The Relation between Corporate Governance and CEOs’ Equity Grants
Lawrence D. Brown
GeorgiaStateUniversity
Email:
Yen-Jung Lee
NationalTaiwanUniversity
Email: ylee@ ntu.edu.tw
September2009
Abstract:We investigate whether the firm’s corporate governance affects the value of equity grants for its CEO. Consistent with the managerial power view, we find thatmore poorly governed firms grant higher values of stock options and restricted stock totheir CEOsafter controlling for the economic determinants of these grants.We show that the negative relation between governance strength and equity grants cannot be attributed to omitted economic factors, the substitution effect between governance strength and equity incentives, or the divergence between the subjective value and the fair value of equity grants. As further evidence consistent with the managerial power view, we show that firms with poorer governance in the pre-Enron era cut back more on the use of employee stock options (ESOs) for their CEOs in the post-Enron era, a period when CEO compensation received intensified public scrutiny and when the transparency and accounting cost of ESOs increased.
Key words:corporate governance, executive compensation, stock options, SFAS 123R
JEL Classifications: G30, G34, J31, J33
We thank Ashiq Ali, Eli Bartov, Dan Collins, Christi Gleason, Paul Healy, Tom Lys, Maria Nondorf, Shiva Rajgopal, Doug Skinner, Abbie Smith, Franco Wong, and workshop participants at the 2007 American Accounting Association Annual Meetings, 2007 Harvard University IMO conference, 2007 University of Okalahoma Accounting Research Conference, Georgia State University, National Taiwan University, University of Iowa, and the University of Texas at Dallas for helpful comments.
The Relation between Corporate Governance and CEOs’ Equity Grants
ABSTRACT
We investigate whether the firm’s corporate governance affects the value of equity grants for its CEO. Consistent with the managerial power view, we find thatmore poorly governed firms grant higher values of stock options and restricted stock totheir CEOsafter controlling for the economic determinants of these grants.We show that the negative relation between governance strength and equity grants cannot be attributed to omitted economic factors, the substitution effect between governance strength and equity incentives, or the divergence between the subjective value and the fair value of equity grants. As further evidence consistent with the managerial power view, we show that firms with poorer governance in the pre-Enron era cut back more on the use of employee stock options (ESOs) for their CEOs in the post-Enron era, a period when CEO compensation received intensified public scrutiny and when the transparency and accounting cost of ESOs increased.
Key words:corporate governance, executive compensation, stock options, SFAS 123R
JEL Classifications: G30, G34, J31, J33
1. Introduction
We investigate the effect of a firm’s corporate governance on its CEO’s equity grants, defined as employee stock options (ESOs) and restricted stock.Corporate governance is a setof mechanisms that helpsmitigate agency problemsbetween managers and shareholders. Theoretically, equity grants can be considered an aspect of corporate governance because it ties managers’ personal wealth to their firms’ stock price performance (Shleifer and Vishny 1997; Core et al. 2003), reducing the possibility that managers takesuboptimal actions to harmshareholder value. However, a number of practitioners, institutional investors, shareholder activists, and academic researchers contend that instead of beingan effectivegovernance mechanism, the option granting practiceis a manifestation of poor corporate governance (e.g., Biggs 2002; Gillan 2005; Guay et al. 2003;Coombes 2005).[1]
While there is evidence that poor governance is associated with excessive cash and total compensation (Core et al. 1999), empirical evidence on the relation between corporate governance and employee equity grants is limited.[2] Unlike cash, stock options and restricted stock are providedto employees for reasons besides justconveying compensation.[3]Both survey and archival evidence suggests thatretainingexisting employees and attracting new employees are the most important reasons for implementing equity grant programs (Ittner et al. 2003 and Oyer and Schaefer 2005).Moreover, researchers demonstrate that stock options and restricted stock are inefficient forms of compensation when compared to cash, raising the question about the relative importance of equity grants’ compensation versus their attracting, retaining and other economic roles.[4]Thus prior research’s cash compensation resultscannot be readily generalized to equity grants.We fill this gap in the literature by examining how the CEO’s equity grants relates to the firm’s governance structure and exploring alternative explanations for the observed relation.
