Is There a Link Between Corporate Governance and CEOs ...
The Relation between Corporate Governance and CEOs’ Equity Grants
Lawrence D. Brown
Georgia State University
Email: ldb@gsu.edu
Yen-Jung Lee
National Taiwan University
Email: ylee@ ntu.edu.tw
September 2009
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 that more poorly governed firms grant higher values of stock options and restricted stock to their CEOs after 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 that more poorly governed firms grant higher values of stock options and restricted stock to their CEOs after 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 set of mechanisms that helps mitigate agency problems between 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 take suboptimal actions to harm shareholder value. However, a number of practitioners, institutional investors, shareholder activists, and academic researchers contend that instead of being an effective governance mechanism, the option granting practice is 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 provided to employees for reasons besides just conveying compensation.[3] Both survey and archival evidence suggests that retaining existing 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 results cannot 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 optimal equity incentives that maximize firm value (e.g., Himmelberg et al. 1999; and Core and Guay 1999). According to the efficient contracting view, there is no systematic relation between corporate governance and the value of equity grants for top executives after 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 them to 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 determine which view is more descriptive of reality, we utilize 8,084 observations representing 1,719 firms from 1998 to 2006 with 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, an equally 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 between governance strength and the value of equity 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 grants explained by governance variables affects future firm performance. Under the managerial power view, governance-motivated equity grants reflect 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 grants explained by governance variables proxies for the demand for a higher level of equilibrium equity compensation, which should not impair future firm performance. Regressing accounting earnings and abnormal stock returns over the three subsequent years on governance- and economic-explained equity grant values, we find that governance-explained (economic-explained) equity grant value is negatively (positively) associated with future accounting and stock return performance, inconsistent with the omitted economic factors explanation.
Given that equity grants are 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 the substitution explanation, we sort all observations into 2(2=4 portfolios based on overall governance strength and abnormal equity grants, which we define 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 grants should outperform firms with weak governance and low abnormal equity grants. We examine future 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 as monitoring 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 approach follows Bebchuk 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 their beginning-of-the-sample-period values (i.e., 1998 values). Our rationale is that since governance structure changes 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 results reveal that past corporate governance has a significantly negative effect on future abnormal equity grants.
As a second way to address endogeneity 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 2002 and listing requirements at the stock 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 (Murphy 1999). ESOs’ zero accounting cost provides 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 of the lack of ESO expensing and award themselves more ESOs than could be justified by economic factors.
The mandatory expensing requirement along with the increased public scrutiny of executive compensation increased the outrage and accounting costs of ESOs, forcing firms that used “too many” ESOs in the pre-Enron period 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 themselves excessive ESOs in the pre-Enron period, providing further evidence on the managerial power view.
We contribute to the literature along several important dimensions. First, we add to the literature examining the relation between corporate governance and agency 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 literature examining substitution effects among governance mechanisms. Sundaramurthy et al. (1997) find that equity market investors react less negatively to 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 incentive compensation 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 their stock 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 presents our 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]
[pic] (1)
where EQUITY is the logarithm of (1 + new equity grants to the CEO for the year). New equity grants to 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 use takeover 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=24(G-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 current managers; (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 employers have a financial relationship with the company, scaled by board size; (6) INTERLOCKDIR, number of outside directors who are interlocked (a director is interlocked if an inside officer of the firm serves on the board of an outside director’s company), scaled by board size; (7) OLDDIR, number of outside directors over age 69, scaled by the number of outside directors; 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 zero otherwise; (11) OUTDIROWN, percent of the firm’s outstanding shares owned by outside directors divided by the number of outside directors; and (12) BLOCKHOLDER, indicator variable equal to one if the firm has an external blockholder who 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 imply a weaker governance structure. Higher values of internal governance variables (9) to (12) indicate stronger non-CEO insider or outside shareholder monitoring, and therefore suggest a 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 another way to measure overall governance strength, we perform a factor analysis on VULNERABILITY and the 12 internal governance variables.[12] The factor analysis does not rely on a priori expectations about the relation between each governance variable and overall governance strength.
The factor analysis is conducted on a pooled cross-sectional and time-series basis. Two factors result from the factor analysis. The first factor explains 16% of total variation in the 13 governance variables, with loadings for OUTDIROWN, CEOHOLDING, INSIDEDIR, NONCEOOWN5, VULNERABILITY, and BOARDSIZE of 0.87, 0.70, 0.56, 0.48, 0.35, and -0.33, respectively. The second factor explains 6% of the total variation and loads on CEOCHAIR and HIREDBYCEO, with loadings of -0.45 and -0.36, respectively. These two factors appear to capture intensity of board monitoring and the inverse of CEO power respectively. We add the two standardized factor scores together to create a governance strength factor, denoted GOVFACTOR.[13] A larger GOVFACTOR indicates stronger corporate governance.[14]
To isolate the abnormal portion of total equity grants, we control for the economic determinants of equity use in executive compensation packages in equation (1). Core and Guay (1999) show that firms grant fewer equity incentives to their CEOs when their CEOs’ pre-existing equity incentives exceed the optimal incentive level. Accordingly, we include the deviation of the CEO’s pre-existing equity incentives from the optimal incentive level, denoted INCENRESID, at the beginning of the year in equation (1). Following Core and Guay (1999), INCENRESID is the residual from a regression of the CEO’s equity incentives derived from stock option and restricted stock holdings on firm size, monitoring difficulty, growth opportunities, CEO tenure, and the free cash flow problem as described in the Appendix. We expect a negative coefficient on INCENRESID.
