms (e.g., Kole, 1997; Gillan, Hartzell and Parrino, 2009). The paper is organized as follows. Section 2 describes the construction of our data set and reports summary statistics. Section 3 compares the design of CEO contracts before versus after the LBO. Section 4 relates our results to contracting theories. Section 5 examines changes in employment contracts in a sample of public firms with large changes in leverage. Section 6 relates our findings to recent discussions of executive compensation reform. Section 7 concludes. 2. Data and sample construction 2.1. Data sources We hand-collect a new data set on CEOs’ compensation, perquisites, severance pay, equity incentives, and vesting conditions.4 We take advantage of the fact that some private equity sponsored private firms are required to file with the 3 Chung, Sensoy, Stern and Weisbach (2012) show that the lifetime income of private equity general partners is also indirectly affected by the effect of the current fund’s performance on the ability to raise capital for future funds. This indirect pay for performance from future fundraising is of the same order of magnitude as compensation from carried interest. Some researchers have however estimated that as much as two-thirds of the expected revenues for large LBO funds come from fees that are not performance-sensitive (Metrick and Yasuda, 2010). 4 Leslie and Oyer (2009, p. 6) study reverse LBOs because such firms have to disclose executive compensation arrangements for two years prior to the initial public offering (IPO). Yet, they conclude that ‘‘[i]t would be ideal to obtain data on all managers’ equity investments, option grants, bonus structures and firing/hiring of managers. [y] By its very nature, it is difficult to obtain information on the practices of private equity firms for a broad sample, let alone such confidential details as managerial incentives.’’

661

Securities and Exchange Commission (SEC) post-LBO as they have public debt [see, e.g., Cotter and Peck (2001) and Helwege and Packer (2009), who follow a similar approach, for details on these filing requirements.] We also take advantage of the 2006 amendments to the disclosure rules under the Securities Exchange Act of 1934, which require firms to provide more details about executive compensation. We collect data from SEC filings, not only 10-Ks or proxy statements, but also employment contracts, equity incentive plans, and equity rollover agreements. These additional documents are most often filed as appendices to 10-K or 10-Q statements. The resulting data set provides a detailed overview of the contracts between shareholders and CEOs before and after the LBO. Data on the LBO transactions are collected from Bloomberg. All accounting data for public firms come from Standard and Poor’s (S&P) Compustat. Accounting data post-LBO are hand-collected from 10-K statements. 2.2. Sample selection issues We study a small and nonrandom sample. We have data on CEO contracts for 20 LBOs of large public firms between 2005 and 2007. Prior to 2005, employment agreements and compensation contracts offer significantly less information. The 2007/2008 financial crisis caused a collapse of the high-yield debt market used by large PE sponsors. As a result, at this time, it is not possible to extend the sample significantly beyond the current number of CEO contracts. Another concern, common to all LBO research, is the nonrandom nature of samples of LBO firms. Our summary statistics in Section 2.3 show that our sample is representative of samples of buyouts in the most recent LBO wave such as those examined by, e.g., Kaplan and ¨ Stromberg (2009) and Guo, Hotchkiss and Song (2011). ¨ In addition, Kaplan and Stromberg (2009) show that the recent LBO wave targeted a set of firms more representative of the entire universe of firms. A related concern is that private equity sponsors select firms which would have redesigned CEO contracts regardless of the appearance of a private equity sponsor. However, we argue that it is unlikely that the boards of the firms we study, in the absence of a private equity sponsor, would have changed CEO contracts at exactly the time of the sponsor’s arrival and to exactly those that the sponsors design. The small sample implies that we have to be cautious in our conclusions. For example, our comparisons in Section 3 focus mostly on the economic significance of our evidence, although we provide p- and Z-values for the differences in means and medians across the LBO transaction for the compensation and governance variables. 2.3. Summary statistics Appendix A lists the private equity transactions in our sample, the announcement dates, and the total deal value of these transactions. Panel A of Table 1 shows summary statistics for total value, premium over the share price at the announcement date, and valuation multiples. The data confirm that our sample consists of large LBOs:

662

H. Cronqvist, R. Fahlenbrach / Journal of Financial Economics 108 (2013) 659–674

Table 1 Summary statistics. Deal characteristics are from Bloomberg. Announced total value is calculated by Bloomberg as offer price multiplied with shares outstanding plus net debt. Premium is defined as the offer price over the stock price at the deal announcement date, expressed in percent. EBITDA multiple is equal to the announced total value divided by trailing 12-month EBITDA. Total assets multiple is announced total value divided by total assets from the last 10-Q prior to the transaction. Target characteristics are measured at the end of the last fiscal year prior to the LBO announcement date and are based on accounting data from Compustat. The sample consists of 20 large leveraged buyouts of publicly listed US firms between 2005 and 2007. Average

Median

St. dev.

9,132.64 18.93 11.43 1.91 0.60 0.60

6,182.69 17.75 10.47 1.56 1.00 1.00

9,642.67 11.04 4.08 1.09

5,001.39 37.03 6,440.87 0.18 16.32 0.11 0.08 0.05 0.24 0.50

2,545.22 13.95 2,481.65 0.15 7.95 0.06 0.07 0.05 0.26 0.50

5,749.98 64.29 8,861.09 0.15 17.73 0.16 0.05 0.03 0.15 0.51

Panel A: Deal characteristics Announced total value ($ million) Premium (%) EBITDA multiple Total assets multiple Club deal (yes ¼1) Same CEO pre- and post-LBO (yes ¼ 1) Panel B: Target characteristics Total sales ($ million) Number of employees (thousands) Total assets ($ million) Book leverage Interest coverage Cash/ assets Return on assets Capital expenditures/ assets Property, plant, and equipment/ assets Dividend payer (yes ¼1)

Table 2 Changes in corporate governance. Data on ownership by private equity sponsors post-LBO come from company filings with the SEC. This information is contained in the section ‘‘security ownership of certain beneficial owners and management.’’ Data on boards of directors come from company filings with the SEC. Data on block ownership pre-LBO come from the Corporate Library. The sample consists of 20 large leveraged buyouts of publicly listed U.S. firms between 2005 and 2007. Statistically significant differences in means and medians in board characteristics pre- and post-LBO are reported in the post-LBO mean and median columns using ***, **, and * to indicate statistical significance at the 1%, 5%, and 10% level, respectively. Average

Median

St. dev.

Min.

Max.

20.0% 2.75

20.2% 2

15.5% 2.15

0.0% 0

61.0% 8

91.8%

97.0%

10.0%

68.1%

100.0%

Pre-LBO Board size Management directors Outside directors Former employee directors

9.6 1.6 7.4 0.6

10 1 8 0

2.3 0.8 2.1 0.9

5 1 3 0

14 3 11 3

Post-LBO Board size LBO sponsor directors Management directors Outside directors Former employee directors

8.3* 5.5*** 1.2 1.4*** 0.2**

9* 5*** 1 1*** 0**

2.6 2.5 0.4 1.6 0.4

4 2 1 0 0

13 11 2 5 1

Panel A: Ownership characteristics Pre-LBO Outside block ownership Number of outside blocks Post-LBO Private equity sponsors Panel B: Board characteristics

the average (median) transaction value is $9.1 billion ($6.2 billion). The largest transaction is the LBO of HCA Inc. with a total deal value of $32 billion. The premium is, on average, 19% (median 18%). Guo, Hotchkiss and Song (2011) find that the average premium, defined as percentage difference of price per share paid and stock price one month prior to the LBO, is 29% in their sample of large buyouts 1990–2006. Some of the difference may

be attributable to the fact that ‘‘club deals’’ have lower premiums (e.g., Officer, Ozbas and Sensoy, 2010), and that 60% of our sample are club deals, i.e., they are financed by a consortium of private equity sponsors. Guo, Hotchkiss and Song (2011) find that the average EBITDA multiple is 11.4 for their LBOs, which is the same as in our sample. Finally, the table shows that 60% of the pre-LBO CEOs remain as CEOs after the PE transactions.

H. Cronqvist, R. Fahlenbrach / Journal of Financial Economics 108 (2013) 659–674

663

Table 3 Design of CEO cash pay, annual bonus, and perquisites. Data on CEO compensation come from the tables and narrative on executive compensation in company filings with the SEC. Tax gross-up is compensation paid to the CEO to cover the tax liability for perquisites or other elements of compensation that trigger a tax liability (such as option exercises). The sample consists of 20 large leveraged buyouts of publicly listed U.S. firms between 2005 and 2007. Statistically significant differences in means and medians in salary, target bonus, and perquisites pre- and post-LBO are reported in the post-LBO column using ***, **, and * to indicate statistical significance at the 1%, 5%, and 10% level, respectively. Pre-LBO

Base salary Target bonus Bonus based on EPS or net income Bonus based on EBITDA Bonus based on other criteria Perquisites: Dollar value Perquisites: Personal use of corporate aircraft (%) Perquisites: Tax gross-up (%)

Post-LBO

Average

Median

Average

Median

% With increase

% With decrease

757,283 1,006,931 60% 10% 30%

815,833 850,000

945,797** 1,272,024** 10% 80% 10%

1,000,000*** 1,166,923**

65% 71%

10% 29%

42,599 42% 60%

17,662

44,078 40% 70%

18,288

38%

31%

Panel B reports target firm characteristics from the last fiscal year prior to the LBO announcement date. Our sample firms have, on average, $5.0 billion in sales and 37,000 employees (median sales of $2.5 billion and median employees of 14,000). They have low book leverage at 18% and interest coverage above 16. They have 11% of cash relative to total assets, a return on assets (ROA) of 8%, and half of the firms are dividend payers.

are consistent with the evidence by Cotter and Peck (2001) for U.S. buyouts in the 1980s. Gertner and Kaplan (1996) and Katz (2009) examine boards around reverse LBOs and also find that private equity sponsors have a large representation on the board. To conclude, PE sponsors have enough ownership and board control to redesign CEO contracts if they find it optimal to do so after the LBO.

