Insider trading and CEO pay

Vanderbilt Law Review

Volumn 64, Issue 2, 2011, Pages 505-555

  Todd Henderson, M ^a  

^a UNIVERSITY OF CHICAGO   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

EID: 79959426920     PISSN: 00422533     EISSN: None     Source Type: Journal    
DOI: None     Document Type: Article

References (115)

1. Darren Roulstone, The Relation Between Insider-Trading Restrictions and Executive Compensation, 41 J. ACCT. RES. 525, 526 (2003) (finding that firms restricting executive trading (using black-out windows) pay a 4-13% premium in terms of total compensation).
2. As described below, Bebchuk and others describe the problem of insiders earning unwarranted and unnoticed profits on informed trades as a systematic problem in need of an across-the-board regulatory response. See infra Section VI. A.
3. See Roulstone, supra note 1, at 527-31.
4. See, e.g., M. Todd Henderson et al., Strategic Disclosure of 10b5-1 Trading Plans (Univ. of Chi. Law & Econ., Working Paper No. 411, 2008), available at http://papers.ssrn. com/sol3/papers.cfm?abstract-id=l 137928 (showing about $25 billion in trades by insiders under Rule 10b5-1 trading plans in first year of use by each executive).
5. See HENRY MANNE, INSIDER TRADING AND THE STOCK MARKET 147-58 (1966).
6. Id. at 132-33, 136-38.
7. That is, a manager of a firm trading on the basis of information about the firm where she works. See United States v. O'Hagan, 521 U. S. 642, 651-52 (1997).
8. See infra notes 84-88 and accompanying text.
9. See Michael C. Jensen & Kevin J. Murphy, CEO Incentives-It's Not How Much You Pay, But How, HARV. BUS. REV., May-June 1990, at 138, 141.
10. For historical data, see, for example, Kevin J. Murphy, Executive Compensation 16-23 (Apr. 1998), available at http://papers.ssrn. com/sol3/papers.cfm?abstract-id=163914. Data for the period of this study is from the ExecuComp database maintained by the University of Pennsylvania.
11. See, e.g., LUCIAN A. BEBCHUK & JESSE M. FRIED, PAY WITHOUT PERFORMANCE: THE UNFULFILLED PROMISE OF EXECUTIVE COMPENSATION 8 (2006) ("We agree that paying generously to provide desirable incentives can be a good compensation strategy for shareholders.... Our concern is simply that executives have partly taken over the compensation machine, leading to arrangements that fail to provide managers with desirable incentives.").
12. See Roulstone, supra note 1, at 526.
13. See J. C. Bettis et al., Corporate Policies Restricting Trading by Insiders, 57 J. FIN. ECON. 191, 199-201, 206 (2000).
14. See id. at 208.
15. Eran Ofek & David Yermack, Taking Stock: Equity-Based Compensation and the Evolution of Managerial Ownership, 55 J. FIN. 1367, 1370 (2000).
16. John Core & Wayne Guay, When Contracts Require Risk-Averse Executives to Hold Equity: Implications for Option Valuation, for Relative Performance Evaluation, and for the Corporate Governance Debate 7, (U. of Pa. Inst, for Law & Econ., Research Paper No. 03-32, 2003), available at http://ssrn. com/abstract=429301.
17. There is no empirical proof for this claim. It is a claim based on faith in markets for labor and capital doing their work.
18. In addition, it may relieve the board from having to calculate with precision the optimal amount of equity to be held at any time. Determining the best mix of pay and the amount of equity an executive should hold at any time is a difficult task for the board. Executives have better information than the board about the optimal structure of their portfolio and how they value options. It may therefore be easier for the board to allocate an amount of options it believes, based on its information at the time, is necessary to give the executive the proper incentives, but then permit the executive to trade (within a given tolerance) shares to locally maximize the executive's wealth.
19. See, e.g., H. NEJAT SEYHUN, INVESTMENT INTELLIGENCE FROM INSIDER TRADING 61 (1998) (finding stock prices rising significantly after insider purchases and falling significantly following insider sales);
20. Steven Huddart & Mark Lang, Information Distribution Within Firms: Evidence from Stock Option Exercises, 34 J. ACCT. & ECON. 3, 6 (2002);
21. H. Nejat Seyhun, The Effectiveness of the Insider Trading Sanctions, 35 J. L. & ECON. 149, 154 (1992) (showing insiders earn nearly 10% in abnormal returns in the year following trades).