We control for the logarithm of total sales, denoted SALES, because Smith and Watts (1992) suggest that the level of CEO compensation is positively associated with firm sales. Core and Guay (1999) show that firms use more equity incentives to mitigate agency problems when monitoring employee efforts is costly. Following Core and Guay (1999), we presume shareholders face greater monitoring costs for firms with greater growth opportunities (Smith and Watts 1992) as proxied by a lower book to market ratio, denoted BM. BM is measured as the book value of assets divided by the market value of assets, and is used to proxy for growth opportunities, where market value of assets is the sum of book value of debt and market value of common equity. We expect a positive coefficient on SALE and a negative coefficient on BM.
Prior studies contend that firms bear tax costs when substituting ESOs for other forms of compensation because there is no tax deduction for incentive ESOs and the tax deduction for nonqualified ESOs is deferred until options are exercised. Therefore, substituting ESOs for other forms of compensation is costlier for firms with higher marginal tax rates. We use NOL to proxy for firms’ marginal tax rates. NOL is an indicator variable set equal to one if the firm has net operating loss carry-forwards in any of the previous three years, and zero otherwise. A positive NOL implies a lower effective tax rate so we expect the coefficient on NOL to be positive. Because granting ESOs requires no cash outlays, firms facing liquidity problems are more likely to use options in lieu of cash compensation to conserve cash. Following Core and Guay (1999), we include SHORTFALL, calculated as the three-year average of cash flow used in investing activities plus common and preferred dividends minus cash flow from operations, all deflated by total assets. The coefficient on SHORTFALL is expected to be positive.
Dechow et al. (1996) and Core and Guay (1999) suggest ESO use is positively related to the extent to which lack of retained earnings constrains a firm’s ability to pay dividends and repurchase shares. We follow Dechow et al. (1996) and Core and Guay (1999) and measure the dividend constraint using an indicator variable, DIV_CONSTRAIN, set equal to one if [(retained earnings + cash dividends + stock repurchases)/the prior year’s cash dividends and stock repurchases] is less than 2.0 in any of the previous three years, and zero otherwise. DIV_CONSTRAIN is also set equal to one if the denominator is zero for all three years. We expect the coefficient on DIV_CONSTRAIN to be positive. Core and Guay (1999) argue that the level of executive compensation is an increasing function of the firm’s stock price performance. We include the prior year and the current year stock returns (LAGRET and RET respectively) to control for the link between stock performance and CEO compensation and expect the coefficients on lagged RET and RET to be positive.
In addition to the incentive, tax, and liquidity considerations, stock options and restricted stock can be used to convey compensation. We control for the factors that have been shown by Core et al. (1999) to be associated with total compensation, including the firm’s investment opportunities (INVOPP), defined as the firm’s year-end book-to-market ratio averaged over the five-year period ended the current year multiplied by negative one. ROA is income before extraordinary items after ESO expense divided by lagged total assets.[15] STDROA is the standard deviation of ROA for the five-year period ending the current year. STDRET is the standard deviation of RET for the five-year period ending the current year.
We include NONEQCOMP, measured as the logarithm of the CEO’s total compensation minus the grant day fair value of stock options and restricted stock. On the one hand, more talented managers could receive greater total and equity grants, implying a positive association between NONEQCOMP and EQUITY. On the other hand, if different forms of compensation substitute for each other, granting non-equity compensation could reduce the need for granting stock options and restricted stock, implying a negative association between EQUITY and NONEQCOMP. Hence, we make no prediction about the sign of NONEQCOMP. Industry dummies are based on the 48 Fama and French (1997) industry classifications and year dummies are created for the years included in our sample period. We use Petersen’s (2009) double cluster procedure to allow inter-correlations of residuals across firms and across time. The expected signs of all variables are presented in Table 4.