2.4. Changes in corporate governance

3. Design of CEO compensation and employment contracts

Panel A of Table 2 reports changes in ownership. Prior to the LBO, the average firm had an aggregate outside block ownership of approximately 20.0%, held by, on average, 2.8 blockholders.5 Not surprisingly, the ownership concentration changes substantially. After the LBO, the private equity sponsors own, on average, 91.8% of the equity of the firm. Panel B shows changes in board control. The average board in our sample decreases by 1.3 directors (14%) to 8.3 directors, statistically significant at the 10% level. Of those, 5.5 are representatives of the private equity sponsors, 1.2 are employees, and 1.4 are outside directors (other than PE sponsors). Private equity sponsors hence have majority control of the board. The relatively small decrease in board size and the large number of PE directors in our sample relative to changes shown in prior studies is likely driven by the fact that many of our deals are club deals, and that each PE firm of the consortium seeks board representation. Compared to prior to the buyout, most of the outside directors are replaced by directors representing the sponsors so that the decrease in outside directors across the transaction is strongly statistically significant. Our results 5 We note that the firms in our sample do not have completely dispersed ownership pre-LBO. Some recent research has found that a blockholder in a public U.S. firm affects CEO pay and other policies (e.g., Cronqvist and Fahlenbrach, 2009; Becker, Cronqvist and Fahlenbrach, 2011). The result that the average firm in our sample has 2.75 blockholders pre-LBO makes it even more interesting to analyze whether there is any marginal effect on CEO contracts after the arrival of the PE sponsor.

In this section, we compare CEO contracts before and after the PE transactions in our sample. This section simply provides empirical facts based on our sample. 3.1. Base salary, bonus, and perquisites Table 3 reports the average and median of CEO salary, bonus, and perquisites. Columns 1 and 2 show data preLBO, while columns 3 and 4 characterize post-LBO contracts. Columns 5 and 6 report the percentage of firms with increases or decreases, respectively, in the compensation variables, if applicable. We first examine CEO salary and target bonus data. We find that prior to an LBO, the average (median) base salary is $757,000 ($816,000). After the arrival of a PE sponsor, the average (median) salary increases to $946,000 ($1,000,000), i.e., a 25% (23%) increase. Both mean and median changes are strongly statistically significant. The CEO’s salary increases in 65% of the firms. We find that target bonuses also increase after a private equity transaction. The average (median) bonus target prior to the LBO is $1,007,000 ($850,000), but it increases to $1,272,000 ($1,167,000) after the LBO, corresponding to a 26% (37%) increase. The CEO’s target bonus increases in 71% of our sample firms.6 6 With 20 sample observations, the reported percentages should change in increments of 5%. However, we do not observe all variables for all sample observations.

664

H. Cronqvist, R. Fahlenbrach / Journal of Financial Economics 108 (2013) 659–674

Table 4 New and retained CEOs: cash pay, annual bonus, and perquisites. Data on CEO compensation come from the tables and narrative on executive compensation in company filings with the SEC. The sample consists of 20 large leveraged buyouts of publicly listed U.S. firms between 2005 and 2007. Observations are used if pre- and post-LBO compensation data are available. Pre-LBO

Post-LBO

Average

Median

Average

Median

Same CEO (N ¼ 12) Base salary Target bonus Perquisites: Dollar value

849,819 1,112,454 35,337

841,500 850,000 11,721

906,954 1,224,355 35,977

845,357 1,105,000 13,200

Newly hired CEO (N ¼ 7) Base salary Target bonus Perquisites: Dollar value

706,833 1,075,405 45,567

815,833 1,100,000 34,146

983,214 1,310,000 13,656

1,000,000 1,350,000 25,492

We also find that bonus criteria are redesigned away from qualitative and nonfinancial measures to EBITDA. CEO contracts therefore shift from earnings-based measures, which Wall Street and analysts care about, to cash flow-based measures which PE sponsors care about. Also, it has been argued that EBITDA is more difficult to manipulate because it allows for less discretion with respect to depreciation, amortization, and taxes. The change in the use of bonus measures is significant: only 10% of CEOs pre-LBO are benchmarked in terms of their bonus against EBITDA, compared to a vast majority (80%) post-LBO. The table also shows changes in perquisites. The median dollar value of perquisites stays unchanged at approximately $18,000 per year. Consistent with a nonchange in perks after LBOs, we find increases and decreases, respectively, in perks for about a third of the CEOs each. Yermack (2006a) focuses on one specific perk, personal use of company aircraft, and finds that such use is a private benefit that cannot be explained by anything but golf club memberships far away from corporate headquarters. Table 3 shows that about 40% of firms allow personal aircraft use both pre- and post-LBO, so the incidence of personal aircraft use does not appear to change significantly with the arrival of a strong principal in the firm.7 We also ask to what extent the increases in base salary and bonus depend on whether the PE sponsor retains the CEO or not. Retained CEOs may bargain for larger increases in salary and bonus, as a ‘‘quid pro quo’’ to ensure an LBO without resistance. On the other hand, outside CEOs of public firms have been found to command a pay premium over insiders. In our sample, 12 firms (60%) retain the former CEO. We report changes in salary, bonus, and perquisites across the LBO for these two groups in Table 4 (but encourage the reader to keep

7 Our results are consistent with the evidence in Edgerton (2012). He studies company aircraft usage prior to and after LBOs, and while he finds that private equity transactions may decrease extreme pre-LBO usage, he finds no average effect. It is also important to recognize that our analysis does not consider substitution from company aircraft use prior to an LBO to CEO usage of the sponsor’s aircraft after an LBO.

the small sample sizes in mind).8 We find that the salary increases are modest for firms who retain the CEO. The median base salary remains essentially unchanged, and the average salary increases by 6.6%. Regarding the target bonus, the average stays approximately the same, but the median increases by 30%. There is little evidence that perks decrease for this group; retained CEOs essentially maintain their perquisites. We find a large salary and bonus increase in firms who appoint new CEOs. For this group, the median salary increases by 22.6%, and the median bonus increases by 22.7%. Interestingly, the median dollar level of perks decreases for newly hired CEOs by 25%, although the dollar decrease of $8,500 is small relative to the increase in cash compensation.

3.2. Severance and separation pay Table 5 reports data on severance/separation contracts upon termination without cause (Panel A) and change of control (Panel B). Columns 1 and 2 show the averages and medians pre-LBO, while columns 3 and 4 characterize post-LBO contracts. We find in Panel A that the average contractual severance pay prior to an LBO is 2X base salary and past annual bonus. Two years of total cash pay corresponds to the average separation pay in large U.S. firms (e.g., Rusticus, 2006). The average contract with respect to cash pay in case of dismissal is not redesigned after the arrival of a strong principal in the firm; it stays at 2X base salary and bonus. However, because salary and bonus increase by 25%, on average, the ex ante dollar amount of severance pay increases as well. One caveat is that our study focuses on ex ante CEO contracts rather than ex post payments upon dismissals, and Yermack (2006b) finds that a significant amount of severance pay is awarded on a discretionary basis by the board. On the other hand, ex ante contracts provide the minimum separation pay, so that as long as the PE sponsors we study do not pay more 8 The founder and CEO of Univision worked prior to the LBO for a salary and bonus of zero, and the newly hired CEO received a regular salary and bonus. We remove this observation from the sample when creating Table 4.