22. Firms could impose internal controls to reduce the prevalence of informed trading. For instance, firms could ban trading until after the executive leaves the firm, could require trades to be executed at random times or at the unchangeable discretion of a third party, or could require all trades to be approved by the board or general counsel.
23. There are other rules too. Section 16 (b) of the Securities Exchange Act of 1934, the socalled short-swing profit rule, requires statutory insiders to disgorge any profits earned on paired buy and sell transactions within six months. Securities and Exchange Act of 1934 §16, 15 U. S. C. § 78p (2006).
24. In addition, Rule 14e-3 prohibits anyone from trading on material, non-public information about a pending tender offer. 17 C. F. R. § 240. 14e-3 (2010).
25. For a summary of the literature and an analysis of legislation designed to increase the ratio of meritorious to meritless lawsuits, with specific focus on the Private Securities Litigation Reform Act of 1995, see generally Marilyn F. Johnson et al., Do the Merits Matter More? The Impact of the Private Securities Litigation Reform Act, 23 J. L. ECON. & ORG. 627 (2007).
26. See, e.g., Dale Morse & Neal Ushman, The Effect of Information Announcements on the Market Microstructure, 63 ACCT. Rev. 247, 249 (1983) (" (T]he bid/ask spread should be positively related to... the likelihood of private information existing....").
27. The SEC tried to do this with Rule 10b5-1, but, as discussed below, it failed. See infra notes 91-94 and accompanying text. Some firms have also tried by making public announcements about restrictions on trading to specific periods, or by requiring authorization by the general counsel or board for any trades.
28. If we observe no bargaining, this would be relatively strong evidence of the managerial power theory of executive compensation. See, e.g., BEBCHUK & FRIED, supra note 11. Evidence of bargaining, on the contrary, does not refute the managerial power theory, but offers a way of testing its strength.
29. See Henderson et al., supra note 4, at 13;
30. Roulstone, supra note 1, at 527.
31. For a summary of the theories on why insiders value the ability to trade at any time independent of their information, see Roulstone, supra note 1, at 529-30.
32. See id.
33. The typical compensation contact also includes, among other things, perquisites, deferred compensation, health and retirement benefits, and contributions to long-term incentive plans.
34. See Roulstone, supra note 1, at 548-49.
35. Id. at 526.
36. Id. at 525.
37. Id. at 526. Roulstone also finds that firms restricting trading "use more incentive-based compensation and their insiders hold larger equity incentives relative to firms that do not restrict insider trading."
38. Id. at 525. This is consistent with the theory above about the cost arising from the inability of insiders to optimize their portfolio. Insiders who cannot trade as liberally should receive more shares than those who can rebalance their portfolio, all else being equal. This is what Roulstone finds.
39. Id.
40. Linda Chatman Thomsen, Director of the SEC Division of Enforcement, stated recently that "the idea [of Rule 10b5-1] was to give executives opportunities to diversify or become more liquid through the use of plans with prearranged trades without facing the prospect of an insider trading investigation. " Linda Chatman Thomsen, Dir., SEC Div. of Enforcement, Opening Remarks Before the 15th Annual NASPP Conference (Oct. 10, 2007), available at www.sec.gov/news/speech/2007/spchl010071ct.htm.
41. Specifically, the Rule provides an insider an affirmative defense if: (A) Before becoming aware of the information, the person had: (1) Entered into a binding contract to purchase or sell the security, (2) Instructed another person to purchase or sell the security for the instructing person's account, or (3) Adopted a written plan for trading securities; (B) The contract, instruction, or plan described in paragraph (c) (l) (i) (A) of this Section: (1) Specified the amount of securities to be purchased or sold and the price at which and the date on which the securities were to be purchased or sold; (2) Included a written formula or algorithm, or computer program, for determining the amount of securities to be purchased or sold and the price at which and the date on which the securities were to be purchased or sold; or (3) Did not permit the person to exercise any subsequent influence over how, when, or whether to effect purchases or sales; provided, in addition, that any other person who, pursuant to the contract, instruction, or plan, did exercise such influence must not have been aware of the material nonpublic information when doing so; and (C) The purchase or sale that occurred was pursuant to the contract, instruction, or plan. A purchase or sale is not "pursuant to a contract, instruction, or plan" if, among other things, the person who entered into the contract, instruction, or plan altered or deviated from the contract, instruction, or plan to purchase or sell securities (whether by changing the amount, price, or timing of the purchase or sale), or entered into or altered a corresponding or hedging transaction or position with respect to those securities. 17 C. F. R. § 240. 10b5-1 (c) (i) (2010).