3. Sample and descriptive statistics
Table 1 presents our sample selection procedures. Our initial sample begins with 14,056 firm-years between 1998 and 2006 that are covered by ExecuComp under the ExecuComp 1992 reporting format. We obtain compensation data from ExecuComp, corporate governance and G-Index data from RiskMetrics, stock price data from CRSP, financial data from Compustat, and institutional ownership data from Thomson Financial. Our sample period starts from 1998 because the director ownership data, which is necessary for constructing ownership structure variables, became available on RiskMetrics in 1998. The sample period ends in 2006 because ExecuComp changed its reporting format in the 2006 fiscal year and the new reporting format discontinues the reporting of several data items necessary to estimate equation (1).[16] [17]
After excluding firms without necessary data, we are left with 8,084 firm-years from 1998-2006. Table 2 reports descriptive statistics for selected variables of interest. Panel A provides an overview of the CEO’s total compensation and compensation components. On average, our sample firm-years granted 3.5 (1.6) million dollars of equity (cash) compensation to their CEOs, with a median of 1.3 (1.1) million. The mean grant-day fair value of stock options is 2.8 million, about four times as large as the grant day fair value of restricted stock. The average (median) CEO receives 5.6 (2.9) million dollars of total compensation. Equity (cash) compensation accounts for 44% (48%) of total compensation for our average sample firm. About 17.29% of our sample firm-years do not grant any stock options or restricted stock. For firm-years that do grant equity-based compensation, the mean equity-based (cash) compensation granted to the CEO is about 53% (39%) of the CEO’s total compensation.
Panel B of Table 2 reports the descriptive statistics on corporate governance variables. About 68% of the CEOs in our sample firm-years serve as chairmen of their firms’ boards. The average board consists of 9.33 directors of which 26% are insiders, 37% are hired by the CEO, 7% are gray, and 1% is interlocked. On average, 12% of outside directors are over age 69 and 12% are classified as busy directors. The average CEO owns 3% of the firm’s outstanding equity, with a median of 1%. The mean ownership per outside director is 2% of the firm’s outstanding shares. About 86% of the sample firms have at least one external shareholder and 12% have at least one internal board member other than the CEO that owns at least 5% of the firm’s outstanding shares. The mean (median) firm has 9.24 (9) anti-takeover provisions.
Panel C of Table 2 describes variables used in the regression analyses. To mitigate the undue influence of outliers, we winsorize all continuous regression variables at the 1% and 99% levels. The median logarithm of equity grants is 7.14, about the same size as the median logarithm of non-equity compensation. The distribution of incentive residuals is skewed slightly to the left, with a mean (median) value of 0.11 (0.08). Although the mean of the incentive residuals from estimating equation (2) should be zero by construction, the mean incentive residual reported in Panel C of Table 2 is not. This is because we estimate the portfolio incentive equation (i.e., equation (2)) using all ExecuComp observations with necessary data and we lose some observations in Table 2 because of our sample selection procedures described in Table 1.
Table 3 reports Pearson and Spearman correlations for selected regression variables. Consistent with the managerial power view, Pearson correlations show that EQUITY are negatively correlated with GOVSCORE and GOVFACTOR, indicating that better-governed firms provide their CEOs with smaller values of stock options and restricted stock. Consistent with firms using new equity grants to manage optimal equity incentives, EQUITY is negatively correlated with lagged INCENRESID. Larger firms, firms with more growth opportunities, firms with net operating loss carry-forwards, and firms with more volatile operating income grant more options and restricted stock to their CEOs. Inconsistent with the notion that stock-based compensation substitutes for other forms of compensation, EQUITY is positively correlated with NONEQCOMP, suggesting that CEOs with more equity awards also receive higher non-equity compensation.
4. Multivariate Analyses
4.1 The association between governance strength and ESO grant values
Panel A of Table 4 presents results from estimating equation (1). Columns 2 and 3 present the expected signs under the efficient contracting and managerial power views respectively. Models (1) and (2) report results when governance strength is measured by GOVSCORE and GOVFACTOR respectively. Consistent with the managerial power view, the coefficients on GOVSCORE and GOVFACTOR are significantly negative, suggesting that firms with stronger governance grant less abnormal equity grants. The coefficient on the incentive residual is significantly negative, suggesting that companies grant less equity-based compensation when the CEO’s existing stock and option holdings provide incentives exceeding economically-justified levels. Most control variables have their expected signs and are significant. More specifically, equity grants are significantly larger for bigger firms, firms which are more cash-and-dividend-constrained, and firms with better growth opportunities and lower marginal tax rates. Equity grants are also positively associated with stock price performance.[18]
To assess the economic significance of governance strength on equity-based compensation, we calculate the marginal effect of all continuous (indicator) independent variables as the percentage change in the dollar amount of equity-based compensation for a change in the independent variable from the 25th percentile to the 75th percentile of the variable in the sample (from 0 to 1), holding other independent variables at their mean values. The marginal effect of GOVSCORE (GOVFACTOR) is -0.398 (-0.544), indicating that the CEO receives 39.8% (54.4%) less equity awards when GOVSCORE (GOVFACTOR) improves from the 25th percentile to the 75th percentile. GOVSCORE and GOVFACTOR are among the most economically significant variables in the model. The only two more influential determinants of equity grants are firm size and non-equity based compensation, with marginal effects of around 70% and 50%, respectively.