H. Cronqvist, R. Fahlenbrach / Journal of Financial Economics 108 (2013) 659–674

665

Table 5 Data on CEO severance payments upon termination without cause and payments in case of change of control come from the tables and narrative on executive compensation in company filings with the SEC or from executive employment contracts also filed with the SEC. The sample consists of 20 large leveraged buyouts of publicly listed U.S. firms between 2005 and 2007. Design of severance and separation pay contracts

Pre-LBO Average

Post-LBO Median

Average

Median

2.3 2.2 29% 71%

2 2

2.1 1.8 61% 46% 71%

2 2

2.8 2.8 12% 90%

3 3

2.4 2.3 16% 100%

2 2

Panel A: Contractual arrangements upon termination without cause/ resignation for good reason Base salary (multiple paid) Past annual bonus (multiple paid) Are unvested options canceled? (yes ¼ 1) Do restricted shares vest immediately? (yes ¼1) Does PE sponsor have right of first refusal? Panel B: Contractual arrangements upon change of control Base salary (multiple paid) Past annual bonus (multiple paid) Are unvested options canceled? (yes ¼ 1) Do restricted shares vest immediately? (yes ¼1)

on a discretionary basis than do boards of public firms, our conclusions still apply. Turning to unvested options and restricted shares, we find differences pre- and post-LBO. First, the vesting conditions become more restrictive for a terminated CEO. After the LBO, unvested options are canceled in the event of a CEO termination in 61% of the firms, compared to only 29% prior to the private equity transaction. Restricted shares vest immediately upon termination of the CEO in 71% of the firms pre-LBO, but only in 46% post-LBO. Second, the principals also redesign contracts to impose additional costs on terminated CEOs, even if some equity vests or has already vested. In 71% of the firms, the private equity sponsors have the right of first refusal to buy back the shares at ‘‘fair market value’’ assessed quarterly by an independent accounting firm. If the sponsors do not exercise this option, a dismissed CEO is stuck with the shares until a liquidity event for the sponsor, because the typical CEO has also signed a shareholder agreement severely restricting the outside parties to which he can sell. It is important to emphasize that even though the PE sponsored firm is private, which restricts resales per se, the PE sponsors apparently further restrict the resale market for vested shares for dismissed executives. We find in Panel B that the average contractual separation pay upon change of control prior to an LBO is 3X base salary and past annual bonus. Fich, Tran and Walkling (2012) report average payments upon change of control of 2.95X total cash pay for large U.S. public firms. We find that the average change of control pay is reduced to 2X post-LBO. An explanation may be that because of the sponsor’s board control post-LBO, there is much less of a requirement to compensate the CEO for not resisting a change of control event, compared to a public firm.

3.3. Equity incentives In this subsection, we report data on the CEOs’ equity ownership pre- and post-LBO, the required buy-in and permitted cash-out in the LBO transaction, the pre- and

post-LBO equity security mix, as well as the change in vesting conditions of the equity grants post-LBO. 3.3.1. CEO equity ownership, buy-in, and cash-Out We first compare percentage ownership and dollar cash-out and buy-in across the LBO transaction.9 Panel A of Table 6 shows statistics for the percentage ownership of CEOs prior to the LBO and immediately after the LBO transaction as well as statistics for the percentage ownership of CEOs of a sample of control firms. For each LBO firm, we find a control firm in the same Fama-French 48 industry with the closest total assets at the end of the last fiscal year prior to the LBO. For these control firms, we record CEO percentage ownership at the end of the fiscal year prior to the LBO and at the end of the fiscal year after the LBO. Table 6, Panel A shows two percentage ownership statistics for the LBO and control sample. The pre- and post-LBO CEO equity ownership (%) numbers include shares owned outright and options and restricted stock which vest in the next 60 days after the record date. The pre- and post-LBO CEO equity ownership (%), fully diluted numbers include shares owned outright and all options and restricted stock, including unvested grants. The fully diluted ownership percentages better reflect potential ownership after the LBO, because unvested options and stock are typically meaningful after the transaction (Kaplan and Stein, 1993). To calculate fully diluted ownership, we assume that all performance goals for performance vesting stock and options are met. Panel A of Table 6 shows that prior to the LBO, the median CEO owns 0.94% of the public firm or 1.13% of the firm on a fully diluted basis. The median CEO percentage ownership of the industry- and size-matched control sample is remarkably close to that of the LBO sample at 1.00% or 1.12% on a fully diluted basis. 9 The statistics presented in Table 6 exclude two sample firms, Aramark and Univision, for ease of exposition. In both of these transactions, the CEOs owned large stakes of the company and used the LBO as a liquidity event. These two observations hence significantly distort the means of the summary statistics.

666

H. Cronqvist, R. Fahlenbrach / Journal of Financial Economics 108 (2013) 659–674

Table 6 CEO equity ownership, buy-in, and cash-out. Data on the equity ownership of the CEO pre- and post-LBO come from the table ‘‘Security Ownership of Certain Beneficial Owners and Management’’ and from the tables ‘‘Grants of Plan-based Awards’’ and ‘‘Outstanding Equity Awards at Fiscal Year-end’’ filed with the SEC either in Forms 10-K or DEF 14A. The pre-LBO and post-LBO CEO equity ownership (%) numbers include shares owned outright and options and restricted stock which vest in the next 60 days. The pre-LBO and postLBO CEO equity ownership (%), fully diluted numbers include shares owned outright and all options and restricted stock, including unvested grants. Panel A shows statistics for CEO equity ownership for 18 sample observations as well as statistics for a control group of firms that were matched by the Fama-French 48 industries and total assets in the year of the LBO. Two observations are excluded from the statistics, because a founder CEO used the LBO as a liquidity event. The overall sample consists of 20 large leveraged buyouts of publicly listed U.S. firms between 2005 and 2007. Panel B shows statistics for 11 sample observations in which the CEO was retained, and Panel C shows the statistics for seven sample observations in which the CEO changed across the LBO. In Panels B and C, two additional variables are presented. Cash-out (dollars) is the total dollar value of the CEO’s stock, restricted stock, and option portfolio, as well as the dollar value of the extra bonus paid for a successful completion of the merger. The information on the cash-out variable comes from form DEFM 14A, in particular the sections ‘‘Interests of the Company’s Directors and Executive Officers in the Merger’’ and ‘‘Security Ownership of Certain Beneficial Owners and Management.’’ Buy-in (dollars) is the total dollar value of the CEO’s equity purchases at the LBO. The information on the buy-in comes from the CEO’s employment contract or the 10-K section on ‘‘executive compensation.’’ Panel A: CEO ownership statistics for sample of LBO firms and sample of industry- and size-matched control firms Average Median Pre-LBO CEO equity ownership (%) (LBO sample) CEO equity ownership (%) (control sample) CEO equity ownership (%), fully diluted (LBO sample) CEO equity ownership (%), fully diluted (control sample) Post-LBO CEO equity ownership (%) (LBO sample) CEO equity ownership (%) (control sample) CEO equity ownership (%), fully diluted (LBO sample) CEO equity ownership (%), fully diluted (control sample)

2.46% 1.87%

0.94% 1.00%

2.74%

1.13%

2.11%

1.12%

2.66% 1.76%

0.80% 1.00%

4.05%

2.52%

1.97%

1.06%

2.70

2.33

0.94

0.96

Comparisons Post-LBO % ownership/ pre-LBO % ownership, fully diluted (LBO sample) Post-LBO % ownership/ pre-LBO % ownership, fully diluted (control sample)

Panel B: CEO ownership statistics for sample of LBO firms that retained the CEO Average

Median

Pre-LBO CEO equity ownership (%) (fully diluted) 3.69% 1.40% Cash-out (dollars) $56,456,751 $43,808,000 Post-LBO CEO equity ownership (%) (fully diluted) 5.59% Buy-in (dollars) $24,605,734 Comparisons Post-LBO % ownership/ pre-LBO % 2.86 ownership (fully diluted) Buy-in (dollars)/ cash-out (dollars) 0.33

3.17% $9,250,025 2.26 0.28

Table 6 (continued ) Panel C: CEO ownership statistics for sample of LBO firms that changed the CEO Average

Median

Pre-LBO CEO equity ownership (%) (fully diluted) 1.24% 0.62% Cash-out (dollars) $35,589,089 $34,003,956 Post-LBO CEO equity ownership (%) (fully diluted) Buy-in (dollars) Comparisons Post-LBO % ownership/ pre-LBO % ownership (fully diluted) Buy-in (dollars)/ cash-out (dollars)

1.64% $5,100,000

1.44% $2,000,000

2.42

2.98

0.14

0.04

The next block of statistics shows the ownership percentages post-LBO for sample firms and the ownership percentage for the sample of control firms in the year after the LBO of the matched firm. While the median ownership percentage of the control sample stays virtually unchanged at 1.06% on a fully diluted basis, the fully diluted CEO ownership percentage in LBO firms increases dramatically to 2.52% after the LBO.10 The mean (median) ratio of fully diluted post-LBO % ownership to fully diluted pre-LBO % ownership is 2.70 (2.33). Panels B and C of Table 6 split the sample into those observations in which the CEO is retained across the transaction (Panel B) and those in which a new CEO is hired (Panel C). Panels B and C contain two additional variables. Cash-out (dollars) is the total dollar value of the prior CEO’s stock, restricted stock, and option portfolio, as well as the dollar value of the extra bonus paid for a successful completion of the merger. The information on the cash-out variable comes from SEC Form DEFM 14A, in particular, the sections ‘‘Interests of the Company’s Directors and Executive Officers in the Merger’’ and ‘‘Security Ownership of Certain Beneficial Owners and Management.’’ Buy-in (dollars) is the total dollar value of the CEO’s equity purchases at the LBO. The information on the buy-in comes from the CEO’s employment contract or the 10-K section on ‘‘executive compensation.’’ Panel B of Table 6 shows that for the sample of retained CEOs, the ratio of post-LBO ownership to preLBO ownership is, on average, 2.86 (median: 2.26). The percentage ownership of the CEO increases substantially after the LBO, consistent with the findings of earlier studies.11 However, we also find in Panel B of Table 6 that those CEOs who are retained in an LBO ‘‘cash-out.’’ While their percentage equity ownership increases, the dollar amount is reduced, often dramatically. The average retained CEO cashes out $56.4 million, and the median

10 Note that the increase in post-LBO percentage equity ownership mostly comes from unvested options and stock. 11 E.g., Kaplan and Stein (1993) report for the years 1988 and 1989, two years with significant LBO activity, median ratios of post- vs. prepercentage ownership of the entire management team of 2.86 (1988) and 2.93 (1989).