42. Henderson et al., supra note 4, at 2. For example, a firm may authorize an executive to enter into a 10b5-1 trading plan on January 1 that commits the insider to trade on March 31, even though the firm may otherwise ban trades in the period right before the firm announces its first-quarter earnings on April 1.
43. The ability to borrow against shares may be limited because the pledgee is likely to value the shares at a large discount. This is because the pledgee steps into the shoes of the pledgor (who may be restricted in the type or amount of sales that can be made) and because default on the loan may be highly correlated with a drop in the value of the stock. In addition, if the pledgee needs to force the sale of large numbers of shares in a fire sale, this may cause the value of the collateral to drop. Finally, pledgees are likely to discount shares or options because of the concern about information asymmetries between the pledgor, who is an insider, and the pledgee. It is likely for these reasons that many firms ban insiders from using shares as collateral. For a recent media account of these issues, see Reed Abelson, Insiders' Share Sales on Margin on the Rise, N. Y. TIMES, Oct. 20, 2008, at B1, available at http://www.nytimes.com/2008/10/20/business/20pay.html.
44. We should expect executives in firms that do not authorize 10b5-1 plans to be paid more than executives in firms that do, since the in-plan insiders can be expected to earn greater profits from trading their shares. In the nomenclature above, T is greater for executives that are authorized to use Rule 10b5-1 than for executives who cannot use rule. Accordingly, holding an executive's reservation wage constant, executives with greater T should see a commensurate reduction in the value of S and/or B.
45. For example, the econometric analysis reported below tries to account for other determinants of pay by controlling for industry, firm size, economic performance, and other variables.
46. See TRUMAN F. BEWLEY, WHY WAGES DON'T FALL DURING A RECESSION 150-52, 160-61, 170-73 (1999) (discussing reasons for and timing of increases and decreases in wages).
47. See Alan D. Jagolinzer, SEC Rule 10b5-1 and Insiders' Strategic Trade, 55 MGMT. SCI. 224, 226-27 (2009) (discussing the features and restrictions of Rule 10b-5 that permit abnormal market returns).
48. Id.
49. Roulstone, supra note 1, at 536-38, 544-46, 548-49 (describing the results of the author's insider trading studies).
50. Henderson et al., supra note 4, at 19, 43-44.
51. See Jagolinzer, supra note 41, at 232-33.
52. See Henderson et al., supra note 4, at 28.
53. As HJM surmise, this disclosure choice is based on a tradeoff between the benefits and costs of disclosure. Id. at 1. Disclosure (and the more the better) increases the value of the litigation deterrence, since only publicly disclosed plans can deter suits from being filed, and only publicly disclosed plans can be admitted (as public documents) at the motion to dismiss stage before discovery costs increase. On the other hand, detailed disclosure increases the commitment value to trade in a particular way, since observers can determine ex post whether insiders have followed through on their plan. In addition, it may allow market participants to front run the insider's planned trades and take any profits for themselves. From this model, we can determine that insiders with a high risk of suit and no plans to change their planned trades-those who are certain of a large negative price drop, for instance-are likely to make detailed disclosures, while those with a lower, but nonzero, risk of suit and valuing the option of being able to change their plans-those with uncertainty about future price drops-are likely to make less specific disclosures.
54. See id. at 9.
55. Id. at 2.
56. See id. at 15.
57. "Similar" here means within 20% of the particular firm at issue.
58. A comparison of median values, not reported here, shows the same result.
59. These results are in unreported t-tests.
60. The data include only CEOs who are in the same position for the two periods examined in the study.
61. Henderson et al., supra note 4, at 3 n. 7, 28.
62. Id. at 14.
63. Id. at 19, 43.
64. Id. at 20.
65. Id. at 22.
66. For the present study, the limited and specific disclosure groups are combined into a single disclosure group, and these firms are compared with firms in the nondisclosure group of HJM.
67. Nondisclosure firms earned about $816 million in operating income on average, compared with $805 million for disclosure firms. The results are, however, not statistically significant.
68. Unreported univariate regressions show that the change in market value is statistically significant and is consistently the best predictor of changes in executive pay across all of the several thousand firms in the HJM database.
