Evolving executive equity compensation and the limits of optimal contracting

Vanderbilt Law Review

Volumn 64, Issue 2, 2011, Pages 611-674

  Walker, David I ^a  

^a Department of Law ^*  (United States)

Author keywords

[No Author keywords available]

Indexed keywords

EID: 79959527120     PISSN: 00422533     EISSN: None     Source Type: Journal    
DOI: None     Document Type: Article

References (249)

1. See infra Part II. B.
2. See infra Part II. B.
3. See infra Part II. C.
4. See, e.g., Judith F. Samuelson & Lynn A. Stout, Are Executives Paid Too Much?, WALL, ST. J., Feb. 25, 2009, at A13 (arguing that the widespread use of stock options in executive compensation encouraged executives to focus on shorter-term goals and take greater risks).
5. See, e.g., Sanjai Bhagat & Roberta Romano, Reforming Executive Compensation: Focusing and Committing to the Long-Term, 26 YALE J. ON REG. 359, 361 (2009) (suggesting that executives not be allowed to dispose of equity compensation prior to retirement);
6. Richard A. Posner, Are American CEOs Overpaid, and, if so, What if Anything Should Be Done About It?, 58 DUKE L. J. 1013, 1045-46 (2009) (suggesting that restricted stock should constitute a minimum fraction of CEO pay);
7. Susanne Craig, Cuomo, Frank Seek to Link Executive Pay, Performance, WALL ST. J., Mar. 13, 2009, at C1 (relating comments of House Financial Services Committee Chairman Barney Frank advocating broader application of rules tying executive pay to long-term performance).
8. See infra Part I. C. I.
9. Of course, whether the playing field remains level and for how long depends on the outcome of efforts to increase regulation of executive pay noted above.
10. See infra text accompanying notes 73-78.
11. See infra Part II. C.l.
12. See infra Part II. C.2.
13. Proxy statements provide various rationales for the use of stock, stock options, or both, but rarely invoke transaction costs.
14. See infra Part II. C.2.b.ii.
15. See infra Part II. C.2.b.iii.
16. See infra Part II. D.2.a. The size of the senior executive team varies firm by firm, but companies are required to report compensation for five executives.
17. See infra Part II. D.3.
18. See Michael C. Jensen & William H. Meckling, Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure, 3 J. PIN. ECON. 305 (1976). In brief, the Jensen and Meckling ("J&M") model begins with a sole proprietor-manager who sells shares of equity to nonmanaging outsiders, which creates a wedge between the manager's private incentives and the incentives of the shareholders generally. The outside shareholders cannot perfectly (or costlessly) observe the manager's effort or focus, and performance results are not completely within the manager's control. Thus, the shareholders cannot ensure perfect fidelity to their objectives, and the manager, who now owns less than 100 percent of the cash flow rights, will tend to consume excessive perks, loaf, and otherwise extract private benefits, since he enjoys 100 percent of the benefit of such activities, but only a fraction of the cost, which is borne pro rata by all shareholders. The optimal contract in this situation would minimize agency costs, which J&M defined as the sum of 1) monitoring costs incurred by the principal, 2) bonding costs incurred by the manager-agent to better ensure loyalty to shareholder wealth maximization, and 3) the cost of the residual divergence between the manager's actual decisions and shareholder wealthmaximizing decisions.
19. Going private transactions are an extreme way of reducing agency costs.
20. The effect is not always salutary. Given their human capital investment in their firms, executives tend to be more risk-averse than shareholders. Compensating executives with equity can increase executive risk aversion.
21. See infra note 24 and accompanying text.
22. The strike price of employee stock options is almost always a fixed price specified at grant, and almost always equal to the fair market value of the stock at grant. A few firms have experimented with indexing strike prices to a basket of competing stocks or to a broad measure of the stock market, such as the S&P 500, with the idea of focusing the option payout on firmspecific performance rather than market movements generally.
23. See Alfred Rappaport, New Thinking on How to Link Executive Pay with Performance, HARV. BUS. REV. Mar.-Apr. 1999, at 91, 101.
24. When graphed, a convex relationship presents a U-shaped curve. The relationship between option value and the price of the underlying shares tracks the right half of the U.
25. For example, an option delta of.75 means that when the price of the underlying shares changes by a small amount, the value of the option changes by seventy-five percent of that amount. See JOHN C. HULL, OPTIONS, FUTURES, AND OTHER DERIVATIVES 251 (6th ed. 2006) (explaining the concept of the option delta).
26. Per dollar of compensation expense, the option would have a delta that was 1.9 times (.75A4) the delta of the stock.
27. While option delta describes the sensitivity of the instrument to small changes in the underlying share price, the degree to which that sensitivity changes as the stock price changes (the second derivative of the value function) provides a measure of convexity, which is generally designated as option gamma.
28. Hull, supra note 20, at 373, 791.
29. All else being equal, executives and other employees whose financial and human capital generally is overinvested in their companies tend to disfavor risky projects relative to diversified shareholders.
30. See Brian J. Hall, Six Challenges in Designing Equity-Based Pay, 15 J. APPLIED CORP. FIN. 21, 29 (2003). In the wake of the recent financial crisis, regulators are concerned that incentive pay packages may have encouraged executives to take on too much risk, although the link is far from clear. The more traditional concern, however, has been a tendency towards conservatism.
31. See Brian J. Hall & Kevin J. Murphy, Stock Options for Undiversified Executives, 33 J. ACCT. & ECON. 3, 5 (2002) (explaining that "[r]estricting the trading and hedging activities of option recipients" causes executives receiving the options to "value the options below their cost to shareholders");
32. see also John E. Core et al., Executive Equity Compensation and Incentives: A Survey, ECON. POL'Y REV., Apr. 2003, at 27, 30 (noting that equity compensation is risky because stock prices are a noisy measure of firm performance and that recipients must be compensated for taking on the non-diversifiable risk).