H. Cronqvist, R. Fahlenbrach / Journal of Financial Economics 108 (2013) 659–674

Table 7 Design of CEO equity incentives. Data on compensation come from the tables and narrative on executive compensation in company filings with the SEC. The first three rows show the equity mix, defined as number of securities divided by total ownership. For the purpose of total ownership, all options are counted with a delta of one, no matter whether they are exercisable or not. The sample consists of 20 large leveraged buyouts of publicly listed U.S. firms between 2005 and 2007. Average

Median

66.9% 0% 4.6% 28.5%

75.6% 0% 0.3% 13.7%

56.6% 0.0% 17.1% 26.3%

74% 0.0% 0% 17.9%

Panel A: Equity security mix pre-LBO Number Number Number Number

of of of of

options/ total ownership indexed options/ total ownership restricted stock/ total ownership unrestricted shares/ total ownership

Panel B: Equity security mix post-LBO Number Number Number Number

of of of of

options/ total ownership indexed options/ total ownership restricted stock/ total ownership unrestricted shares/ total ownership

retained CEO cashes out $43.8 million. In contrast, the average (median) dollar buy-in at the LBO is only $24.6 million ($9.3 million). Hence, the ratio of total dollar buyin to total cash-out stemming from the sale of equity and options at the LBO and contingent bonus payments is 0.33 (median: 0.28). The CEO invests much less than half of what he received for his pre-LBO equity in the firm postLBO. This result is consistent with Kaplan and Stein (1993), who find that the median ratio of dollar equity ownership post-LBO vs. pre-LBO was 0.46 in the 1980s LBO wave. It is also consistent with Jensen’s (1989) observation that leveraging and shrinking the equity piece is an important way to provide executives with strong percentage participation without outlandish dollar values. Panel C of Table 6 shows the same statistics for the sample of LBOs in which the CEO turns over during the transaction. Several observations are interesting. First, prior to the LBO, the median fully diluted percentage ownership of these CEOs is generally lower than that of CEOs who are retained in the deal (0.62% vs. 1.40%). Second, the mean and median ratio of fully diluted postLBO percentage ownership to pre-LBO percentage ownership are roughly comparable across the two panels. Third, new CEOs are required to buy-in less equity in dollar terms than those CEOs who are retained (median of $2,000,000 vs. a median of $9,250,025). Hence, the ratio of buy-in (of new CEO) to cash-out (of old CEO) is much lower than the same ratio for retained CEOs: the median ratio is 0.04 for Panel C and 0.28 for Panel B. 3.3.2. Equity security mix Panel A of Table 7 reports that, prior to the LBO, the average (median) CEO receives equity incentives from the following mix of securities: 67% (76%) options, 5% (0%) restricted stock, and 29% (14%) unrestricted shares, computed by dividing the number of options, restricted shares, and unrestricted shares, respectively, by total

667

ownership, on a fully diluted basis. Panel B of Table 7 shows that we observe little changes in the equity security mix across the LBO at the median, and a slight reduction in the use of options at the mean. We find no evidence of indexed executive stock options either before or after the LBO. PE sponsors do not use indexed CEO compensation contracts that filter out industry or market risk.12 3.3.3. Vesting conditions Turning to vesting conditions, it may be instructive to first provide two specific examples of management incentive plans. Appendix B reports plan details for HCA Inc. and Alltel Corp., which are representative of our sample of firms. We find that most private equity sponsored firms provide their CEOs with a mix of performanceand time-vesting options or restricted shares. In HCA, the CEO was granted three tranches of options: 1/3 timevest, 1/3 vest if the firm achieves certain annual EBITDA performance targets, and 1/3 vest if the sponsors achieve a certain multiple of the price paid in the LBO (50% of the tranche vests at 2X; 50% vests at 2.5X). In Alltel, 69% of the CEO’s options time-vest like standard option grants in public firms, while 31% performancevest, with the threshold being 1.5X or 2X the price paid in the LBO. In Table 8, we find that the vast majority (90%) of CEO contracts contain some performance-vesting equity grants post-LBO. The fraction of performance-vesting equity is dramatically higher than that observed in public firms. In a comprehensive study of 14,000 firmyear observations of U.S. publicly listed firms, Bettis, Bizjak, Coles and Kalpathy (2010) find evidence of performance-vesting equity in only 7% of those firmyears. In our sample, the most common condition for the performance-vesting is the achievement of certain returns for the private equity sponsors, measured either as multiples of the price paid in the LBO, or IRRs. Unless the CEO produces certain multiples or IRRs for the principal, performance-vesting equity will not vest until much later.13 For the average CEO in our sample, the vesting conditions of the equity security mix after the LBO are: 22% (7%) performance-vesting options (restricted stock), 5% premium options, 25% (10%) timevesting options (restricted stock), 5% options granted conditional on minimum buy-in, and the rest is unrestricted shares. We find no evidence that the EBITDA, multiple-of-money, or other targets used in the contracts 12 Meulbroek’s (2002) case study of the Outperform Stock Option (OSO) Plan at Level 3 Communications states that indexed CEO compensation is not common. Level 3’s plan was implemented to ‘‘adequately motivate already financially-secure managers,’’ which seems relevant also for some of the CEOs in our sample that cash-out significantly in the LBO but who are also retained by the LBO sponsor. Yet, we do not find any evidence of indexed CEO compensation contracts among such firms. 13 Technically, these options are of the accelerated vesting type, in that if the performance criteria are not met, they will time-vest. Bettis et al. (2010) report similar results for performance-vesting options in a sample of public firms, and explain the eventual time-vesting of these options with a more favorable accounting treatment.

668

H. Cronqvist, R. Fahlenbrach / Journal of Financial Economics 108 (2013) 659–674

Table 8 Design of CEO equity vesting conditions post-LBO. Data on compensation come from the tables and narrative on executive compensation in company filings with the SEC. The sample consists of 20 large leveraged buyouts of publicly listed U.S. firms between 2005 and 2007. Average Median % With performance vesting equity Number of performance-vesting options/ total ownership Number of performance-vesting restricted stock/ total ownership % With performance-vesting indexed to industry/market Number of premium options/ total ownership Number of time-vesting options/ total ownership Number of time-vesting restricted stock/ total ownership Number of options conditional on buy-in/ total ownership

90.0% 21.7%

100.0% 22.9%

7.1%

0%

0%

0%

4.9% 24.7% 10.0%

0% 28.7% 0%

5.3%

0%

for performance-vesting are indexed. Performancevesting is based on absolute firm performance, not performance relative to the firm’s industry or the market. 4. Comparison with contracting theories There is little empirical evidence that compares actual CEO contracts to the ones predicted by contracting theories, ¨ with the exception of the study by Kaplan and Stromberg (2003). In this section, we compare the design of CEO contracts post-LBO to predictions from contracting theories. 4.1. Principal-agent theories ¨ In classical principal-agent models, e.g., Holmstrom (1979), the setup is that a principal wants to design a contract with an agent to maximize value. The problem is that the agent is effort-averse and effort is unobservable to the principal. As a result, the optimal contract involves compensation contingent on a set of signals which are observable and correlated with the agent’s effort.14 Informativeness principle: The ‘‘informativeness principle’’ implies that any signal that, at the margin, reveals information about an agent’s effort should be included in ¨ the contract (e.g., Harris and Raviv, 1979; Holmstrom, 1979; Innes, 1990). Consistent with such theories, we find that PE sponsors contract on several measures which correlate with the CEO’s effort, e.g., EBITDA for year-toyear performance and IRRs or multiples for longer-term performance. As Prendergast (1999, p. 14) points out: ‘‘The most common example of the use of [the informativeness principle] has been application of Relative Performance Evaluation, where the performance of one agent is compared to another when choosing compensation.’’15 From this perspective, a surprising characteristic of the contracts is the absence of 14 For a review of contracting theories relevant for our discussion, see Gibbons (1998) and Prendergast (1999). 15 For research on relative performance evaluation for CEOs, we refer to Gibbons and Murphy (1990).