The extant executive compensation literature offers two distinct views on the relation between corporate governance and the CEO’s equity awards: efficient contracting and managerial power. The efficient contracting view contends that boards grant optimalequity incentives that maximize firm value(e.g., Himmelberg et al. 1999; and Core and Guay 1999). According to the efficient contracting view,there is nosystematic relation between corporate governance and the value ofequity grants for top executivesafter controlling for the economic determinants of equity grants.The managerial power view contends that CEOs exert substantial influence over corporate boards and structure compensation contracts to benefit themselves rather than outside shareholders (e.g., Shleifer and Vishny 1997; Murphy 1999; Bebchuk and Fried 2003). As such, weaker governance gives CEOs relatively more power vis-à-vis the board, allowing themto receive more compensation than justified economically. The managerial power view predicts a negative association between governance strength and the value of the CEO’s equity grants.
To determinewhich view is more descriptive of reality, we utilize 8,084 observations representing 1,719 firms from 1998 to 2006with necessary compensation data from ExecuComp and governance data from RiskMetrics to examine the relation between firms’ corporate governance and the grant day fair values of new equity grants for their CEOs.Because governance mechanisms do not operate independently to resolve agency problems, we consider jointly the structuring of internal and external governance. We use Core et al.’s (1999)12 board and ownership structure variables to measure internal monitoring and Gompers et al.’s (2003) G-index to capture external monitoring. We use two methods, anequally weighted sum of the standardized governance variables and a factor analysis on all governance measures, to collapse the 13 governance variables into one parsimonious measure of overall governance strength.
Consistent with the managerial power view, we find a negative relation betweengovernance strength and thevalue ofequity grants. This negative relation, however, is consistent with two alternative explanations: (1) omitted economic factors affecting equilibrium equity compensation(hereafter, omitted economic factors explanation);and (2) substitution effects between governance mechanisms and equity incentives(hereafter, substitution explanation). To test for validity of the omitted economic factors explanation, we examine how the portion of the CEO’s equity grantsexplained by governance variables affects future firmperformance. Under the managerial power view, governance-motivated equity grantsreflect unresolved agency conflicts, which should not have a favorable effect on future firm performance.In contrast, the omitted economic factors explanation suggests that the portion of equity grantsexplained by governance variablesproxies for the demand for a higher level of equilibrium equity compensation, which should not impair future firm performance. Regressing accounting earnings and abnormal stock returnsoverthe three subsequent yearson governance- and economic-explained equity grant values, we find that governance-explained (economic-explained) equitygrant valueis negatively (positively) associated with future accounting and stock return performance,inconsistent with the omitted economic factors explanation.
Given that equity grantsare used to align managers’ and shareholders’ interests, firms are likely to substitute equity incentives for active monitoring in cases where direct monitoring by outside directors is less cost-beneficial. To explore thesubstitution explanation, we sort all observations into 22=4 portfolios based on overall governance strength and abnormal equity grants, which wedefine as the component of equity grants not explained by economic determinants of these grants.[5] If the substitution explanation is valid, firms with weak governance and high abnormal equity grantsshould outperform firms with weak governance and low abnormal equitygrants. We examinefuture accounting and stock return performance for each of the 4 portfolios and do not find evidence consistent with the substitution explanation.
One obstacle hindering the interpretation of our results is that both corporate governance and equity grants are endogenously determined by such factors asmonitoring costs, firm performance, and growth opportunities. Comparing compensation decisions between firms with good and bad governance may capture the effects of differences in the environments wherein firms operate rather than the effects of their governance.However, finding a good instrument for governance that is exogenous in the equity grant decision is extremely difficult so we provide two other analyses to mitigate this endogeneity concern.