33. See John Core & Jun. Qian, Option-Like Contracts for Innovation and Production 2 (Jan. 2000) (unpublished manuscript), available at http://papers.ssrn. com/sol3/papers.cfm? abstract-id=207968. In practice, more convex pay contracts means more option-heavy pay packages.
34. See, e.g., Chongwoo Choe, Leverage, Volatility and Executive Stock Options, 9 J. CORP. FIN. 591, 593 (2003) (hereinafter Choe, Leverage];
35. Chongwoo Choe, Maturity and Exercise Price of Executive Stock Options, 10 REV. FIN. ECON. 227, 229 (2001).
36. See, e.g., Choe, Leverage, supra note 27, at 593. Although increased volatility increases the value of an option, the value of an option at grant reflects expected volatility. Because executives cannot diversify away option risk, as firm risk increases, the cost to executives of added convexity increases.
37. See Richard A. Lambert & David F. Larcker, Stock Options, Restricted Stock, and Incentives 3 (Apr. 2004) (unpublished manuscript), available at http://papers.ssrn. com/sol3/papers.cfm?abstract-id=527822.
38. See, e.g., id.
39. Studies examining risk aversion alone or in combination with other factors include Ingolf Dittmann & Ernst Maug, Lower Salaries and No Options? On the Optimal Structure of Executive Pay, 62 J. FIN. 303, 308 (2007);
40. Gerald A. Feltham & Martin G. H. Wu, Incentive Efficiency of Stock Versus Options, 6 REV. ACCT. STUD. 7 (2001);
41. Brian J. Hall & Kevin J. Murphy, Optimal Exercise Prices for Executive Stock Options, 90 AM. ECON. REV. 209, 209 (2000);
42. Hall & Murphy, supra note 25, at 7;
43. Yisong S. Tian, Too Much of a Good Incentive? The Case of Executive Stock Options, 28 J. BANKING & FIN. 1225, 1226 (2004);
44. @ Lambert & Larcker, supra note 29, at 23;
45. Yisong S. Tian, Optimal Contracting, Incentive Effects and the Valuation of Executive Stock Options 4 (Apr. 30, 2001) (unpublished manuscript) [hereinafter Tian, Contracting], available at http://papers.ssrn. com/sol3/papers.cfm?abstract-id=268738.
46. Other characteristics that have been modeled include loss aversion, effort aversion, overall wealth, firm equity held, and outside investment opportunities.
47. See, e.g., Anna Dodonova & Yuri Khoroshilov, Optimal Incentive Contracts for Loss-Averse Managers: Stock Options Versus Restricted Stock Grants, 41 FIN. REV. 451, 452 (2006) (loss aversion);
48. Oded Palmon et al., Optimal Strike Prices of Stock Options for Effort-Averse Executives, 32 J. BANKING & FIN. 229, 230 (2008) (effort aversion);
49. Dittmann & Maug, supra at 308 (effort aversion);
50. Feltham & Wu, supra at 7 (effort aversion);
51. Tian, Contracting, supra at 40 (effort aversion, overall wealth, firm equity held, and outside investment opportunities).
52. See, e.g., Tian, Contracting, supra note 31, at 32;
53. see also Hall, supra note 24, at 31 (noting that under plausible assumptions, the "value-to-cost discount for stock is two to three times less than that of at-the-money options).
54. Compare Hall & Murphy, supra note 31, at 26-27 (concluding that "when existing compensation is adjusted, incentives are maximized through restricted stock grants rather than options")
55. and Dittmann & Maug, supra note 31, at 305 (reporting results of a model indicating that CEOs should receive restricted stock instead of options)
56. with Lambert & Larcker, supra note 29, at 2 ("[E]xercise price in the optimal contract is frequently far 'out of the money.'").
57. See Tian, Contracting, supra note 31, at 32-33.
58. He suggests, for example, that options are more likely to be optimal for younger executives who are less likely to be risk averse than older executives nearing retirement.
59. Id.
60. See John Core & Wayne Guay, The Use of Equity Grants to Manage Optimal Equity Incentive Levels, 28 J. ACCT. & ECON. 151, 152 (1999) (using delta as the measure of equity incentives).
61. See Wayne R. Guay, The Sensitivity of CEO Wealth to Equity Risk: An Analysis of the Magnitude and Determinants, 53 J. FIN. ECON. 43, 43 (1999) (using a vega-type measure of sensitivity of wealth to risk).
62. David I. Walker & Victor Fleischer, Book/Tax Conformity and Equity Compensation, 62 TAX L. REV. 399, 403-04, 410-11 (2009).
63. See Am. Inst, of Certified Pub. Accountants, Accounting for Stock Issued to Employees, Accounting Principles Bd. Opinion No. 25 (1972) [hereinafter APB 25].
64. See id. para. 12.
65. The FASB attempted to rationalize equity compensation accounting in the 1990s, but they succeeded only in implementing an elective regime that effectively left the 1972 standard in place while requiring firms to include pro forma earnings statements reflecting "fair value" accounting for options in the footnotes to their Financials.
66. See Fin. Accounting Standards Bd., Accounting for Stock-Based Compensation, Statement of Fin. Accounting Standards No. 123, § 5 (1995) [hereinafter SFAS 123].
67. Fair value was and is defined as the value arrived at through use of the Black-Scholes option pricing model or another appropriate model.