69. The methodology and dataset of disclosure firms is taken from Henderson et al., supra note 4, at 11-15.
70. (t+1) - (t-1) is the change in the percentage of stock owned by the insider from the pre-to the post-plan years.
71. Compensation data in the sample are reported in thousands of dollars, and the variable for this coefficient is a dummy (either 0 or 1). So a coefficient of 1, 220 translates to about $1.2 million.
72. See supra Part IV (comparing disclosure and non-disclosure firms and compensation using 10b5-1 trading plans data).
73. See supra Part IV.
74. The median CEO of a plan/nondisclosure firm had an increase of about $500, 000 in total pay over the period, while the median CEO of a nonplan firm had an increase of about $350, 000. The performance of the median plan/nondisclosure firm was an increase of market value of about $300 million, while the median nonplan firm increased by about $200 million.
75. In the pre-10b5-1 period, the median disclosure firm grew 2.6 times faster than the median nondisclosure firm; in the 10b5-l period, the median disclosure firm grew 2.2 times faster than the median nondisclosure firm.
76. The coefficient for "market value" is positive (0.16) and is statistically significant at the 1% level (p-value < 0.01); the coefficient for "assets" is positive (0.36) and is statistically significant at the 1% level (p-value < 0.01); and the coefficient for the "disclosure" dummy is not statistically significant (0.61). The other coefficients are not statistically significant. This means the typical determinants of pay (that is, market value and asset size) are significant determinants of pay across the two groups of firms in the pre-10b5-1 period, but there is no correlation between the fact of being in the groups and changes in pay over the period in question.
77. See Lucian Arye Bebchuk et al., Managerial Power and Rent Extraction in the Design of Executive Compensation, 69 U. CHI. L. REV. 751, 831 (2002) (noting that the ability and freedom of CEOs to earn undisclosed profits is better explained under a managerial power approach then from an optimal contacting perspective).
78. Id.
79. Id.
80. This may be true even if the board knows that insiders earn illegal profits on average, since it may not know the profits made by particular executives and therefore may set the wage too high.
81. Bebchuk et al., supra note 70, at 831.
82. This may be based on particular facts or based on a general average.
83. This argument is consistent with the view that information about a firm is the property of the firm, and that insiders should not be able to appropriate it for their use without the consent of the board.
84. For a useful summary of the debate, see BEWLEY, supra note 40, at 1-2 (concluding based on interviews with over 300 business and labor figures that firms balance gains from reduced wages against decreases in morale that might result).
85. See Henderson et al., supra note 4, at 43 tbl.2 (reporting supporting data on insiderspecific statistics).
86. $8 million times 12% equals $960, 000.
87. See supra Table 5.
88. See SEYHUN, supra note 19, at 5.
89. Chiarella v. United States, 445 U. S. 222, 228-31 (1980) (summarizing settled case law on question of company insiders trading on material informational advantage).
90. Id. at 228-29
91. (quoting Speed v. Transamerka Corp., 99 F. Supp. 808, 829 (D. Del. 1951)).
92. Another argument for regulation is firm based. Firms might prefer broad insider trading laws because of the inability of firms to commit to not trade on inside information. As a result of this inability, all firms would face higher costs of capital, including firms whose insiders were not trading.
93. See United States v. O'Hagan, 521 U. S. 642, 653 (1997) (stating that a person commits fraud under the misappropriation theory by his "undisclosed, self-serving use of a principal's information" when he owes a fiduciary duty to that principal) (emphasis added).
94. Id. at 652.
95. Id. at 659 n. 9.
96. This assumes the Court did not require disclosure to take advantage of this defense. In the Rule 10b5-1 context, the firm need not disclose. But as HJM show, the firms that do not disclose generally do not see insiders earn abnormal returns. See Henderson et al., supra note 4, at 23-27 (discussing correlation between insiders' returns from trading and the level of insiders' disclosure).
97. For an argument that diversification cannot account for this risk for investors, see Alicia J. Davis, Are Investors' Gains and Losses from Securities Fraud Equal Over Time? Theory and Evidence 4-6 (Univ. of Mich. Law & Econ., Olin Working Paper No. 09-002, 2010), available at http://ssrn. com/abstract=1121198.