indexing of a CEO’s performance relative to the performance of firms in the same industry. The strong owners do not filter out common risks, at least not explicitly in the contract.16 Our evidence suggests that the absence of relative performance evaluation may not necessarily be a result of weak governance, but that indexing in many cases is inefficient (e.g., Meulbroek, 2001; Garvey and Milbourn, 2003; Dittmann, Maug and Spalt, 2011). Precision of signals: One explanation for not using a large number of signals in the contracts is that the relevance of a particular signal is inversely related to its noise, so imprecise signals should receive a relatively lower weight in the contract. We find that the PE sponsors contract on more precise signals than what is commonly observed in public firms. After the arrival of a strong owner in the firm, the vast majority of contracts have conditions related to the achievement of cash flow and EBITDA, not qualitative and nonfinancial measures. Also, the PE sponsors prefer not to contract upon signals that can be easily manipulated (e.g., earnings management changing earnings per share (EPS)). In the longer run, value creation in the private equity setting can be measured more precisely (Was there a liquidity event for the sponsor? What was the multiple or IRR to the sponsor?) and the contracts do, as predicted by classical principal-agent theories, involve such signals. It is a particularity of the contracts we observe that almost all of them involve performance-vesting equity where the performance is measured at an exit event for the PE sponsor. However, multiples or IRRs cannot be measured without significant noise year to year because there is no public market for the firm’s equity, so accounting-based signals are used for shortterm performance evaluation and compensation.17 Screening and selection effects: Another implication of principal-agent models is that contracts result in selection effects. Principals may design CEO contracts not only to induce an agent’s effort, but also to affect which agents will select the contract. If agents differ in their skills (e.g., Rosen, 1981; Gabaix and Landier, 2008), performance-sensitive contracts will be preferred by better agents (e.g., Lazear, 1986). We find that the principals design CEO contracts consistent with screening models. Contracts become significantly more performance sensitive post-LBO, even for executives who are retained across the going-private transaction. Percentage equity ownership increases, vesting conditions become more restrictive, and CEOs are required to ‘‘buy-in’’ at the LBO, causing a separating equilibrium with endogenous self-selection of better CEOs. Screening through contract selection is a complement to detailed assessments of CEO

16 One caveat is that it is still possible that there is significant subjective relative performance evaluation in that the principal terminates a CEO who performs worse than other CEOs/firms in the same industry. Benchmarking must not be explicit and contractual. However, it is not completely clear from a theoretical perspective why relative performance evaluation should be subjective, rather than based on contracting. 17 A large body of work in the accounting literature examines the use of accounting- vs. stock performance-based measures in the context of standard agency models. For examples, see Core, Guay, and Verrecchia (2003), Ittner, Larcker, and Rajan (1997), or Lambert and Larcker (1987).

H. Cronqvist, R. Fahlenbrach / Journal of Financial Economics 108 (2013) 659–674

candidates, as studied by Kaplan, Klebanov and Sorensen (2012). More surprising from the perspective of screening models is that the average cash-out is significantly larger than the buy-in for CEOs who are retained. However, this is consistent with Kaplan and Stein (1993) who also find that senior executives, on average, invest less than half in post-LBO equity as they cashed out for their pre-buyout equity. The large cash-out may be compensation to the CEO of a public firm for not resisting the buyout by using poison pills or other takeover defense mechanisms.18 Sensitivity of compensation to signals: Classical principalagent theories imply that contracts should maximize the sensitivity of the agent’s compensation to the signals. We find that the strong owners design contracts such that the CEO’s compensation is sensitive to the signal. In particular, the performance-vesting feature of equity grants guarantees a large payoff to executives when the PE sponsor, at the exit event, can measure a precise signal (IRR; multiple). Yet, PE sponsors do not rely exclusively on bonuses and performance-vesting equity because CEOs have a base salary and a significant percentage of the equity does not performance-vest. Theory predicts several reasons for muted incentives. One is simply that the agent is risk averse, which reduces the optimal sensitivity. Multitasking and subjective performance evaluation: Another theoretical explanation for muted equity incentives is what is commonly referred to as ‘‘multitasking’’ in contracting theory. In multi-tasking models, an effort-averse agent will exert effort only on the tasks whose signals pay off the most. Clearly, it involves multiple tasks to manage an LBO firm. More subjective performance evaluation then becomes the equilibrium ¨ contract (e.g., Holmstrom and Milgrom, 1991; Baker, 1992). Time-vesting of equity incentives may be interpreted as subjective performance evaluation by the principal, because a portion of the agent’s pay is contingent on the principal’s decision not to dismiss the agent. Even PE sponsorswho use performance-vesting based on, e.g., IRRs, find it optimal to simultaneously use some time-vesting. This result appears consistent with Baker, Gibbons and Murphy’s (1994) model which implies that, in an optimal contract, objective measures and subjective performance evaluation are complements rather than substitutes. Recent principal-agent theories: In general, classical principal-agent models are concerned with the level and sensitivity of pay, rather than vesting periods, security mix, and restrictions on unwinding of equity incentives. More recent models focus on those aspects, and some of our results are consistent with predictions from these models. For example, Edmans, Gabaix, Sadzik and Sannikov (2012) predict that time-vesting is optimal, and that restricted stock is preferred to options (their optimal contract can be implemented completely with stock). Dittmann, Maug and Spalt (2011) examine 18 Such an explanation seems consistent with anecdotal evidence that CEO cash-outs are significantly smaller in private-to-private transactions, where concerns about CEOs using takeover defense mechanisms are not relevant.

669

indexing of CEO compensation, and calibrate that the average benefits are only 3% of the compensation costs because for about 80% of the CEOs in their sample, indexing destroys incentives as it reduces option deltas and increases the likelihood of poor outcomes; if the benchmark is the stock market index, the benefits from indexing decline even more. Their model can explain the absence of indexing in the CEO contracts we study.

4.2. Perquisite theories There are two main theoretical perspectives regarding CEO perquisites. One is ‘‘perks as private benefits,’’ i.e., perquisites are the result of principal-agent problems and are valuable only to the CEO, not the firm’s shareholders. If perks are a private benefit, then we would expect LBO firms to provide significantly fewer CEO perks (e.g., Jensen and Meckling, 1976; Grossman and Hart, 1980; Jensen, 1986). Another perspective is ‘‘perks as productivity.’’ Perks may improve the executive’s (and the firm’s) productivity. For example, the CEO may not internalize the value to the firm from the perk, perquisites may be cost effective, and there may be scale economies in providing a perk. We find that PE sponsors do not significantly redesign CEO contracts regarding perquisites, not even related to CEOs’ personal use of corporate aircraft or tax gross-ups, two perks with a particularly bad reputation in media, nor do they redesign the perk mix. In this respect, our results are consistent with the conclusion by Rajan and Wulf (2006, p 31): ‘‘We do not see a systematic pattern that is fully consistent with an [y] agency explanation of perks.’’

4.3. Severance pay theories Models of severance pay include Almazan and Suarez (2003) and Inderst and Mueller (2010). These models analyze moral hazard problems that arise when agents change their behavior because of the risk of being terminated by the principal. A severance pay contract is the optimal contract used to insure an agent ex ante against dismissal. One of the predictions of Almazan and Suarez’s (2003) model is that CEO severance contracts are more favorable to a CEO when governance is more lax and CEOs may entrench themselves. Such contracts provide incentives to a CEO to resign when it is optimal to do so. We find that the PE sponsors redesign CEO contracts consistent with such a model. The contractual provisions for termination without cause prior to the LBO are more favorable than those designed by a private equity sponsor. In Inderst and Mueller’s (2010) model, the CEO privately observes an interim signal about his match quality with the firm and about the expected value of the firm. Severance pay provides incentives for poor CEOs to resign, but makes it costly to induce CEO effort. A model prediction is that an increase in equity incentives is accompanied by a simultaneous increase in severance pay, but our data do not seem consistent with such a prediction.

670

H. Cronqvist, R. Fahlenbrach / Journal of Financial Economics 108 (2013) 659–674

5. Can changes in leverage explain the changes in CEO contracts? While several central aspects of a firm’s contracting environment (e.g., the industry in which a firm operates) do generally not change with an LBO, one aspect that changes substantially is the firm’s capital structure. Panel A of Table 9 shows changes in book leverage, defined as total long-term debt (including current portion of longterm debt) divided by total assets, from the last 10-K prior to the LBO to the first 10-K post-LBO. The average (median) change in book leverage is 0.411 (0.392) in our sample, and book leverage is about 60% after the LBO. We now examine whether any firm with a similar significant increase in leverage carries out similar CEO contract changes. If it did, it would suggest that changes in capital structure and not the PE sponsor as a strong principal were responsible for changes in CEO contracts. To examine this question we analyze changes in CEO contracts in publicly listed firms across years with large increases in leverage, following two approaches. The first approach involves matching each firm in our sample by fiscal year and industry (based on the 48 Fama and French industries) to large Compustat firms with total assets of more than $500 million and to retain a public firm as control if: DBook leverage sample LBO firm  Dbook leverage public firm r25% DBook leverage sample LBO firm

and Post-LBO leverage sample firm  leverage public firm r 25%: Post-LBO leverage sample firm