68. Id. § 19.
69. See id. § 18. The accounting expense for restricted stock is equal to the full fair market value of the stock at grant assuming that the employee is required to pay nothing explicitly for the stock, as is typical.
70. See generally Fin. Accounting Standards Bd., Share-Based Payment, Statement of Fin. Accounting Standards No. 123 (rev'd 2004).
71. I say 'largely" because some potential for distortion remains. The fair value of an option is determined using the Black-Scholes or binomial model and is manipulable. Thus, options provide some accounting flexibility that stock compensation does not provide.
72. See Walker & Fleischer, supra note 37, at 418-21 (describing potential for option expense manipulation).
73. See Ross L. WATTS & JEROLD L. ZIMMERMAN, PCSITIVE ACCOUNTING THEORY 133 (1986). The general idea is that in the presence of transaction costs, both renegotiation of earnings-based contracts to adjust for cosmetic changes and failure to do so can be costly.
74. See David I. Walker, Financial Accounting and Corporate Behavior, 64 WASH. & LEE L. REV. 927, 927 (2007);
75. 8see also Gregg D. Polsky, Controlling Executive Compensation Through the Tax Code, 64 WASH. & LEE L. REV. 877, 923-24 (2007) (arguing that section 162 (m) of the Internal Revenue Code, which disallows tax deductions for certain executive pay in excess of $1 million per year that is not performance based, encourages firms to adopt objective, formulaic bonus structures that can be manipulated through cosmetic adjustments to earnings).
76. See generally Walker, supra note 45, at 935-43.
77. However, in calculating alternative minimum taxable income, the spread on an ISO at exercise is included. I. R. C. § 56 (b) (3).
78. See generally David I. Walker, The Non-option: Understanding the Dearth of Discounted Employee Stock Options, 89 B. U. L. REV. 1505 (2009).
79. Apparently, deductibility under § 162 (m) of the Internal Revenue Code represents a second order tax consideration. Stock options qualify fairly easily as performance-based pay for purposes of this section and, thus, option payouts generally are fully deductible. Conventional, time-vested restricted stock is not considered performance-based pay and deductibility may be limited. One way to ensure deductibility of restricted stock is to condition vesting on achievement of performance objectives as well as continued employment. Many firms do so, but many other companies continue to grant conventional time-vested restricted stock.
80. See infra text accompanying note 67. As firms commonly report in their proxy statements, deductibility under § 162 (m) is a consideration, not a prerequisite, in compensation design.
81. As discussed infra note 58, compensation consultant Frederick W. Cook & Co. reports that since 2005, none of the largest 250 members of the S&P 500 has issued explicitly discounted options.
82. I also do not mean to suggest that incremental transaction costs associated with concurrent grants of stock and options are large. To the contrary, I suspect that the incremental administrative costs are modest, although the cost of complexity could be significant.
83. See infra Part II. C.2.b.i.
84. See Brian J. Hall, The Pay to Performance Incentives of Executive Stock Options 32 (Nat'l Bureau of Econ. Research, Working Paper No. 6674, 1998), available at http://papers. ssrn. com/sol3/papers.cfm?abstract-id=108563 (finding a "bias toward valuing options according [to] what they would be worth if exercised today").
85. See generally Lucian Arye Bebchuk et al., Managerial Power and Rent Extraction in the Design of Executive Compensation, 69 U. CHI. L. REV. 751 (2002).
86. The ultimate source for the data analyzed is individual company proxy statements. However, as discussed in the Appendix, most of the data is taken from S&Fs Compustat databases.
87. See infra Appendix A.
88. This view is consistent with the process described in most large company proxy statement disclosures regarding executive pay.
89. See infra note 117 and accompanying text.
90. See, e.g., Core et al., supra note 25, at 35-38.
91. See infra Part II. D.
92. For a more detailed overview of long-term executive incentive compensation practice at large U. S. public companies, see FREDERICK W. COOK & Co., THE 2009 TOP 250: LONG-TERM INCENTIVE GRANT PRACTICES FOR EXECUTIVES (2009).
93. The non-qualified and incentive labels applied to stock options refers to their federal income tax treatment.
94. See MYRON S. SCHOLES ET AL., TAXES AND BUSINESS STRATEGY: A PLANNING APPROACH 191-95 (2d ed. 2001) (discussing option tax treatment). Although these differences are important in some cases, given various limitations and current tax rates, the large majority of options granted are non-qualified options.
95. See Brian J. Hall & Jeffrey B. Liebman, The Taxation of Executive Compensation, 14 TAX POL'Y & ECON. 7 (2000) (estimating that about five percent of options granted are incentive stock options). Employee stock options generally become exercisable, or vest, in installments, often ratably across the period beginning on the first anniversary of the grant and ending on the fourth anniversary of the grant.
96. See FREDERICK W. COOK & Co., supra note 58, at 13 (providing data indicating vesting schedules of three to five years for ninety-eight percent of the executive stock options analyzed). If employment is terminated prior to vesting, options generally are forfeited.
97. As an example, in 2007 the CEO of Home Depot received an option grant that only vests if the company's share price exceeds the grant date price by twenty-five percent for thirty consecutive trading days.
98. See Home Depot, Proxy Statement (Form DEF 14A), at 32 (Apr. 11, 2008).
99. See FREDERICK W. COOK & Co., supra note 58, at 7, 18-23.
100. See id. at 7. Cook's 2009 survey omits the category of discounted options. Cook's 2008 survey reports no instances of companies granting discounted options since 2005.