98. See, e.g., MANNE, supra note 5, at 61 (stating that outsiders benefit from insider trading since "the market value of the shares held by people who may know nothing whatever of the new information has increased" as a result of the trading);
99. Stephen M. Bainbridge, Insider Trading, in 3 ENCYCLOPEDIA OF LAW AND ECONOMICS 772, 774 (Boudewijn Bouckaert & Gerritt De Geest eds., 2000) (noting that "the market reacts to the insiders' trades and gradually moves toward the correct price").
100. Henderson et al., supra note 4, at 39 fig.2.
101. There may be some threshold confidence level (of certainty times magnitude) below which entering into a plan is too costly, but given the ease and low cost of entering into these plans, this must be very, very low.
102. See Henderson et al., supra note 4, at 39 fig.2 (showing a strong correlation between Cumulative Abnormal Returns and specific-disclosure sales).
103. See Henderson et al., supra note 4, at 27 (summarizing empirical analyses and noting that "trading plans may actually enhance insiders' strategic trade potential because of the reduced litigation risk").
104. Executive Compensation and Related Person Disclosure, Securities Act Release No. 33-8732A, Exchange Act Release No. 34-54.302A, Investment Company Act Release No. IC-27.444A, 71 Fed. Reg. 53, 158, 53, 158. (Nov. 7, 2006), available at http://www.sec.gov/rules/final/2006/33-8732a.pdf.
105. 17 C. F. R. § 229. 402 (c) (2010).
106. See FIN. STANDARDS ACCOUNTING BD., SUMMARY OF STATEMENT NO. 123 (rev. 2004), available at http://www.fasb.org/summary/stsuml23r.shtml. Rule 123R allows firms discretion to choose among a variety of valuation methodologies, but the Black-Scholes vale is the most common.
107. If Firm A, Firm B, and Firm C all report compensation in cash and securities to their CEOs of $1 million, but do not publicly disclose the trading restrictions they impose, we cannot know the CEOs' expected compensation. We can reasonably assume that the ability to trade is worth something to executives, either based simply on liquidity or the ability to time trades based on private information. So, if Firm C imposes no restrictions on trading, Firm B imposes standard blackout windows, and Firm A bans all trades, holding all else equal, we can predict that the CEO of Firm C will be able to earn more compensation in a given year than the CEO of Firm B, who will in turn be able to earn more compensation in a given year than the CEO of Firm A. But if this is not disclosed, the market cannot know this.
108. BEBCHUK & FRIED, supra note 11, at 8.
109. These expected values are simply the stock price for each state of the world times the probability of that state of the world (that is, for Project A: 70% x $15 + 30% x $8 = $13; and for Project B: 70% x $20 + 30% x $0 = $14).
110. For Project A, Sue earns $500 in the good state of the world ($15-$10 x 100 shares), which occurs 70% of the time; and -$200 in the bad state of the world ($8-$10 x 100 shares), which occurs 30% of the time. The sum of these is expected values is $290 ($350 + (-$60) = $290). For Project B, Sue earns $1000 in the good state of the world ($20-$10 x 100 shares), which occurs 70% of the time; and -$1000 in the bad state of the world ($0-$10 x 100 shares), which occurs 30% of the time. The sum of these expected values is $400 ($700 + (-$300) = $400).
111. For Project A, Sue earns the same as in the pre-Rule case for the good state of the world ($350), but in the bad state of the world, she can avoid the $60 in expected losses by planning trades at $10 per share in advance, and letting them execute when the probabilities resolve themselves (but are not publicly disclosed) in a negative way. Sue's expected value from Project A is thus $350. For Project B, Sue earns the same as in the pre-Rule case for the good state of the world ($1000), but in the bad state of the world, she can avoid the $300 in expected losses through insider insurance. Sue's expected value from Project B is thus $700.
112. A related point is that insiders may be less likely to surreptitiously unwind equity incentives they have been given through derivatives or other hedging transactions. For these reasons, many of which will be firm-specific, any general reform may be overinclusive and destructive of social welfare.
113. In unreported regressions, the independent variable "percent insider ownership" is statistically insignificant.
114. See Ronald J. Gilson & Reinier Kraakman, Investment Companies as Guardian Shareholders: The Place of the MSIC in the Corporate Governance Debate, 45 STAN. L. REV. 985, 1006-09 (1993) (reporting empirical studies purporting to show "mildly positive relationship between active large-block shareholders and corporate performance").
115. In unreported regressions, the independent variable "percent institutional ownership" is statistically insignificant.