For completeness, we replace GOVSCORE and GOVFACTOR with the 13 governance variables used to construct GOVSCORE and GOVFACTOR. We report these results in Panel B of Table 4. Model (1) presents results for equation (1) when only internal governance mechanisms are included. Consistent with expectations, we find firms with the following governance characteristics provide their CEOs with less abnormal grants: (1) fewer outside busy directors; (2) CEO owns a higher percentage of the company’s outstanding shares; (3) at least one non-CEO insider owns 5% of the firm’s outstanding shares, and (4) higher percentage ownership per outside director. Consistent with Core et al. (1999), we find that CEOs of firms with a higher percentage of inside directors receive less abnormal equity grants. Inconsistent with our prediction that older directors are less effective monitors, we find firms with more outside directors that are over age 69 grant less abnormal equity grants.
4.2 Alternative Explanations
We interpret our results in Section 4.1 as weaker corporate governance enabling managers to exercise influence over boards and extract rents in the form of excess equity grants. The negative association between governance variables and the amount of equity grants is also consistent with two other explanations. First, governance variables could proxy for some aspects of economic determinants of equity grants not modeled or fully captured in equation (1), such as the CEO’s job complexity or the CEO’s quality. Second, they could reflect substitution between governance structure and incentive compensation. We next perform tests to determine the descriptive validity of these two plausible explanations.
4.2.1 Omitted Economic Factors
To test for the omitted economic factors explanation, we examine how the component of equity grants arising from governance factors relates to future performance. Bartov and Mohanram (2004) show that excessive ESOs provide perverse incentives for managers to inflate short-term earnings and stock prices to increase cash payouts of ESO exercises. Bebchuk and Fried (2003) argue that the largest cost arising from managers' ability to influence the boards’ compensation decisions is firm value loss due to suboptimal incentives rather than the rents captured by managers. If the association between equity grants and governance strength reflects managerial power, we expect a zero or negative association between governance-explained equity grants and subsequent firm performance.[19] In contrast, if governance structure captures omitted economic factors that affect the equilibrium level of equity grants, we expect a favorable effect of governance-predicted equity grants on subsequent firm performance.
We decompose equity grants into its three components – the portion of equity grants that is explained by (1) governance structure (EQUITY_GOV); (2) economic factors (EQUITY_ECO); and (3) neither economic nor governance factors (EQUITY_RESID), and we estimate the following equation:
[pic] (2)
where FUTABROA is industry-adjusted return on assets, calculated as income before extraordinary items after ESO expense summed over the subsequent three years scaled by total assets at the end of the current year, denoted FUTROA, minus the industry median FUTROA, where the industry assignments are based on the 48 Fama-French industry classifications.[20] FUTABRET is future abnormal stock return performance, measured as the intercept coefficient, or alpha, from the firm-specific Carhart’s (1997) four factor model estimated over the three-year period following the end of the current year as follows:
[pic] (3)
where R is firm return minus risk-free rate, RMRF is the value-weighted market return minus the risk-free rate, and the terms SMB (small minus big), HML (high minus low), and Momentum are the monthly returns on zero-investment factor-mimicking portfolios capturing size, book-to-market, and momentum effects, respectively.
EQUITY_GOV and EQUITY_ECO are the predicted components of equity grants arising from governance and economic factors, respectively.[21] Specifically, we compute predicted equity grants due to governance factors for each firm-year, after controlling for the economic determinants of ESO grants as follows:
[pic] (4)
where (1 is the estimated coefficient on GOVSCORE or GOVFACTOR reported in model (1) or model (2) in Panel A of Table 4.
We compute the predicted component of equity grants that is related to economic determinants for each firm-year, after controlling for the corporate governance factors as follows:
[pic] (5)
where (j is the estimated coefficient on each economic determinant j reported in either model (1) or model (2) of Panel A in Table 4. The residual equity grant value explained by neither economic nor governance factors, denoted EQUITY_RESID, is calculated as EQUITY minus the sum of EQUITY_GOV and EQUITY_ECO. Given that EQUITY_GOV, EQUITY_ECO, and EQUITY_RESID are predicted values based on equation (1), the standard errors from the regular OLS for equation (2) are invalid. We use the bootstrap procedure described by Cameron et al. (2007) to calculate bootstrapped clustered standard errors in both the firm and time dimension.
Barber and Lyon (1996) argue that it is important to control for past performance when testing the association between corporate events and future firm performance because earnings are serially correlated and corporate events (granting ESOs in our case) may be correlated with recent firm performance. Therefore, we include current operating and abnormal stock return performance – ABROA or ABRET – in equation (2), where ABROA is ROA minus the median ROA for all firms in the same Fama-French 48 industry group and ABRET is the intercept coefficient from the firm-specific Carhart’s (1997) four factor model estimated over the 12-month period ending the current year end. We control for firm size (SIZE), investment opportunities (INVOPP), earnings volatility (STDROA), stock return volatility (STDRET) and firm leverage (LEVERAGE) in the operating performance equation, where LEVERAGE is total debt divided by total assets and SIZE, INVOPP, STDROA, and STDRET are as defined in Section 2.