101. See FREDERICK W. COOK & CO., THE 2008 TOP 250: LONG-TERM INCENTIVE GRANT PRACTICES FOR EXECUTIVES (2008).
102. Data on file with author.
103. Restricted stock awards may vest in installments or "cliff vest" on a single date. As in the case of options, most senior executive stock awards vest on a three-to five-year schedule.
104. FREDERICK W. COOK & Co., supra note 58, at 13.
105. Performance-vested restricted stock is analogous to performance-vested options. For example, in 2007 Moody's granted restricted stock to senior executives that vests relatively slowly, or relatively quickly, depending on growth in the company's annual operating income.
106. See Moody's Corp., Proxy Statement (Form DEF 14A), at 24-25 (Mar. 19, 2008).
107. Performance shares were formerly known as phantom stock. See, e.g., Eli Ofek and David Yermack, Taking Stock-Equity-Based Compensation and the Evolution of Managerial Ownership, 55 J. FIN. 3, 7 (2000).
108. The difference between the two devices is that restricted stock is granted at the time of the award and is forfeited if the shares fail to vest, while performance shares are not issued until performance criteria are met. But this difference is not significant economically. For example, under either type of plan, participants may be entitled to dividends.
109. Northern Trust Corporation's fiscal year 2007 performance share awards are typical. Each participant was assigned a target number of shares. If the company achieves average three-year earnings per share (EPS) growth of 10 percent, 100 percent of the target shares will vest at the end of three years. If EPS growth is between 8 percent and 10 percent, a fraction of the shares will vest. If EPS growth exceeds 10 percent, a multiple of target shares, up to 125 percent at 12 percent average EPS growth, will vest.
110. See Northern Trust Corp., Proxy Statement (Form DEF 14A), at 46 (Apr. 15, 2008).
111. Data on file with author. As discussed supra note 47 and accompanying text, deductibility of conventional time-vested restricted stock payouts may be limited under § 162 (m) of the Internal Revenue Code, but deductibility of conventional time-vested stock option payouts generally is not limited. This difference likely explains the greater use of performance-vested stock than performance-vested options.
112. The analysis was limited to the S&P 500 group of companies because of the labor intensive process of determining total stock grants in the pre-2006 period.
113. See infra Appendix A. A panel approach was used to ensure that changes in aggregate compensation were not driven by changes in the membership of the S&P 500. However, an analogous graph based on the equity grants of the historic S&P 500 membership each year would be very similar.
114. As discussed in Appendix B, Compustat data for stock grants and total compensation is not directly comparable before and after 2006. Appendix B describes how the data were adjusted to increase comparability. Nonetheless, while the relative contribution of stock and options in each period should be comparable, the absolute contributions of both to total compensation may not be fully comparable pre-2006 and post-2007. Non-equity compensation includes salary, annual bonuses, long-term incentive compensation that is not equity based, perquisites, and other compensation such as earnings on deferred compensation that are treated as compensation.
115. Because equity compensation in practice consists of binary combinations of stock and atthe-money options, convexity is essentially a function of the option-heaviness of the equity pay package; henceforth the terms "option-heaviness" and "convexity" will be used interchangeably.
116. See supra Part LB.
117. In his comment on this Article, Some Thoughts on the Evolution of Executive Equity Compensation, 64 VaND. L. REV. EN BANC (forthcoming 2011), Professor Herwig Schlunk provides a much more sophisticated analysis of the change in perceived growth opportunities based on changes in price/earnings ratios ("P/E") relative to the riskless return. His analysis indicates that the reduction in perceived growth opportunities across the period is greater than the simple analysis provided here would suggest. However, Schlunk does not argue that the corresponding shift from option to stock compensation is evidence of optimal contracting. Rather, he argues that the shift reflects managerial opportunism.
118. Recall that increasing volatility increases the value of an option but also increases the discount placed on an option by a non-diversified, risk-averse executive.
119. See supra note 25 and accompanying text.
120. All data is taken from the Compustat datasets. P/E is based on twelve-month trailing basic earnings per share and reflects averages of quarterly data. M/B and D/E reflect averages of annual data. D/E is defined as long term debt plus preferred stock divided by total shareholder equity. Volatility is sixty-month average volatility as reported in the Compustat database and utilized therein to calculate option values
121. Volatility data is not reported after 2006, so the volatility figures for the latter period are for the sixty-month period beginning on January 1, 2002 and ending December 31, 2006. There has been a large increase in firm-level volatility over the last fifty years, but it is difficult to account for the dramatic shift in equity design in the present decade based on that long-term trend.
122. See Diego Comin & Thomas Philippon, The Rise in Firm-Level Volatility: Causes and Consequences (Nat'l Bureau of Econ. Research, Working Paper No. 11388, 2005).
123. Volatility index data is available at CBOE, http://www.cboe.com/data/ Historical Volatility.aspx (1ast visited Mar. 2, 2011).
124. See supra Figure 3.
125. See, e.g., Core et al., supra note 25, at 154, 180.
126. See J. Carr Bettis et al., Exercise Behavior, Valuation, and the Incentive Effects of Employee Stock Options, 76 J. FIN. ECON. 446, 447 (2005) (finding for a sample of 140, 000 option exercises by executives at almost 4000 firms between 1996 and 2002 that, on average, options were exercised a little over two years following vesting and more than four years prior to expiration);
127. Jennifer N. Carpenter, The Exercise and Valuation of Executive Stock Options, 48 J. FIN. ECON. 127, 138 (1998) (finding for a sample of forty firms (mainly large manufacturers) that executive stock options granted between 1983 and 1984 were, on average, exercised after 5.8 years);
128. Steven Huddart & Mark Lang, Employee Stock Option Exercises: An Empirical Analysis, 21 J. ACCT. & ECON. 5, 20 (1996) (finding that the median fraction of option life elapsed at the time of exercise ranged from 0.21 to 0.38 for options granted by seven public companies to a wide range of employees).
