Executive Compensation in Widely-Held US Firms ESNIE 2007 Jesse Fried Boalt Hall School of Law U.C. Berkeley

Overview of Presentation Why study U.S. CEO pay? 2 conflicting views classical “optimal contracting” “managerial power” approach Managerial power approach: in depth Sources of power “Outrage Constraint” & “Camouflage” Costs to shareholders Policy implications

A picture is worth 1000 words

Why study U.S. CEO pay? Economic importance Amounts Incentive effects Managerial effort Decision-making Window into functioning of U.S. corporate governance system generally Theoretical interest Setting for testing agency/governance/labor market theories Good data (public firms) Interaction among economics/social norms /political & legal institutions Public fascination

Two views of CEO pay Optimal contracting approach E.g., Murphy (1999); Core, Guay & Larcker (2001); Gabaix & Landier (2007); Kaplan (2007) Managerial power approach E.g. Bebchuk & Fried (2002,03,04,05); Yermack (1997); Bertrand & Mullainathan (2001); Blanchard & Lopez-de-Silanes & Shleifer (1994)

Shared premise of both views Managerial agency problem Managers of widely-held public firms have significant power Berle & Means (1932) Jensen & Meckling (1976): “agency problem” Managers use power to benefit selves Empire building (Jensen, 1974; Williamson, 1964) Failure to distribute excess cash (Jensen, 1986) Entrenchment (Shleifer & Vishny, 1989)

Optimal K Approach Classical financial economic view Executive pay is remedy to agency problem Boards design pay scheme to Compensate and retain executives Incentivize managers to increase shareholder value Main flaw: due to political limitations on pay amounts, CEOs pay may be insufficiently high powered Jensen & Murphy, 1990; Kaplan 2006

Managerial Power Approach Executive compensation is potential remedy to managerial agency problem But it is also part of the agency problem itself Managers use their positional power to get pay excessive too decoupled from own performance weakens incentives to generate shareholder value perverts incentives Arrangements deviate from optimal K

Managerial power approach Sources of Managerial Power “Outrage constraint” “Camouflage” Pay Distortions Going forward: what should be done

Sources of Managerial Power (1) Optimal K assumes arm’s length bargaining b/w board & CEO But why? if they assume executives not “hard-wired” to serve shareholders, why should they presume directors will automatically seek to do so?

Sources of Managerial Power (2) CEOs have power over directors Economic Incentives directorship: $200K, perks, prestige, connections Until now, CEOs control renomination to board Social factors Collegiality Loyalty Cognitive dissonance (directors are current/former executives) Personal costs of favoring executives are small

Result of Managerial Power Managers want pay that is Higher More decoupled from performance (easier to get) Boards routinely approve executive pay deals that do not serve shareholders Pay likely too high Pay decoupled from performance Dilutes incentives Distort incentives

Only a slight exaggeration…

Evidence of power-pay effects CEO pay higher, less performance-based when board weaker Larger board, more “independent” directors appt’d by CEO, directors serving on multiple boards, CEO is board chair (Core, Holthausen & Larcker, 1999) no large outside shareholder eg Lambert, Leicker, Weigelt 1993 fewer “pressure-resistant” institutional shareholders David, Kochar, Levitas 1998 more anti-takeover provisions Borokhovich, Brunarski, Parrino 1997

Constraints on Managerial Power? (1) Corporate law? State corporate law defers to board compensation decisions under “business judgment rule” (e.g. Disney) Takeover market discipline? Staggered boards Courts allow “poison pills” Hostile takeovers expensive, rare

Constraints on Power? (2) Election of new directors? Corporation sends out proxy materials with names of board nominees Check “yes” or “withhold” Send proxy back to company to be voted “yes” or “withhold” Unless competing proxy, 1 “yes” vote gets director elected under “plurality voting rule” Don’t need approval of majority of votes, just plurality No competing proxies Costs of mailing competing proxy high Managers won’t release shareholder list Collective action problem Result: 99% elections uncontested

“Outrage Constraint” Outrage Constraint Boards’ main constraint: adverse publicity and “outrage” Outrage imposes social and economic costs embarrassment shareholders more likely to support (rare) challenge to management Evidence of publicity’s effect Thomas & Martin (1999) Dyck & Zingales (2004) Wu (2004)

“Camouflage” Fear of outrage leads firms to “camouflage” pay Pay designers try to obscure and legitimize amount of pay performance-insensitivity

Camouflage pre-1992: An SEC official describes the pre-1992 state of affairs as follows: “The information [in the executive compensation section] was wholly unintelligible.... Depending on the company’s attitude toward disclosure, you might get reference to a $3,500,081 pay package spelled out rather than in numbers. ………. Someone once gave a series of institutional investor analysts a proxy statement and asked them to compute the compensation received by the executives covered in the proxy statement. No two analysts came up with the same number. The numbers that were calculated varied widely.”[1][1] [1].Linda C. Quinn, Executive Compensation under the New SEC Disclosure Requirements, 63 U. Cin. L. Rev. 769, (1995).