Panel A (Panel B) of Table 5 presents the results for equation (2) when governance strength is measured using GOVSCORE (GOVFACTOR). Models 1 and 2 (3 and 4) report the results when future performance is measured by FUTABROA (FUTABRET). The coefficient on EQUITY_GOV (EQUITY_ECO) is significantly negative (positive) across both panels, consistent with the managerial power view but inconsistent with the omitted economic factor explanation.
4.2.2 Substitution
The negative association between governance structure and abnormal equity grants is also consistent with the substitution explanation, which suggests a tradeoff between governance monitoring and incentive pay. Gillan et al. (2006) show that board structures and anti-takeover provisions substitute for each other. The authors also contend that boards of directors monitor less in riskier/noisier environments because it is more difficult for outsiders to determine appropriateness of managers’ actions. In cases when active monitoring by shareholders is costly or ineffective, firms may substitute equity incentives such as ESOs for active monitoring. If the substitution explanation is valid, firms with weak governance structure and more equity incentives should outperform firms with weak governance structure and less equity incentives. Similarly, firms with low equity incentives and stronger governance structure should outperform firms with low equity incentives and weaker governance structure.
To test for the substitution explanation, we examine how overall governance strength and incentive compensation interact to affect future firm performance. We individually classify all firms into 2(2=4 portfolios based on governance strength and excessive equity grants.[22] We define excessive (or abnorm) equity grants, denoted EXCESS_EQUITY, as the portion of equity grants not explained by economic determinants and industry dummies specified in equation (1).[23] Specifically, EXCESS_EQUITY is the residual from the following regression:
[pic] (6)
All variables are as defined previously. Panel A of Table 6 presents the mean and median FUTABROA for each of the 4 portfolios. Holding overall governance strength constant at the weak level, firms with the high EXCESS_EQUITY report a mean (median) FUTABROA of 0.1243 (0.0570), which is significantly lower than 0.1828 (0.0834) reported by firms with the lowest EXCESS_EQUITY, suggesting that firms granting large EXCESS_EQUITY does not outperform firms granting small EXCESS_EQUITY, inconsistent with the substitution explanation. Keeping the overall excessive compensation at the low level, the mean FUTABROA for firms with strong corporate governance is significantly higher than that for firms with weak corporate governance. However, strong corporate governance is associated with a higher future operating performance regardless of the level of EXCESS_EQUITY. Similar results can be obtain from Panel B of Table 6 where future performance is measured by abnormal stock return relative to the Carhart’s (1997) four factor model. Taken together, our results suggest that the substitution effect is not a valid explanation for the negative association between governance strength and ESO grant values.[24]
5. Additional Analyses
5.1 Endogeneity
The negative association between corporate governance and abnormal equity grants could be driven by unmodeled or imperfectly measured factors that affect both governance structure and equity grants. Ideally, we would adopt an instrumental variable approach to help mitigate endogeneity effects. But the choice of appropriate instruments, while never easy, is very challenging in our context. Prior studies (Fama and Jensen 1983; Demsetz and Lehn 1985; Coles et al. 2008) show a firm’s governance structure is affected by its growth potential, operational complexity, recent performance, and uncertainty in its business and information environments. Virtually all of these factors have been shown to influence a firm’s decision to use equity grants (e.g., Core et al. 1999; Ittner et al. 2003), making it extremely difficult to find instruments that identify governance strength in the equity grants equation. To mitigate endogeneity concerns, we perform two additional tests to supplement our primary analyses in Section 4.
First, following Bebchuk, Cohen, and Ferrell (2008) and Dittmar and Mahrt-Smith (2007), we repeat the analysis in Panel A of Table 4 retaining only observations from the second half of our sample period (2003-2006) and replacing GOVSCORE and GOVFACTOR by their 1998 (beginning of our sample period) values, denoted GOVSCORE98 and GOVFACTOR98 respectively.[25] The results confirm our previous findings: the coefficients on GOVSCORE98 and GOVFACTOR98 remain significantly negative, suggesting that governance structures in earlier years affect the compensation decisions in later years. The control variables behave similarly to those reported in Panel A of Table 4.
As a second way to address endogeneity concerns, we examine ESO changes post versus pre Enron. Following Enron and other major U.S. corporate scandals, public attitude turned against ESOs because excessive option grants were blamed for creating perverse incentives for managers to commit accounting frauds. The post-Enron era was characterized by increased media attention on executive compensation, institutional shareholder activism, and various corporate reforms such as Sarbanes-Oxley Act of 2002 and new listing requirements at stock exchanges. In addition, the FASB passed SFAS 123R on December 16, 2004, requiring firms to recognize a compensation expense equal to the grant-day fair value of ESOs on their income statements.
Prior to SFAS 123R, GAAP allowed firms to avoid any compensation expense if they granted ESOs with an exercise price not lower than the grant-date market price of the underlying stock (known as the intrinsic value method) as long as they disclosed information about the grant date fair value of ESO grants in the financial statement footnotes.[26] ESOs bore zero accounting cost as long as they were not granted in the money. Bodie et al. (2003) argued that failure to recognize ESO expense provided an “accounting subsidy to stock options”, encouraging firms to use them to excess because virtually all other forms of compensation required expensing. Making compensation decisions based on ESOs’ zero accounting cost rather than their true economic costs is costly to shareholders because it results in compensation arrangements that provide worse incentives than those provided by arm's length contracts.