129. See, e.g., Paul M. Healy & Krishna G. Palepu, The Fall of Enron, 17 J. ECON. PERSPECTIVES 3, 13 (2003) (noting that "[h]eavy use of stock option awards linked to short-term stock price may explain the focus of Enron's management on creating expectations of rapid growth and its efforts to puff up reported earnings to meet Wall Street's expectations").
130. Although credit for discovering backdating properly belongs to finance professor Erik Lie, the scandal received public attention after it was exposed in the Wall Street Journal.
131. See Erik Lie, On the Timing of CEO Stock Option Awards, 51 MGMT. SCI. 802, 810 (2005) (providing convincing evidence that options were backdated);
132. Charles Forelle & James Bandler, The Perfect Payday: Some CEOs Reap Millions by Landing Stock Options When They Are Most Valuable, WALL ST. J., Mar. 18, 2006, at A1 (reporting evidence of option backdating to a broad readership).
133. See generally RICHARD A. BREALEY ET AL., PRINCIPLES OF CORPORATE FINANCE (8th ed. 2005).
134. The Jobs and Growth Tax Relief Reconciliation Act of 2003 cut the top marginal federal income tax rate applicable to dividends from thirty-five percent to fifteen percent. For evidence on the impact on dividend payouts
135. see Jeffrey R. Brown et al., Executive Financial Incentives and Payout Policy: Firm Responses to the 2003 Dividend Tax Cut, 62 J. FIN. 1935, 1935 (2007) (reporting that thirty-five percent of S&P 1500 firms increased dividend payouts in 2003 compared with twenty-seven percent increasing payouts in the two prior years and that the rate of firms newly adopting dividend programs increased from about one in one hundred in 2001 and 2002 to one in ten in 2003).
136. See Kevin J. Murphy, Executive Compensation, in 3 HANDBOOK OF LABOR ECONOMICS 2485, 2509-10 (Orley Ashenfelter & David Card eds., 1999).
137. See, e.g., Christine Jolls, Stock Repurchases and Incentive Compensation (Nat'l Bureau of Econ. Research, Working Paper No. W6467, 1998) (finding that companies that rely heavily on stock option compensation for executives are more likely than other firms to repurchase shares, presumably as an alternative to paying dividends).
138. See Brown et al., supra note 85.
139. I thank Dhammika Dharmapala for this suggestion.
140. About seventy-five percent of S&P 500 firms have a fiscal year ending in December.
141. See generally STANDARD & POOR'S 500 GUIDE (13th ed. 2010).
142. See Fin. Accounting Standards Bd., Invitation to Comment: Accounting for Stock-Based Compensation: A Comparison of FASB Statement No. 123, Accounting for Stock-Based Compensation, and Its Related Interpretations, and IASB Proposed IFRS, Share-based Payment, FASB Index No. 1102-001 (Nov. 18, 2002).
143. See SFAS 123, supra note 40;
144. Expensing Stock Options: Can FASB Prevail? KNOWLEDGEIGWHARTON, http://knowledge.wharton. upenn. edu/article.cfm?articleid=975 (last visited Feb. 2, 2011) ("The last FASB effort to require an options-expense treatment, back in 1994, foundered in the face of political and industry opposition that threatened the Board's very existence.").
145. See Lawrence D. Brown & Yen-Jung Lee, The Impact of SFAS 123R on Changes in Option-Based Compensation (May 2007) (unpublished manuscript), available at http://ssrn. com/abstract=930818.
146. See Mary Ellen Carter et al., The Role of Accounting in the Design of CEO Equity Compensation, 82 ACCT. REV. 327 (2007).
147. See id. at 353.
148. See id. at 354.
149. Stock value is the value of the stock at grant with no adjustment for restrictions. Option value is the Black-Scholes ex ante value as reported in Compustat. The x-axis labels in the figure represent the midpoint of ratio ranges. For example, the eighty-nine observations at x-axis label 0.55 represent ratios greater than 0.50 up to and including 0.60.
150. Eighty-six percent of the executives at panel firms received some amount of equity compensation during 1997.
151. A third possible type would be firms that limited equity compensation to stock. However, the frequency of stock-only grants is more plausibly explained as censoring of the distribution than is the frequency of option-only grants. See infra notes 103-07 and accompanying text for more on the censoring possibility.
152. Setting aside stock-only and option-only grants, the grant-year based distribution of the ratio of stock grants to total equity pay conferred for panel firms for 2007 is normal with a mean of fifty-three percent stock. Although ninety-three percent of panel firm executives received an equity grant in 2007, a focus on a single year's equity grants tends to overstate the clustering at the extremes of the distribution. Some executives who received only stock or options in 2007 may have received the other form of compensation in 2006 or 2005. Accounting data can be used to measure the distribution of equity holdings, rather than annual grants, but the distribution based on accounting data remains censored or trimodal. Moreover, given the overall shift in emphasis from option compensation to stock in recent years, accounting-based data may overstate the extent to which firms are granting a mix of stock and options. Eli Lilly, for example, consistently issued options to its senior executives through 2003. In 2004 through 2006, it issued a mix of stock and options. In 2007 and since, it has issued stock exclusively. Apparently, 2004 through 2006 were transition years. However, accounting data for 2007 and later years continues to reflect mixed grants issued in 2004 through 2006.
153. The middle two columns of the histogram represent stock ratios ranging from just over forty percent up to and including sixty percent.
154. The extent to which firms grant the same equity compensation mix to the various members of the executive team is explored further in the following section.