1992: Summary Pay Table (SEC) Firms required to clearly report most forms of compensation in tables with dollar amounts Salary Bonus Stock options (number) Long-term incentive compensation Comp table became focus of Media, economists, shareholders

Post 1992 Camouflage Pay designers began relying heavily on forms of compensation not reportable in any column in comp table post-exit payments (e.g. pensions, golden parachutes) low-interest loans (Worldcom: $400 million) performance-insensitive compensation that can be reported as something other than “salary” E.g.: “guaranteed bonus”

Other CEO pay distortions (1) Non-equity pay weakly linked to performance often driven by luck (e.g. oil company earnings) bonuses have low “goalposts” often tied to manipulable metrics (accounting earnings)

Other CEO pay distortions (2) Equity pay Option plans fail to filter out windfalls Most stock price increases do not reflect Firm-specific factor CEO’s contribution Firms could use market/sector-based indexing but don’t Backdating accentuates windfalls Few restrictions on unwinding Managers not required to hold shares (diluting incentives) Can sell on inside information (perverting incentives)

Costs to shareholders Direct Top-5 pay = 10% of aggregate corporate earnings during Bebchuk & Grinstein (2005) up from 5% during Indirect Perverted incentives, e.g Size justifies pay: incentive to acquire Manipulate earnings to sell at high price Fannie Mae spent $1 billion cleaning up accounting

Going Forward: What Should Be Done ? (1) Transparency Outrage constraint currently main check on managerial power Constraint depends on transparency SEC must track efforts by pay designers to get around new disclosure rules

2006 Disclosure Rules (SEC) Improved summary table reporting Annual change in actuarial value of pension Total amount More detail More transparent reporting of Outstanding equity Post retirement payouts More detailed rationale for pay package Result: harder to camouflage compensation

What should be done (2) Increase shareholder power Problem: managerial power Must counterbalance with more shareholder power Should make it easier to replace directors SEC could make companies turn proxy material into corporate ballot with both management and shareholder candidates (like political election) Dramatically lower cost so shareholders can cheaply replace bad directors

Some setbacks 2003: SEC chair supports “shareholder access” to proxy statement Business execs pressure White House, SEC chair resigns But fight is not over Pressure many companies to adopt “majority vote” for individual directors So shareholders can “punish” individual directors by withholding votes Hedge funds becoming active

The future Further empowering shareholders best hope for improving Executive compensation US corporate governance generally THE END

Boards behaving better CEO pay increases moderating Kaplan, 2007 CEO turnover increasing Kaplan & Minton, 2007

Congress 1993: Tax: Section 162(m) 2002: Sarbox Prohibition on loans Clawback

1993: IRC Section 162(m) Outcry in early 1990s that pay decoupled from performance Congress: “Non-performance” pay over $1m not deductible by company At-the-money options qualify as performance pay Problem: does not address managerial power Some managers continue to get more than $1m salary Who is hurt?

Unintended effect of 162(m) Signals acceptability of Salary up to $1 million Large option grants (Congress deems it “performance comp”) Used to justify total pay increase Below $1 million salaries rise to $1 million Option grants skyrocket Bull market turns options into windfalls Huge increase in actual non-performance pay

Congress: SarbOx 2002 Prohibition on loans Outrage over huge, hidden low-cost loans 1920s proposal to ban loans resurrected Effect: disrupts efficient contracting Clawback provision Return bonuses, stock proceeds Following earnings misstatement caused by “misconduct” Shareholder-serving boards should have done this on their own Not yet applied

End

Pay without Performance Jesse Fried March 7, 2006 Berkeley For fuller exposition of views on the subject: Pay without Performance (Harvard University Press, 2004)

Going Forward: Making Directors More Accountable to Shareholders We should make it easier for shareholders to replace directors E.g., giving shareholders access to corporate ballot would reduce costs of challenging current board Not a panacea – still collective action problem but increasing probability of shareholder revolt will improve incentives