The managerial power view predicts that firms with poorer corporate governance are more likely to exercise power over weak corporate governance and extract excessive compensation in the form of ESOs. The intensified public scrutiny and the mandatory ESO expensing requirement in the post-Enron period increased the visibility of ESOs and drew more attention to the real cost of stock options, making it harder for firms to justify excessive executive pay (Guay et al. 2003). If poor corporate governance facilitated the excessive ESO use, one would expect to observe a greater decrease in ESO use in the post-Enron period for firms with poorer corporate governance. To test this prediction, we estimated the following OLS regression:
[pic] (7)
The dependent variable, DIFFOPTION, is the average OPTION over the post-Enron period (2002-2006) minus the average OPTION over the pre-Enron period (1998-2001), where OPTION is the logarithm of (1+grant day fair value of ESOs for the CEO).[27] A smaller DIFFOPTION indicates a greater decrease in ESO grant values in the post-Enron period. GOVSCORE9801 and GOVFACTOR9801 are the mean GOVSCORE and GOVFACTOR over the pre-Enron period (1998-2001), respectively. We include the change in the grant day fair value of restricted stock from the pre- to the post-Enron period, denoted DIFFRSTK, to control for the potential substitution between ESOs and restricted stock. DIFFRSTK is calculated as the average RSTK over the post-Enron period (2002-2006) minus the average RSTK over the pre-Enron period (1998-2001), where RSTK is the logarithm of (1+ grant day fair value of restricted stock for the CEO). DIFF_Y is average Y over 2002-2006 minus average Y over 1998-2001, where Y is one of the 14 independent variables in equation (1).
Table 7 shows the coefficients on GOVSCORE9801 and GOVFACTOR9801 are significantly positive, suggesting that firms with weaker (stronger) corporate governance cut back more (less) on ESO use after controlling for changes in determinants of ESOs including changes in governance structure from the pre- to the post-Enron period.[28] In sum, our evidence is consistent with the prediction of the managerial power view that more poorly governed firms took greater advantage of ESO accounting prior to SFAS 123R to award their CEOs with more ESOs than are justifiable economically. [29]
5.2 Subjective Valuation of Equity Grants
We now explore how the subjective value of equity grants affects our results. We define subjective value as valuation the executive places on ESOs and restricted stock she receives. Because employees are risk-averse and under-diversified, Lambert et al. (1991); Meulbroek (2001), Hall and Murphy (2002), and Ingersoll (2006) among others show that the lack of diversification makes the subjective value of incentive compensation such as stock options and restricted stock less than their fair market value. Anecdotal evidence suggests that employees value cash compensation much more highly than ESOs and restricted stock awards. For example, Hall and Murphy (2002, pgs.15-16) report that firms conducting explicit exchanges of cash for stock-based compensation typically gave participating executives a “risk premium” for accepting stock-based pay.[30]
If firms compensate their CEOs for the subjective valuation discount that employees apply to stock-based compensation by granting higher nominal values of equity grants, the coefficient estimate on corporate governance will be unreliable when the divergence between the fair value and the subjective value of equity grants is correlated with governance strength.[31] To examine how the subjective valuation of the equity grants (i.e., subjective value discount) affects our results, we adopt the Ingersoll (2006) model to estimate the subjective value of equity grants.[32] Ingersoll (2006) shows that the subjective value of options can be obtained via a downward adjustment to the Black-Scholes formula. The downward adjustment is a function of employee risk aversion, the idiosyncratic volatility of the stock, and the proportion of the executive wealth tied to the company’s stock. In Ingersoll (2006), the subjective value of an ESO (VIngersoll) to a risk adverse employee is given by the following equation:
[pic] (8)
[pic]
[pic]
[pic]
[pic]
where S is the price of the underlying stock on the grant day, q is the dividend yield on the stock, T is the expected life of the option, K is the exercise price of the employee stock option, r is the risk-free interest rate, N(.) is the cumulative normal distribution function, [pic]is the annualized total variance of the stock, σ is the expected annual stock price volatility, [pic]>0 is the relative risk aversion parameter, [pic]is the residual firm specific variance of the stock, and [pic]is the fraction of his own wealth an executive is constrained to hold in his firm’s stock beyond that represented in the market portfolio. Without loss of generality, we set the expected life equal to the maturity of the option grant. We obtain the risk-free rate, dividend yield, and total variance of the stock from Execucomp. The residual firm specific variance is the annualized variance of the residual from a firm-specific rolling 36-month market model regression.
The subjective value of a restricted stock (SIngersoll) is given by the following equation:
[pic] (9)
where S is the market price of restricted stock. We compute VIngersoll (SIngersoll) for each individual option (stock) grant. The subjective value of the CEO’s equity grants, denoted SUBJECTIVE, is calculated as the sum of VIngersoll and SIngersoll for all option and stock grants that the CEO received in a particular year.