155. See infra Part II. C.2.b.
156. See supra notes 33-34 and accompanying text.
157. See Hall, supra note 52, at 32 (finding a "bias toward valuing options according [to] what they would be worth if exercised today").
158. On the salience bias, see generally A. Tversky & D. Kahneman, Availability: A Heuristic for Judging Frequency and Probability, 5 COGNITIVE PSYCHOL. 207 (1973).
159. The managerial power model of the compensation-setting process explains the dearth of out-of-the-money options as outrage management. In-the-money options would produce outrage on the part of investors and the financial press. Nondiscounted options, whether at or out of the money, are likely to produce similar investor and financial press response. If so, compensation value per unit of outrage is maximized by granting options at the money.
160. See Bebchuk et al., supra note 53.
161. For twenty-two percent of the executives who received stock grants but not options in 2007, the value of the stock accounted for a relatively modest thirty percent or less of total compensation. For executives receiving both stock and options, in only nine percent of the cases did equity pay account for thirty percent or less of total compensation. However, executives who received options but not stock were more likely than executives in either of the other two groups to receive pay packages that were light on equity compensation. In thirty-one percent of these cases, equity value accounted for thirty percent or less of total compensation. Thus, this data is as or more consistent with the intuitive idea that firms that place relatively less emphasis on equity pay are more likely to utilize a single pay instrument rather than both stock and option pay than with truncation of equity mix at the stock-only end of the distribution.
162. See infra Appendix C.
163. See Walker & Fleischer, supra note 37, at 424-26 (discussing inadequacy of the Black-Scholes model for valuing long-term, non-transferable options).
164. Ford Motor Co., Proxy Statement (Form DEF 14A), at 39 (Apr. 4, 2008).
165. See SEC Regulation S-K, Item 402, 17 C. F. R. § 229. 402 (2010).
166. Exxon Mobil Corp., Proxy Statement (Form DEF 14A), at 23 (Apr. 10, 2008).
167. See, e.g., Consol. Edison Inc., Proxy Statement (Form DEF 14A), at 26 (Apr. 11, 2008);
168. Lennar Corp., Proxy Statement (Form DEF 14A), at 20 (Mar. 7, 2008).
169. See, e.g., Pinnacle W. Capital Corp., Proxy Statement (Form DEF 14A), at 19 (Apr. 10, 2008);
170. Verizon Commc'ns Corp., Proxy Statement (Form DEF 14A), at 26 (Mar. 17, 2008).
171. See, e.g., Bemis Co., Proxy Statement (Form DEF 14A), at 18 (Mar. 19, 2008).
172. See, e.g., Progress Energy Inc., Proxy Statement (Form DEF 14A), at 27 (May 14, 2008);
173. Int'l Paper Co., Proxy Statement (Form DEF 14A), at 52 (Apr. 8, 2008).
174. See, e.g., Verizon Commc'ns Corp., Proxy Statement (Form DEF 14A), at 25 (Mar. 17, 2008);
175. Lennar Corp., Proxy Statement (Form DEF 14A), at 20 (Mar. 7, 2008).
176. See, e.g., Unisys Corp., Proxy Statement (Form DEF 14A), at 24 (June 18, 2008);
177. Lennar Corp., Proxy Statement (Form DEF 14A), at 20 (Mar. 7, 2008).
178. See, e.g., Analog Devices, Proxy Statement (Form DEF 14A), at 27 (Feb. 6, 2008);
179. U. S. Bancorp, Proxy Statement (Form DEF 14A), at 22 (Mar. 4, 2008);
180. Ecolab Inc., Proxy Statement (Form DEF 14A), at 29 (Mar. 19, 2008).
181. See, e.g., U. S. Bancorp, Proxy Statement (Form DEF 14A), at 22 (Mar. 4, 2008);
182. Schlumberger Ltd., Proxy Statement (Form DEF 14A), at 22 (Feb. 29, 2008).
183. See, e.g., Mattel, Inc., Proxy Statement (Form DEF 14A), at 37 (Apr. 24, 2008);
184. Int'l Game Tech., Proxy Statement (Form DEF 14A), at 25 (Feb. 27, 2008).
185. See, e.g., Mattel, Inc., Proxy Statement (Form DEF 14A), at 37 (Apr. 24, 2008);
186. Int'l Game Tech., Proxy Statement (Form DEF 14A), at 25 (Feb. 27, 2008).
187. See, e.g., Mattel, Inc., Proxy Statement (Form DEF 14A), at 37 (Apr. 24, 2008) (dividing value equally into stock and options);
188. Coca-Cola Co., Proxy Statement (Form DEF 14A), at 39 (Mar. 3, 2008) (allocating sixty percent of value to options and forty percent to stock);
189. Kimberly Clark Corp., Proxy Statement (Form DEF 14A), at 43 (Mar. 4, 2008) (dividing equally into timevested restricted stock, performance-vested restricted stock, and options).
190. Shlomo Benartzi & Richard H. Thaler, Naive Diversification Strategies in Defined Contribution Savings Plans, 91 AM. ECON. REV. 79, 80 (2001).
191. See id. at 82 (reporting that thirty-four percent of respondents chose an exact 50/50 mix and, on average, allocated fifty-four percent of funds to the stock fund).
192. Id.
193. Given Compustat's coding practices, we cannot know whether "restricted stock" grants in 2005 consisted of conventional, time-vested restricted stock or of performance-vested restricted stock, and it is possible that some executives received both. Given the relative scarcity of performance-vested restricted stock, however, we can safely assume that cases in which firms made grants of both time-vested restricted stock and performance-vested restricted stock were few.