Our first approachfollowsBebchuk et al. (2008) and Dittmar and Mahrt-Smith (2007) wherein we repeat our analysis retaining firms from the second half of our sample period (2003-2006) and replace the governance variables by theirbeginning-of-the-sample-period values (i.e., 1998 values). Our rationale is that since governance structurechanges slowly over time, the initial value of governance structure is less affected by future business and operating environments, making it more exogenous to future compensation decisions than is contemporaneous governance structure.[6]Consistent with our primary analysis, the resultsreveal that past corporate governance has a significantly negative effect on future abnormal equity grants.
As a second way to addressendogeneity concerns and to add validity to our managerial power interpretation, we examine changes in CEOs’ESO grant values in the post-Enron era, a period characterized by an increased public attention to executive compensation, various regulatory actions such as the Sarbanes-Oxley Act of 2002and listing requirements at thestock exchanges, and issuance of Statement of Financial Accounting Standards 123R (SFAS 123R) - Share-Based Payment (FASB 2004) that requires that ESOs be expensed on income statements.
Prior to SFAS 123R, firms were not required to record any ESO expense as long as the exercise price of their ESO was not below the grant day market price of the underlying stock. Not surprisingly, virtually all ESOs are granted “at-the-money”- that is, their exercise price is set to the grant-date market price (Murphy1999). ESOs’ zero accounting costprovides firms with opportunities to extract excessive compensation via ESOs.Under the managerial power view,CEOs of more poorly-governed firms are more likely to take advantage ofthe lack of ESO expensing and award themselves more ESOs than could be justified by economic factors.
The mandatory expensing requirement along with theincreased public scrutiny of executive compensation increased the outrage andaccounting costs of ESOs, forcing firms that used “too many” ESOsin the pre-Enronperiod to scale back their ESO use in the post-Enron era.[7] We show that firms with weaker corporate governance in the pre-Enron period cut back more on the use of ESOs post-Enron after controlling for changes in governance structures and other economic determinants of equity grants during the period. This evidence is consistent with CEOs exercising power over poor governance structure to grant themselvesexcessive ESOsin the pre-Enron period, providing further evidence on the managerial power view.
Wecontribute to the literature along several important dimensions. First, we add to the literature examining the relation between corporate governanceandagency problems. While prior studies show that managers exercise power to their own benefits (e.g., expropriating funds; empire building; consuming perquisites), direct evidence relating corporate governance to equity grants is limited. Using proprietary data for 205 firms from 1982-1984, Core et al. (1999) find that poorly-governed firms award their CEOs with more cash and total compensation.We complement and extend Core et al. (1999) by focusing on the equity component of the CEO compensation and by using a broader sample and data from a more recent period when ESOs become a major component of executive compensation.[8]
Second, we extend the literatureexamining substitution effects among governance mechanisms. Sundaramurthy et al. (1997) find that equity market investors react less negativelyto antitakeover provisions adopted by firms whose CEOs do not chair their boards than to anti-takeover provisions adopted by firms whose CEOs do chair their boards, consistent with the market perceiving stronger board monitoring substitutes for weak monitoring from the takeover market. Gillan et al. (2006) find firms with more independent boards have a higher G-Index, suggesting that the market for corporate control plays a less important monitoring role when the firm has a more powerful board. We extend this literature by showing that firms do not use incentivecompensation to substitute for weak governance. In fact, our evidence is inconsistent with incentive compensation serving as an effective governance mechanism.
Third, we contribute to the literature examining CEOs’ opportunistic behavior with respect to stock option grants. Yermack (1997), Aboody and Kasznik (2000), and Lie (2005) show that managers time corporate disclosures or information flows around ESO grants to increase the value of theirstock option grants. Choudhary et al. (2009) reported that some firms accelerated the vesting of outstanding ESOs to reduce accounting expenses that would need to be reported under SFAS 123R. Our findings suggest weak corporate governance allowed CEOs to take advantage of opaque ESO accounting to enrich themselves via ESOs.The rest of our paper proceeds as follows. Section 2 presentsour research design. Section 3 describes our data and reports the descriptive statistics. We show results of our multivariate analyses in section 4 and of additional analyses in section 5. Section 6 concludes.