Following Hall and Murphy (2002), we use relative risk aversion parameters of 2.0 and 3.0 which are at the low end of the reasonable range of estimates in the literature. We assume that 50% of the CEO’s wealth is constrained to hold the company’s stock beyond that represented in the market portfolio. We create SUB_FV, defined as SUBJECTIVE divided by the grant day fair value of the option and stock awards, to measure the subjective valuation discount. The higher the SUB_FV, the closer the CEO’s subjective valuation is to the fair value of his/her equity grants (and the lower the subjective valuation discount). We set SUB_FV equal to one if the firm did not grant any equity grants during the year. We re-estimate equation (1) including SUB_FV and expect the coefficient SUB_FV to be negative because the larger the subjective valuation discount (i.e., a smaller SUB_FV), the greater the nominal value of equity grants that the company has to provide in order to make up for the divergence between the subjective and objective values of the equity grants.
Table 8 reports results for a risk aversion parameter equal to 2.0.[33] In our sample, the mean subjective value of new ESO (restricted stock) grants is about 52% (90%) of their fair value. The mean subjective value of a dollar of total equity grants is about 52 cents (not tabulated). Consistent with our Table 4 results, the coefficients on GOVSCORE and GOVFACTOR are significantly negative. As expected, the coefficient on SUB_FV is significantly negative, suggesting companies pay greater equity grants when CEOs place a greater discount on their equity awards. In sum, our results indicate the negative relation between governance strength and abnormal equity grants cannot be attributed to the subjective value discount employees attach to stock-based compensation.
6. Concluding Remarks
We examine whether and why corporate governance affects the CEO’s equity grant values. Although it has often been argued that equity incentives are a useful component of a firm’s governance structure, empirical evidence as to whether it actually aligns managers’ and shareholders’ interests is lacking. We consider the two predominant views on the relation between governance and equity grants: efficient contracting and managerial power. The efficient contracting view maintains that boards always optimize their equity-granting decisions. Under this view, there is no systematic relation between equity grants and governance strength. In contrast, the managerial power view maintains that managers in firms of weak corporate governance exercise power over the boards’ compensation decisions to extract excessive compensation. Under this view, there should be a negative relation between governance strength and abnormal equity grants.
Our evidence, covering 8,048 firm-year observations from 1,719 firms over the years 1998-2006, suggests that weaker corporate governance as proxied by lower takeover vulnerability and weaker board and shareholder monitoring is associated with greater abnormal equity grants for the CEO after controlling for the economic determinants of equity grants. We measure takeover vulnerability (external governance) based on Gompers et al.’s (2003) G-Index and the intensity of board and shareholder monitoring (internal governance) using the 12 governance variables of Core et al. (1999). Although our principal finding of a negative association between ESOs and corporate governance is consistent with the managerial power view, it is also consistent with two alternative explanations: (1) corporate governance variables proxy for unmodeled (omitted) economic factors affecting ESO use; or (2) incentive compensation substitutes for governance mechanisms to mitigate agency problems. We perform additional tests to determine the validity of these alternative explanations for our findings.
To test for the omitted economic factors explanation, we decompose equity grants into three components – the portion of equity grants explained by (1) governance strength; (2) economic factors; (3) neither governance strength nor economic factors. Inconsistent with the omitted economic factors explanation but consistent with the managerial power view, the predicted component of ESO grant value attributable to governance factors is inversely related to future operating performance.
To examine the validity of the substitution explanation, we sort firms into 2(2=4 portfolios based on overall governance strength and the value of abnormal equity grants, and compared future operating and stock return performance across the four portfolios. In contrast to the prediction under the substitution explanation, we find that firms with larger excessive ESO grant values do not outperform firms with smaller excessive ESO grant values for portfolios with the weakest corporate governance.
We further validate our managerial power interpretation of the relation between governance strength and equity grants by showing that poorly-governed firms cut back more on ESO grant values in the post-Enron era, a period characterized by greater investors’ scrutiny of CEO compensation, major regulatory actions including the enactment of Sarbanes-Oxley Act of 2002, and adoption of a new accounting rule requiring firms to report the fair value of ESOs on their income statements as an expense. This finding is consistent with CEOs in poorly governed firms being more likely to abuse the favorable ESO accounting treatment to deliver stealth compensation in the form of ESOs in the pre-Enron period. The increase in public attention on executive compensation along with the mandatory ESO expensing requirement increased the transparency and accounting cost of ESOs, made it harder for poorly governed firms to justify excessive ESO packages in the post-Enron period. Consistent with the managerial power view, poorly-governed firms granted the most excessive compensation and cut back the most on ESO use in the post-Enron era.