194. To be sure, performance shares and performance-vested restricted stock can provide an element of optionality that is not present in conventional time-vested restricted stock. If minimum performance targets are not met, the stock is lost. Generally, however, plans provide for a range of performance targets and payouts and cap payouts on the high side as well, so the asymmetric payouts that are achievable with options generally are not achievable with performance shares or performance-vested stock. The 2005 data suggest that firms consider performance shares and restricted stock to be roughly equivalent. When paired singly with options, restricted stock constituted forty-six percent of the mix, on average, while performance shares constituted forty-eight percent of the mix, on average.
195. Similarly, the inclusion of a third equity compensation instrument did not appear to increase total compensation. Among executives receiving two equity instruments, mean total ex ante compensation was $5.4 million ($11.6 million for CEOs only); among executives receiving three instruments, mean total ex ante pay was $5.2 million ($10.7 million for CEOs only).
196. See supra Part I. D (discussing the managerial power view of the compensation-setting process).
197. Of thirty-one S&P 500 companies that did not utilize a compensation consultant in 2007, fifteen (forty-eight percent) issued stock or options exclusively or almost exclusively (over ninety-five percent) to their senior executives in 2006 and 2007. Less than twenty-five percent of companies that employed a consultant relied exclusively or almost exclusively on stock or options. This difference is statistically significant at the one percent level. Data on file with author.
198. Ruth Bender, Paying For Advice: The Role of the Remuneration Consultant in U. K. Listed Companies, 64 VAND. L. REV. 361, 368-69 (2011).
199. Kedia and Rajgopal have recently found that headquarters location explains variation in broad-based option plans, but the authors found no evidence that location affected senior executive options plans.
200. See Simi Kedia & Shiva Rajgopal, Neighborhood Matters: The Impact of Location on Broad Based Stock Option Plans, 92 J. FIN. ECON. 109, 125 (2009).
201. The limited degree of individualization will not be surprising to practitioners. The primary aim of this section is to describe intra-firm grant practices and offer thoughts on why we do not observe a higher degree of individualization of equity mix. I do not mean to make any claims regarding the "right" level of individualization in the presence of transaction costs and other limitations.
202. See Tian, Contracting, supra note 31, at 32.
203. See supra notes 35-36 and accompanying text.
204. Core & Guay, supra note 35, at 152-54.
205. I am principally interested in within-firm variation in recent years (given much greater inter-firm variation) and thus there was no reason to limit my analysis to the panel of firms for which data is available back to 1992.
206. The one percent standard deviation cutoff is arbitrary, but is meant to count as lockstep cases with minor deviations in the mix of stock and options granted resulting from rounding the number of shares in equity grants or from grants occurring on different dates. For example, the five top executives of Assurant, Inc. received grants of both stock and options for 2006.
207. Assurant, Inc., Proxy Statement (Form DEF 14A), at 22 (Apr. 12, 2007). The ratios of the values of their stock grants to the sums of those values and the values of their option grants were 21.9 percent, 20.4 percent, 21.4 percent, 20.9 percent, and 22.5 percent, with a standard deviation of 0.9 percent. This spread in ratios is at the high end of what was considered to represent a lockstep grant. To provide further context, the one percent standard deviation cutoff is about five percent of the standard deviation of the stock percentage (nineteen percent) for the entire population of 1, 064 executives who received a mix of stock and options.
208. See, e.g., supra notes 111-15 and accompanying text.
209. McGraw-Hill Cos., Proxy Statement (Form DEF 14A), at 30 (Mar. 20, 2008).
210. See, e.g., Gen. Dynamics Corp., Proxy Statement (Form DEF 14A), at 18 (Mar. 21, 2008) C[A]s a matter of principle, we do not consider the value of past equity grants when determining current compensation. Our responsibility in setting compensation is to ensure that the value of the equity grants, at the time they are received, is reasonable.").
211. The economic literature on signaling posits that inferences can be drawn from choices that have differing costs for differing types.
212. See Michael Spence, Job Market Signaling, 87 Q. J. ECON. 355, 358-59 (1973).
213. Of course, a high degree of inherent risk aversion might in itself be incompatible with certain executive roles in certain industries (e.g., chief development officer of a high tech company), while in other cases it might be viewed as a plus (e.g., COO of a utility company). This analysis assumes that a range of inherent risk aversion levels would be compatible with the job role and that the information problem lies in distinguishing inherent risk aversion from low skill. At some companies executives are given a choice between receiving equity compensation in the form of stock or options.
214. See, e.g., Best Buy Co., Proxy Statement (Form DEF 14A), at 34-36 (May 12, 2008) (describing plan that allows senior executives to choose between equal value packages including one-hundred-percent options, a 50/50 mix of restricted stock and options, and two other combinations).
215. Signaling aside, if the company set the price differential between the two instruments in such a way that shareholders would be indifferent, providing executives with a choice would seem to increase welfare. To my knowledge, however, very few companies employ this approach, perhaps because of the potential negative signals sent by those who do not elect option-heavy packages.
216. Coca-Cola Co., Proxy Statement (Form DEF 14A), at 39 (Mar. 3, 2008).
217. I thank Dan Shaviro for this suggestion. The value of equity grants post-vesting would, of course, depend on individual decisions regarding exercise. These decisions could disrupt the ordinal ranking of compensation within the executive suite even if all compensation were granted in lockstep. However, executives might reasonably view post-vesting gains and losses as being personal and distinct from pre-vesting value changes.
218. See infra Part II. C.2.iii.
219. See infra Parts II. C.1-.2.
220. This final factor is somewhat ambiguous, however, as increased shareholdings also reduce the subjective value of high-powered incentives to the executive relative to their cost.