2. Research Design
We estimate the following Tobit model on a pooled cross-sectional and time-series basis to assess the relation between governance strength and equity grants for the CEO:[9]
(1)
where EQUITY is the logarithm of (1+new equity grantsto the CEO for the year). New equity grantsto the CEO is calculated as the sum of the grant day fair value of new ESOs as computed by Execucomp plus the grant day fair value of restricted stock.[10]
GOVSCORE and GOVFACTOR are measures of governance strength. Corporate governance mechanisms used to mitigate agency conflicts between shareholders and managers can be either internal (i.e., the board of directors, insider ownership, and blockholders) or external (e.g., the market for corporate control). We usetakeover vulnerability,measured as a transformation of Gompers et al.’s (2003) G-Index, to proxy for external governance strength, and use board monitoring and ownership structure, measured following Core et al. (1999), to proxy for internal governance strength. G-Index is based on 24 anti-takeover provisions from RiskMetrics and is determined by adding one point for each anti-takeover provision the firm has in place and zero otherwise. A higher G-index imposes higher costs on takeover activities. We transform G-Index to takeover vulnerability using a linear transformation, denoted VULNERABILITY=24G-Index so that a higher value of VULNERABILITY indicates stronger external governance.[11]
Core et al. (1999) used the following 12 variables to measure internal governance strength: (1) CEOCHAIR, an indicator variable equal to one if the board chair is the firm’s CEO and zero otherwise; (2) BOARDSIZE, total number of board directors; (3) INSIDEDIR, percent of the board who are managers, retired managers, or relatives of currentmanagers; (4) HIREDBYCEO, number of outside directors appointed by the CEO, scaled by board size; (5) GRAYDIR, number of outside directors who are former employees or whose employershave a financial relationship with the company, scaled by board size; (6) INTERLOCKDIR, number of outside directors who are interlocked (a director isinterlocked if an inside officer of the firm serves on the board of an outside director’scompany), scaled by board size; (7) OLDDIR, number of outside directors over age 69, scaled by the number of outsidedirectors; and (8) BUSYDIR, number of outside directors who serve on three or more other boards (six or more for retired outside directors), scaled by the number of outside directors.(9) CEOHOLDING, percent of the firm’s outstanding shares owned by the CEO; (10) NONCEOOWN5, indicator variable set equal to one if at least one inside board member other than the CEO owns 5% or more of the firm’s outstanding shares, and zerootherwise; (11)OUTDIROWN, percent of the firm’s outstanding shares owned by outside directors divided bythe number of outside directors; and (12) BLOCKHOLDER, indicator variable equal to one if the firm has an external blockholderwho owns at least 5% of the its outstanding shares, and zero otherwise.
Because different governance mechanisms interact to achieve overall monitoring effectiveness, we use an equally weighted sum of the standardized internal and external governance variables, denoted GOVSCORE, to capture overall governance strength. Consistent with Core et al. (1999), we expect that higher values of internal governance variables (1) to (8) provide more power to the CEO in its relationship with the board, and thus implya weaker governance structure.Higher values of internal governance variables (9) to (12) indicate stronger non-CEO insider or outside shareholder monitoring, and thereforesuggesta stronger governance structure. Variables (1) to (8) are multiplied by negative one before entering into the calculation of GOVSCORE. A higher value of GOVSCORE indicates stronger corporate governance.
One limitation of GOVSCORE is the need to form expectations about the direction that each governance variable affects governance strength. As anotherway to measure overall governance strength, we perform a factor analysis onVULNERABILITY and the 12internal governance variables.[12]The factor analysis does not rely on a prioriexpectations about the relation between each governance variable and overall governance strength.