We find matches for five of our 20 sample LBOs. One match is a voluntary leveraged recapitalization and three matched firms undertake large debt-financed acquisitions. Panel B of Table 9 shows the changes in book leverage for these firms. The median change is 0.336 and comparable to the median change in leverage for our LBO firms. Are there comparable CEO contract changes in these matching firms? We focus on three changes: EBITDA-based measures for accounting bonuses instead of EPS-based measures, performance-vesting equity, and restrictions on the resale of vested equity. We find that only one firm redesigns CEO contracts away from earnings-based to cash flow-based measures. Two of the firms use performance-vesting equity. One of them explicitly states that the vesting is based on synergies from the debt-financed acquisition, and the leveraged recap firm uses restricted stock units that vest based on performance targets (1/3 revenue-, 1/3 EBITDA-, and 1/3 free cash flowbased). Only one matching firm has any restrictions on the sale of equity by the CEO, in which case the CEO can only sell a maximum of 10% of his holdings each year. One firm has adopted ‘‘a [non-binding] guideline encouraging senior executives to retain at least 50% of their vested equity in the firm.’’ The other firms have no restrictions or very low (five times base salary) minimum ownership requirements. The second approach involves compiling a sample of large firms that underwent voluntary leveraged recapitalizations or leveraged share buybacks in 2006 and 2007. We find seven such firms. Panel C of Table 9 shows that the median firm increases book leverage by 0.438, to a leverage ratio of

Table 9 Changes in book leverage for LBOs versus voluntary capital structure changes. The table shows annual levels and changes in book leverage, defined as book value of long-term debt (including current maturities) divided by total assets. Panel A shows levels and changes for a sample of leveraged buyouts (LBOs). The sample consists of 20 large leveraged buyouts of publicly listed U.S. firms between 2005 and 2007. Panel B shows levels and changes in book value for a sample of five publicly listed firms that operate in the same industry as the sample LBOs and had similar changes in book leverage and leverage levels in the same year. Panel C shows levels and changes in book value for a sample of seven leveraged recapitalizations and leveraged stock repurchases with significant changes in leverage in 2006 and 2007. Mean

Median

0.183 0.594 0.411

0.149 0.536 0.392

0.196 0.553 0.358

0.195 0.561 0.336

0.467 1.027 0.560

0.222 0.715 0.438

Panel A: (N ¼ 20) Book leverage pre-LBO Book leverage post-LBO Change in book leverage across transaction Panel B: (N ¼5) Book leverage pre-large change Book leverage post-large change Change in book leverage Panel C: (N ¼7) Book leverage pre-recap/ levered buyback Book leverage post-recap/ levered buyback Change in book leverage across transaction

more than 70%. We find that only two firms redesign CEO contracts away from earnings-based measures. In addition to the firm in our other matched sample, one additional firm uses performance-vesting (the vesting of 50% of the equity is reduced from five to 2.5 years if the stock price increases by 50%), and for another firm, the restricted stock units vest conditional on 8% annual return or a return equivalent to at least the S&P 500. No firm has large ownership requirements or restrictions on the sale of vested equity. In conclusion, boards of public firms that undertake capital structure changes comparable in magnitude to those of LBOsponsored firms do not simultaneously change CEO contracts as dramatically as our sample firms. Therefore, changes in leverage alone cannot explain the changes in CEO contracts that we have reported and discussed previously in the paper.

6. Relation to discussion of executive compensation reform In this section, we discuss how the CEO compensation practices designed by private equity sponsors compare to some of the most criticized compensation practices in U.S. public firms.19 19 The basis for such criticism is Bebchuk and Fried (2004), who identify compensation practices that appear inconsistent with incentivecompatible contracts. For other evidence that CEO contracts are not always aligned with shareholder interests, see, e.g., Bertrand and Mullainathan (2001), Heron and Lie (2007), and Bebchuk, Grinstein, and Peyer (2010). In contrast, other economists have concluded that CEO ¨ compensation contracts overall appear efficient (e.g., Holmstrom and Kaplan, 2003; Murphy and Zabojnik, 2004; Gabaix and Landier, 2008; Edmans and Gabaix, 2009).

H. Cronqvist, R. Fahlenbrach / Journal of Financial Economics 108 (2013) 659–674

6.1. Pay for luck ‘‘Pay for luck’’ comes from equity that compensates CEOs for increases in the firm’s stock price even if those increases are the result of industry/market factors beyond the CEO’s control. Importantly, we find no explicit indexing in CEO contracts designed by private equity sponsors. This seems to contradict the notion that the absence of CEO compensation indexing is the result of the absence of a strong principal (e.g., Bertrand and Mullainathan, 2001). Johnson and Tian (2000a, b) examine indexed options, and discuss practical implementation issues. Also, Meulbroek (2001) argues that as the incentive alignment increases because of filtering out industry/market risks, the CEO becomes increasingly exposed to firm-specific risk, and values indexed options at only a fraction of their market value. Our finding is consistent with models such as Dittmann, Maug and Spalt (2011) who show that the benefits from indexing are small. If pay for luck was generally inefficient, we would have expected at least some firms with sophisticated and financially savvy principals to use indexed CEO compensation contracts. But we do not find any evidence that these principals use such contracts. Our evidence suggests that the absence of indexing is the result of insignificant benefits to explicit indexing in the first place. 6.2. Uniform use of at-the-money options Bebchuk and Fried (2004) show that 95% of options to executives in public U.S. firms are granted at-the-money, and they ask why not more CEOs are granted premium options, with a strike price above the current stock price. We find that premium options are not a common feature of CEO contracts designed by strong principals. 6.3. Performance-vesting versus time-vesting Critics have raised concerns that equity in U.S. public firms does not always vest based on performance. While the percentage of firms using performance-vesting has increased in recent years, it is still small (e.g., Bettis, Bizjak, Coles and Kalpathy (2010)). We find that the PE sponsors use more performance-vesting than do boards of public firms, but they do not rely completely on performance-vesting. 6.4. Unrestricted freedom to unwind equity incentives Public firms take few steps to restrict a CEO’s freedom to unwind his equity incentives. For example, they generally do not have restrictions on the resale of shares from exercise, except a small minimum stock ownership threshold (e.g., Core and Larcker, 2002). Indeed, Ofek and Yermack (2000) find that executives sell shares when additional equity is granted. In contrast, the strong principals in our sample contractually restrict the resale market for vested shares for CEOs. Even though the PE sponsored firm is private, which restricts resales per se, the PE sponsors further restrict the resale market by a right of first refusal and by limiting the set of parties to which executives can sell. These restrictions are effective at preventing an executive from unwinding incentives. It may not be possible to implement a right of first refusal in publicly listed firms, but implementing

671

restrictions on how much equity can be sold prior to retirement/ turnover of the CEO are implementable. 6.5. Another strong principal: executive compensation in TARP firms In 2009, the U.S. Treasury Department issued standards for CEO contracts in firms receiving Troubled Asset Relief Program (TARP) funding. Among them are that salaries above $500,000 should be paid as vested stock, bonuses should be restricted to 1/3 of total compensation and should be paid in restricted stock, clawbacks are imposed on bonuses, perquisites are restricted and tax gross-ups are prohibited, and severance and change in control payments are prohibited (e.g., Core and Guay, 2010).20 The CEO contract design choices of the U.S. government are very different from those of sophisticated PE sponsors, although both apply to firms with substantial leverage. The TARP standards for CEO contracts may of course have been designed to protect all stakeholders including the government, and not to uniquely align the incentives of CEOs and shareholders. Still, Mr. Feinberg, the Special Master for TARP Executive Compensation (‘‘Pay Czar’’) has encouraged all public firms to look at the standards. Regulators and proponents of CEO pay reform should be aware that the market-based CEO contracts we have studied differ markedly from those designed by the U.S. government as another strong principal. 7. Conclusion We have contributed a new perspective on executive compensation research by examining changes to CEO contracts carried out when a firm transitions from having dispersed shareholders to having private equity sponsors, as ‘‘strong principals.’’ Most existing work on incentives in the LBO setting has focused on executives’ percentage equity ownership, but we find that a much broader set of features of CEO contracts change. While we confirm some results from studies from the early 1980s LBO wave, several other results have not been previously reported. We find that CEO contracts become more performance-sensitive. About 50% of equity grants to CEOs are performance-vesting, with performance thresholds that are typically only attained at a reverse LBO or sale of the firm. We also find that ex ante severance contracts are stricter with respect to unvested equity, which is often forfeited. CEOs are generally prevented from unwinding equity if they do not lead the firm to success. PE sponsors also redesign contracts away from qualitative, nonfinancial and earnings-based measures, to cash flow-based measures. Perhaps to compensate a risk-averse CEO for these additional contractual provisions and for a potential increased risk of turnover while working for a highly levered firm and a strong principal with board control, we find that base salaries and bonuses increase. We also find that PE sponsors do not use premium options or 20 Dittmann, Maug, and Zhang (2011) analyze several proposals to restrict CEO compensation and find that many restrictions would have unintended consequences.