221. See Hall & Murphy, supra note 31, at 210-11 (finding that the value of at-the-money options to executives relative to the cost to shareholders is "decreasing in risk aversion, increasing in nonfirm-related wealth, and decreasing in holdings of company stock").
222. After eliminating observations in which each executive at a firm received only stock or only options, I was left with 3, 303 S&P 500 executive equity grants for 2006 and 2007, consisting of 634 CEO grants and 2, 669 grants to subordinate executives. The mean ratio of stock value to total equity value for CEOs was 0.508 and for subordinate executives was 0.525.
223. A two-tailed ttest was used to compare the means. The test yielded a t-statistic of-1.27 and a probability that there is in fact no difference in means of 20.5 percent.
224. CFO equity packages were also examined. One might expect CFO equity packages to be less option heavy given the nature of the role and the purported link between options holdings and earnings management. However, average S&P 500 CFO equity mix for 2006 and 2007 was virtually indistinguishable from the mix granted to the remaining senior executives. Only the CEO and CFO positions are identified in Compustat, so no analysis was performed for the COO or chief technology officer roles.
225. See, e.g., Bhagat & Romano, supra note 5, at 363, 366-67 (suggesting that executives not be allowed to dispose of equity compensation prior to retirement);
226. Posner, supra note 5, at 1045-46 (recommending that a "significant share" of executive compensation should be backloaded and suggesting that restricted stock should constitute a minimum fraction of CEO pay);
227. Craig, supra note 5 (relating comments of House Committee on Financial Services Chairman Barney Frank advocating broader application of rules tying executive pay to long-term performance).
228. American Recovery and Reinvestment Act of 2009, Pub. L. No. 111-5, § 7001 (b) (2) - (3) (A) (2009).
229. Id. Under the Act, the stock cannot vest before the government loans are repaid.
230. See id.
231. See, e.g., Craig, supra note 5 (reporting comments of House Financial Services Committee Chairman Barney Frank suggesting more widespread linkage of executive pay to long-term company performance);
232. Press Release, H. Comm. on Fin. Servs., 111th Cong., Chairman Frank Holds News Conference to Discuss the Committee Agenda and Priorities for the Coming Year (Feb. 2, 2009), available at http://www.knowledgeplex.org/news/3000221.html (indicating in question and answer session that the potential extension of executive compensation regulation "doesn't just go for TARP recipients").
233. See, e.g., Bhagat & Romano, supra note 5, at 361 (advocating significantly greater holding periods for equity-based incentive compensation).
234. See David I. Walker, The Challenge of Improving the Long-Term Focus of Executive Pay, 51 B. C. L. REV. 435, 468-69 (2010) (suggesting that any coercive pay regulation aimed at remedying short-term thinking by executives should be limited to restricting the holding period of pay-and not the methods or instruments-in order to minimize inefficiencies);
235. see also Bhagat & Romano, supra note 5, at 371-72 (recommending that executive pay consist of some combination of long-term stock and long-term option compensation).
236. "Say on pay" proposals uniformly involve an advisory up-or-down shareholder vote on executive compensation.
237. See Jeffrey N. Gordon, "Say on Pay": Cautionary Notes on the U. K. Experience and the Case for Shareholder Opt-In, 46 HARV. J. ON LEGIS. 323, 324-25, 338-40, 348-53 (2009) (discussing U. S. legislative proposals and corporate governance activists' attempts to implement "say on pay").
238. The Dodd-Frank financial regulatory reforms enacted in 2010 mandate non-binding shareholder voting on executive pay beginning in 2011. Dodd-Frank Wall Street Reform and Consumer Protection Act, Pub. L. No. 111-203, § 951 (2010).
239. See Fabrizio Ferri & David Maber, Say on Pay Votes and CEO Compensation: Evidence from the UK 3-4, 23-25 (Oct. 15, 2010) (unpublished manuscript), available at http://ssrn. com/abstracts1420394 (finding that 2002 U. K. "say on pay" legislation had no effect on the level or growth rate of CEO pay, but finding an increase in pay sensitivity to poor performance).
240. But see Gordon, supra note 157, at 345 (suggesting that the effect documented by Ferri and Maber could result in excessive executive conservatism).
241. As noted above, it is possible that companies optimize equity mix for CEOs and then apply the same mix in granting equity to subordinate executives.
242. See supra notes 148-51 and accompanying text. It is clear, however, that firms generally do not optimize for each member of the executive team.
243. Actually, even this requirement has evolved somewhat. The current rule requires disclosure for the CEO, CFO, and the three most highly compensated employees other than the CEO and CFO; prior rules required disclosure for the firm's CEO and the four most highly compensated employees other than the CEO.
244. Compare 17 C. F. R. § 229. 402 (a) (3) (ii) - (iii) (2007) (new rule)
245. with 17 C. F. R. § 229. 402 (a) (i) - (ii) (2006) (old rule). However, for most firms the two rules produce the same list of executives.
246. See Fin. Accounting Standards Bd., supra note 42.
247. See SEC Reg. S-K, 17 C. F. R. § 229. 402 (c) (2) (vi) (2010) (detailing requirements for summary compensation table);
248. SEC Reg. S-K, 17 C. F. R. § 229. 402 (d) (2) (viii) (2010) (detailing requirements for table of plan-based awards).
249. The Compustat database is not error-free. We found numerous examples of miscoded data, and surely many more cases eluded us since we focused on errors that produced outlier results, such as the 2006 $2.2 million restricted stock grant to an executive of Baxter International that is coded in Compustat as a $2.2 billion grant. However, I have no reason to think that I could do a better job than S&P of coding this data even if I had several lifetimes to accomplish the task.