Making Directors More Accountable By making boards accountable to shareholders and attentive to their interests, such reform would: Make reality more like official story of arm’s length negotiations Improve executive compensation arrangements Improve corporate governance more generally

Decoupling Pay from Performance (1) Rise in executive compensation has been justified as necessary to strengthen incentives Financial economists have applauded: Shareholders should care more about incentives than about the amount paid executives. “It’s not how much you pay, but how” (Jensen & Murphy, 1990) Institutional investors have accepted higher pay as price of improving managers’ incentives

Decoupling Pay and performance (2) But the devil is in the details: managers’ compensation is less linked to performance than is commonly appreciated. Managers’ own performance does not explain much of the cross-sectional variation in managers’ compensation. Firms could have generated the same increase in incentives at much lower cost, or used the same amount to generate stronger incentives

Decoupling Pay from performance (3) Factors contributing to the weak link between pay and managers’ own performance: (1) The historically weak link between bonus payments and long-term stock returns. (2) The large amounts given through performance- insensitive retirement benefits. (3) The large fraction of gains from equity-based compensation resulting from market-wide and industry-wide movements.

Decoupling Pay from Performance (4) (4) Practices of “back-door re-pricing” and reload options that enable gains even when long-term stock returns are flat. (5) Executives’ broad freedom to unload vested options/restricted stock. (6) “Soft landing” arrangements for pushed out executives that reduce the payoff differences between good performance and failure. And more …

Paying for performance (1) Reduce windfalls from equity-based compensation:  Filter out some or all of the gains resulting from market-wide or sector-wide movements.  Can be done in various ways; indexing is only one option.  Move to restricted stock increases windfalls – restricted stock is an option with an exercise price of zero.

Paying for performance (2) Reduce windfalls from bonus compensation:  Filter out some or all of the improvements in accounting performance resulting from market-wide or sector-wide movements. [E.g., look at increase in earnings relative to peers.]

Paying for performance (3) Tie equity-based compensation to long-term values:  Separate vesting and freedom to unload: require holding for several years after vesting (even until/after retirement).  Prohibit contractually any hedging or other scheme that effectively unloads some of the exposure to firm returns.  Limit the ability of serving executives to time sales.

Paying for performance (4) Tie the performance-based component of non- equity compensation to long-term values:  Assuming it is desirable to link pay to improvement in some accounting measures, don’t link to short-term (e.g., annual) changes – can lead to gaming and distortions or at least to decoupling of pay from long- term changes in value.  Claw-back provisions that reverse payments made on the basis of restated financial figures: “if it wasn’t earned it must be returned.”

Paying for Performance (5)  Rethink termination arrangements: Current arrangements provide “soft landing” in any termination that is not for fault, defined extremely narrowly. This is costly – reduces the payoff difference between good and poor performance. Consider: -- Broadening the definition of “for cause” termination -- Making the severance payment depend on the performance during the executive’s service.

 Easy for companies to fix:  Company should include in annual proxy statement:  increase in value of retirement entitlement from last year and its current value Improving Transparency

Improving transparency (2) Another important instance of opaqueness: deferred compensation arrangements. Benefit executives by providing tax-free buildup of investment gains. Outsiders cannot make even a rough approximation of value Firms can easily make these benefits transparent.

Board Accountability  Recent reforms emphasize strengthening director independence from executives. Strengthened independence is beneficial but it is an insufficient foundation for board accountability. For each company, vast number of individuals could be considered “independent directors.” Two key questions (1) Who is selected from this vast pool? (2) What will their incentives be once appointed? Strengthened independence eliminates some people from pool, reduces bad incentives for those appointed. But does not fully answer (1) and (2)

Board Accountability (2) We should make directors not only more independent of executives, but also make them more accountable to shareholders What we need is reduced insulation from shareholders. Can be done in a way that does not provide distraction and short-terms focus

Improving Board Accountability Make shareholder power to remove directors real (even if weak). Election reform: (1) Adopt procedure for shareholder nomination of directors (2) Provide company reimbursement for shareholders whose nominees receive sufficient shareholder support.

Improving board accountability (2)  Remove charter-based staggered boards, which prevent shareholders from ever replacing a majority of the directors in one vote. Evidence staggered boards are associated with 4-5% lower firm value [ Bebchuk and Cohen, The Costs of Entrenched Boards, JFE, 2005]

Conclusion There is much that can be done – and should be done – to link pay more closely to performance.