672

H. Cronqvist, R. Fahlenbrach / Journal of Financial Economics 108 (2013) 659–674

indexed options, do not condition vesting on relative industry performance, and do use time-vesting equity. When comparing the actual CEO contracts in our sample to those predicted by contracting theories, we find that several principal-agent, perquisite, and severance pay theories do well in describing how the CEO contract changes to take the goals of the private equity sponsor into account. Appendix A. Sample of private equity transactions Data are from Bloomberg. Announced total value is calculated by Bloomberg as offer price multiplied with shares outstanding plus net debt. Firm

HCA Inc First Data Corp Alltel Corp Freescale Semiconductor Inc Univision Communications Inc Biomet Inc SunGard Data Systems Inc ARAMARK Corp Toys R US Inc Dollar General Corp Neiman Marcus Group Inc West Corporation Education Management Corp Aleris International Inc United Surgical Partners HealthMarkets Inc Yankee Candle Co Inc Linens ‘n Things Inc SS&C Technologies Inc Metals USA Inc

Announcement date

Announced total value ($ million)

07/24/2006 04/02/2007 05/21/2007 09/15/2006 06/27/2006 12/18/2006 03/28/2005 05/01/2006 03/17/2005 03/12/2007 05/02/2005 05/31/2006 03/06/2006 08/08/2006 01/08/2007 09/15/2005 10/25/2006 11/08/2005 07/28/2005 05/18/2005

32,193.46 27,497.13 27,149.00 16,222.22 12,605.50 11,426.96 10,591.52 7,998.57 7,544.39 7,321.03 5,044.34 3,713.71 3,123.64 2,301.62 1,781.41 1,712.55 1,640.08 1,274.82 800.54 710.32

Appendix B. Examples of equity incentive plans HCA Inc.: On November 17, 2006, the Board of Directors approved and adopted the 2006 Stock Incentive Plan for Key Employees of HCA Inc. and its Affiliates (the ‘‘2006 Plan’’). The purpose of the 2006 Plan is to promote our long term financial interests and growth by attracting and retaining management and other personnel and key service providers with the training, experience and ability to enable them to make a substantial contribution to the success of our business; to motivate management personnel by means of growth-related incentives to achieve long range goals; and to further the alignment of interests of participants with those of our shareholders through opportunities for increased stock or stock-based ownership in the Company. In January 2007, the Committee approved grants to Messrs. Bovender, Bracken, Johnson, Hazen and Rutledge of options to purchase 399,604 shares, 349,654 shares, 249,753 shares, 159,841 shares and 139,861 shares, respectively, of our common stock. The options are divided so that 1/3 are time vested options, 1/3 are EBITDA-based performance vested options and 1/3 are performance options that vest based on investment return to the Sponsors, each as described below.

The time vested options vest and become exercisable in equal increments of 20% on each of the first five anniversaries of the date of grant. The time vested options have a strike price equivalent to fair market value on the date of grant (as determined reasonably and in good faith by the Board of Directors after consultation with the Chief Executive Officer). The EBITDA-based performance vested options are eligible to vest and become exercisable in equal increments of 20% at the end of fiscal years 2007, 2008, 2009, 2010 and 2011, but will vest on those dates only if we achieve certain annual EBITDA performance targets, as determined in good faith by the Board (in consultation with the CEO). The EBITDA-based performance vested options also vest and become exercisable on a ‘‘catch up’’ basis, if at the end of fiscal years 2008, 2009, 2010 or 2011, the cumulative total EBITDA earned in all prior completed fiscal years or the 2012 fiscal year exceeds the cumulative total of all EBITDA targets in effect for such years. Similar to 2006 performance-based awards, we do not intend to publicly disclose the specific EBITDA performance targets for these options. However, we intend to set these targets at levels designed to be generally consistent with the level of difficulty of achievement associated with prior year performance-based awards. The options that vest based on investment return to the Sponsors are eligible to vest and become exercisable with respect to 10% of the common stock subject to such options on each of the first five anniversaries of the closing date of the Merger if the Investor Return (as defined below) is at least equal to two times the price paid to shareholders in the Merger (or $102.00), and with respect to an additional 10% on each of the first five anniversaries of the closing date if the Investor Return is at least equal to two-and-a-half times the price paid to shareholders in the Merger (or $127.50). ‘‘Investor Return’’ means, on any of the first five anniversaries of the closing date of the Merger, or any date thereafter, all cash proceeds actually received by affiliates of the Sponsors after the closing date in respect of their common stock, including the receipt of any cash dividends or other cash distributions (but including the fair market value of any distribution of common stock by the Sponsors to their limited partners), determined on a fully diluted, per share basis. The Sponsor investment return options also may become vested and exercisable on a ‘‘catch up’’ basis if the relevant Investor Return is achieved at any time occurring prior to the expiration of such options. The combination of time, performance and investor return based vesting of these awards is designed to compensate executives for long term commitment to the Company, while motivating sustained increases in our financial performance and helping ensure the Sponsors have received an appropriate return on their invested capital.

Alltel Corporation Upon closing of the Merger, the post-Merger Board of Directors replaced Alltel’s existing equity incentive plans with the 2007 Stock Option Plan. This plan provides for

H. Cronqvist, R. Fahlenbrach / Journal of Financial Economics 108 (2013) 659–674

the award of stock options with respect to up to 6.5% of our stock on a fully diluted basis. Following the Merger and the approval of the 2007 Stock Option Plan, Alltel granted time-based and performance-based options to certain executive officers, including the Named Executive Officers, other than Mr. Beebe. Approximately 69% of the equity awards granted to each executive vest based on employment continually through the vesting period by the optionee, and approximately 31% vest through the attainment of company-based performance goals. The time-based options vest over a 5-year period (vesting 20% on each anniversary of the grant date) and are intended to enable greater leverage in retaining seasoned executives critical to the future success of the Company. The performance-based options seek to align the efforts and interests of the executives with those of the stockholders by rewarding executives if the value of the company measured by investment returns achieved by stockholders increases by certain threshold amounts. The event in which such a return is achieved occurs when control is transferred to new owners. By requiring that options be held until such a transaction occurs, the options also encourage retention of critical executives until stockholders are able to realize these returns. Each performance-based option vests and becomes exercisable (i) upon the Sponsors attaining a multiple of their equity investment calculated as cash or liquid securities received, divided by the purchase price and (ii) subject to the optionee’s employment on the date the performance condition is met. Of the Performance Options granted to each Named Executive Officer in 2007, one-half require a return multiple of at least 1.5 and one-half require a return multiple of at least 2.0. References Almazan, A., Suarez, J., 2003. Entrenchment and severance pay in optimal governance structures. Journal of Finance 58, 519–547. Andrade, G., Kaplan, S.N., 1998. How costly is financial (not economic) distress? Evidence from highly leveraged transactions that became distressed. Journal of Finance 53, 1443–1493. Baker, G.P., 1992. Incentive contracts and performance measurement. Journal of Political Economy 100, 598–614. Baker, G.P., Gibbons, R., Murphy, K.J., 1994. Subjective performance measures in optimal incentive contracts. Quarterly Journal of Economics 109, 1125–1156. Baker, G.P., Wruck, K.H., 1989. Organizational changes and value creation in leveraged buyouts: the case of the O.M. Scott & Sons Company. Journal of Financial Economics 25, 163–190. Bebchuk, L.A., Fried, J.M., 2004. Pay Without Performance: The Unfulfilled Promise of Executive Compensation, Boston, MAHarvard University Press. Bebchuk, L.A., Grinstein, Y., Peyer, U., 2010. Lucky CEOs and lucky directors. Journal of Finance 65, 2363–2401. Becht, M., Franks, J., Mayer, C., Rossi, S., 2009. Returns to shareholder activism: evidence from a clinical study of the Hermes UK Focus Fund. Review of Financial Studies 22, 3093–3129. Becker, B., Cronqvist, H., Fahlenbrach, R., 2011. Estimating the effects of large shareholders using a geographic instrument. Journal of Financial and Quantitative Analysis 46, 907–942. Bengtsson, O., Hand, J.R.M., 2010. CEO compensation in venture-backed firms. Journal of Business Venturing 26, 391–411. Bertrand, M., Mullainathan, S., 2001. Are CEOs rewarded for luck? The ones without principals are. Quarterly Journal of Economics 116, 901–932.

673

Bettis, J.C., Bizjak, J.M., Coles, J.L., Kalpathy, S.L., 2010. Stock and option grants with performance-based vesting provisions. Review of Financial Studies 23, 3849–3888. Chung, J., Sensoy, B.A., Stern, L.H., Weisbach, M.S., 2012. Pay for performance from future fund flows: the case of private equity. Review of Financial Studies 25, 3259–3304. Cole, R.A., Mehran, H., 2008. What can we learn from privately held firms about executive compensation? Federal Reserve Bank of New York Staff Report No, 314. Core, J.E., Guay, W.R., 2010. Is there a case for regulating executive pay in the financial services industry? Unpublished working paper. Massachusetts Institute of Technology and University of Pennsylvania. Core, J.E., Guay, W.R., Verrecchia, R.E., 2003. Price versus non-price performance measures in optimal CEO compensation contracts. Accounting Review 78, 957–981. Core, J.E., Larcker, D.F., 2002. Performance consequences of mandatory increases in executive stock ownership. Journal of Financial Economics 64, 317–340. Cotter, J.F., Peck, S.W., 2001. The structure of debt and active equity investors: the case of the buyout specialist. Journal of Financial Economics 59, 101–147. Cronqvist, H., Fahlenbrach, R., 2009. Large shareholders and corporate policies. Review of Financial Studies 22, 3941–3976. DeAngelo, H., DeAngelo, L., 2000. Controlling stockholders and the disciplinary role of corporate governance payout policy: a study of the Times Mirror Company. Journal of Financial Economics 56, 153–207. Denis, D.J., 1994. Organizational form and the consequences of highly leveraged transactions: Kroger’s recapitalization and Safeway’s LBO. Journal of Financial Economics 36, 193–224. Denis, D.J., Denis, D., 1993. Managerial discretion, organizational structure, and corporate performance: a study of leveraged recapitalizations. Journal of Accounting and Economics 16, 209–236. Dittmann, I., Maug, E., Spalt, O., 2011. Indexing executive compensation contracts. Unpublished working paper. University of Mannheim, Erasmus University, and Tilburg University. Dittmann, I., Maug, E., Zhang, D., 2011. Restricting CEO pay. Journal of Corporate Finance 17, 1200–1220. Edgerton, J., 2012. Agency problems in public firms: evidence from corporate jets in leveraged buyouts. Journal of Finance 67, 2187–2213. Edmans, A., Gabaix, X., 2009. Is CEO pay really inefficient? A survey of new optimal contracting theories. European Financial Management 15, 486–496. Edmans, A., Gabaix, X., Sadzik, T., Sannikov, Y., 2012. Dynamic CEO compensation. Journal of Finance 67, 1603–1647. Fich, E.M., Tran, A.L., Walkling, R.A., 2012. On the importance of golden parachutes. Journal of Financial and Quantitative Analysis Forthcoming. Gabaix, X., Landier, A., 2008. Why has CEO pay increased so much? Quarterly Journal of Economics 123, 49–100. Gao, H., Lemmon, M.L., Li, K., 2012. Are CEOs in the U.S. public firms overpaid? New evidence from private firms. Unpublished working paper. Nanyang Technological University, University of Utah, and University of British Columbia. Garvey, G., Milbourn, T., 2003. Incentive compensation when executives can hedge the market: evidence of relative performance evaluation in the cross-section. Journal of Finance 58, 1557–1581. Gertner, R., Kaplan, S.N., 1996. The value-maximizing board. Unpublished working paper. University of Chicago. Gibbons, R., 1998. Incentives in organizations. Journal of Economic Perspectives 12, 115–132. Gibbons, R., Murphy, K.J., 1990. Relative performance evaluation for chief executive officers. Industrial and Labor Relations Review 43, 30–51. Gillan, S.L., Hartzell, J., Parrino, R., 2009. Explicit versus implicit contracts: evidence from CEO employment agreements. Journal of Finance 64, 1629–1655. Grossman, S.J., Hart, O.D., 1980. Takeover bids, the free-rider problem and the theory of the corporation. Bell Journal of Economics 11, 42–64. Guo, S., Hotchkiss, E.S., Song, W., 2011. Do buyouts (still) create value? Journal of Finance 66, 479–517. Harris, M., Raviv, A., 1979. Optimal incentive contracts with imperfect information. Journal of Economic Theory 20, 231–259. Helwege, J., Packer, F., 2009. Private matters. Journal of Financial Intermediation 18, 362–383. Heron, R.A., Lie, E., 2007. Does backdating explain the stock price pattern around executive stock option grants? Journal of Financial Economics 83, 271–295.

674

H. Cronqvist, R. Fahlenbrach / Journal of Financial Economics 108 (2013) 659–674

Holderness, C.G., Sheehan, D.P., 1985. Raiders or saviors? The evidence on six controversial investors. Journal of Financial Economics 14, 555–579. Holderness, C.G., Sheehan, D.P., 1991. Monitoring an owner: the case of Turner Broadcasting. Journal of Financial Economics 30, 325–346. ¨ Holmstrom, B., 1979. Moral hazard and observability. Bell Journal of Economics 10, 74–91. ¨ Holmstrom, B., Kaplan, S.N., 2003. The state of U.S. corporate governance: what’s right and what’s wrong? Journal of Applied Corporate Finance 15, 8–20. ¨ Holmstrom, B., Milgrom, P., 1991. Multitask principal-agent analyses: incentive contracts, asset ownership, and job design. Journal of Law, Economics, and Organization 7, 24–92. Inderst, R., Mueller, H.M., 2010. CEO replacement under private information. Review of Financial Studies 23, 2935–2969. Innes, R., 1990. Limited liability and incentive contracting with ex-ante action choices. Journal of Economic Theory 52, 45–67. Ittner, C.D., Larcker, D.F., Rajan, M.V., 1997. The choice of performance measures in annual bonus contracts. Accounting Review 73, 231–255. Jensen, M.C., 1986. Agency costs of free cash flow, corporate finance, and takeovers. American Economic Review 76, 323–329. Jensen, M.C., 1989. The eclipse of the public corporation. Harvard Business Review 67, 61–74. Jensen, M.C., Meckling, W., 1976. Theory of the firm: managerial behavior, agency costs and ownership structure. Journal of Financial Economics 3, 305–360. Johnson, S.A., Tian, Y.S., 2000a. Indexed executive stock options. Journal of Financial Economics 57, 35–64. Johnson, S.A., Tian, Y.S., 2000b. The value and incentive effects of nontraditional executive stock option plans. Journal of Financial Economics 57, 3–34. Kaplan, S.N., 1989. The effects of management buyouts on operating performance and value. Journal of Financial Economics 24, 217–254. Kaplan, S.N., Klebanov, M.M., Sorensen, M., 2012. Which CEO characteristics and abilities matter? Journal of Finance 67, 973–1007. Kaplan, S.N., Stein, J.C., 1993. The evolution of buyout pricing and financial structure in the 1980s. Quarterly Journal of Economics 108, 313–357. ¨ Kaplan, S.N., Stromberg, P., 2003. Financial contracting theory meets the real world: an empirical analysis of venture capital contracts. Review of Economic Studies 70, 281–315. ¨ Kaplan, S.N., Stromberg, P., 2009. Leveraged buyouts and private equity. Journal of Economic Perspectives 23, 121–146. Katz, S.P., 2009. Earnings quality and ownership structure: the role of private equity sponsors. Accounting Review 84, 623–658. Kole, S.R., 1997. The complexity of compensation contracts. Journal of Financial Economics 43, 79–104.

Lambert, R.A., Larcker, D.F., 1987. An analysis of the use of accounting and market measures of performance in executive compensation contracts. Journal of Accounting Research 25, 85–125. Lazear, E., 1986. Salaries and piece rates. Journal of Business 59, 405–431. ¨ Lerner, J., Sorensen, M., Stromberg, P., 2011. Private equity and long-run investment: the case of innovation. Journal of Finance 66, 445–477. Leslie, P., Oyer, P., 2009. Managerial incentives and value creation: evidence from private equity. Unpublished working paper. Stanford University. Maug, E., 1997. Boards of directors and capital structure: alternative forms of corporate restructuring. Journal of Corporate Finance 3, 113–139. Metrick, A., Yasuda, A., 2010. The economics of private equity funds. Review of Financial Studies 23, 2303–2341. Meulbroek, L.K., 2001. The efficiency of equity-linked compensation: understanding the full cost of awarding executive stock options. Financial Management 30, 5–44. Meulbroek, L.K., 2002. Compensation at Level 3 Communications. Harvard Business School Case # 9-202-084. Murphy, K.J., Zabojnik, J., 2004. CEO pay and appointments: a marketbased explanation for recent trends. American Economic Review 94, 192–196. Muscarella, C.J., Vetsuypens, M.R., 1990. Efficiency and organizational structure: a study of reverse LBOs. Journal of Finance 45, 1389–1413. Ofek, E., Yermack, D.L., 2000. Taking stock: equity-based compensation and the evolution of managerial ownership. Journal of Finance 55, 1367–1384. Officer, M.S., Ozbas, O., Sensoy, B.A., 2010. Club deals in leveraged buyouts. Journal of Financial Economics 98, 214–240. Prendergast, C., 1999. The provision of incentives in firms. Journal of Economic Literature 37, 7–63. Rajan, R.G., Wulf, J., 2006. Are perks purely managerial excess? Journal of Financial Economics 79, 1–33. Rosen, S., 1981. The economics of superstars. American Economic Review 71, 845–858. Rusticus, T.O., 2006. Executive severance agreements. Unpublished working paper. Northwestern University. Wruck, K.H., 1994. Financial policy, internal control, and performance: Sealed Air Corporation’s leveraged special dividend. Journal of Financial Economics 36, 157–192. Yermack, D.L., 2006a. Flights of fancy: corporate jets, CEO perquisites, and inferior shareholder returns. Journal of Financial Economics 80, 211–242. Yermack, D.L., 2006b. Golden handshakes: separation pay for retired and dismissed CEOs. Journal of Accounting and Economics 41, 237–256.