We extend prior research examining the association between CEO compensation and corporate governance. Although Core et al. (1999) showed that poor governance contributes to excessive cash and total compensation, empirical evidence on how governance relates to equity grants is limited. It is inappropriate to infer the association between corporate governance and equity grants from Core et al. (1999) because cash compensation constitutes the majority of CEO compensation during Core et al.’s (1999) sample period (1982-1984), suggesting that Core et al.’s total compensation results could be driven by cash compensation. We contribute to the literature by explicitly examining alternative explanations for the negative association between governance strength and abnormal equity grants. We show that incentive compensation neither substitutes for weak governance nor is it an effective governance mechanism. We add to research that examines CEOs’ opportunistic behavior regarding stock option grants (Yermack, 1997; Aboody and Kasznik, 2000; Lie, 2005; Choudhary et al. 2006) by showing that weak corporate governance along with opaque ESO accounting prior to SFAS 123R enabled CEOs to abuse ESO grants for their own benefits.
Appendix: Estimation of the Deviation of the CEO’s Pre-existing Equity Incentives from the Optimal Equity Incentives
This appendix describes the estimation of the deviation of the CEO’s pre-existing equity incentives from the optimal equity incentives. Core and Guay (1999) show that firms use new option and restricted stock grants to manage the optimal level of equity incentives. In particular, they find that annual grants of new stock options and restricted stock to CEOs are negatively associated with CEOs’ incentive residual prior to the equity grants, where the incentive residual is the portion of pre-existing equity incentives that is in excess of the optimal incentive level. Following Core and Guay (1999), we model the optimal level of the CEOs' portfolio holdings of equity incentives as a function of firm size, monitoring difficulty, growth opportunities, CEO tenure, and the free cash flow problem.[34] The incentive residual, denoted INCENRESID, is the residual from the following regression model.
[pic] (10)
where EQPORT is the CEO's total portfolio incentives from options and restricted stock at the end of the year. The incentives provided by the CEO’s holding of restricted stock are estimated by multiplying the year-end stock value by 1%. We use the procedures outlined in Core and Guay (1999) to estimate the incentives provided by the CEO’s option holding. Core and Guay (1999) break the total option holding into those options granted in the current year and those options received from previous years. The incentives provided by newly granted options are estimated as the sensitivity of the Black-Scholes value of the new option grant to a 1% change in the year-end stock price. All the inputs required to calculate the Black-Scholes value – year-end stock price, option exercise price, time-to-maturity, stock return volatility, dividend yield, and the risk-free rate are available on Execucomp. The incentives provided by options granted in previous years are measured using weighted average exercise price and weighted average time-to-maturity as described in Core and Guay (1999).
MVE is logarithm of the market value of equity and is a proxy for firm size. IDRISK is logarithm of the idiosyncratic risk, measured as the standard deviation of the residual from a rolling 36-month market model regression and is used to measure the monitoring difficulty. BM is book value of assets divided by the market value of assets and is used to proxy for growth opportunities, where market value of assets is the sum of the book value of debt and the market value of common equity. TENURE is CEO tenure, measured as the number of years between the year when the CEO first took office and the current year-end. FCF is the degree of the free cash flow problem, measured as the three-year average of (operating cash flow - common and preferred dividends) / total assets if the firm has a book-to-market assets ratio greater than one (a proxy for low growth opportunities), and zero otherwise.
REFERENCES
Aboody, D., and R. Kasznik. 2000. CEO stock option awards and the timing of corporate voluntary disclosures. Journal of Accounting and Economics 29 (1): 73-100.
Adams, R. 2003. What do boards do? Evidence from board committee and director compensation data. Working paper, University of Queensland and European Corporate Governance Institute (ECGI).
Anderson, R., and J. Bizjak. 2003. An empirical examination of the role of the CEO and the compensation committee in structuring executive pay. Journal of Banking and Finance 27 (7): 1323-1348.
Barber, B., and J. Lyon. 1996. Detecting long-run abnormal operating performance: The empirical power and specification of test statistics. Journal of Financial Economics 41 (2): 359–399.
Bartov, E., and P. Mohanram. 2004. Private information, earnings manipulations, and executive stock-option exercises. Accounting Review 79 (4): 889-920.
Bebchuk, L., A. Cohen and Ferrell. 2008. What matters in corporate governance? Review of Financial Studies: 22(2): 783-827.
Bebchuk, L., and J. Fried. 2003. Executive compensation as an agency problem. Journal of Economic Perspectives 17 (3): 71-92.
Biggs, J. 2002. Testimony on Stock Options before the U.S. Senate Finance Committee. Available on
Bodie, Z., R. Kaplan, and R. Merton. 2003. For the last time: Stock options are an expense. Harvard Business Review 81 (3): 63-71.
Cameron, A. J. Gelbach, and D. Miller, 2008. Bootstrap-Based Improvements for Inference with Clustered Errors. The Review of Economics and Statistics 90 (3): 414-427.
Choudhary, P, S. Rajgopal, and M. Venkatachalam. 2009. Accelerated vesting of employee stock options in anticipation of FAS 123-R. Journal of Accounting Research 47 (1): 105-146.
Coles, J., N. Daniel, and L. Naveen. 2006. Managerial incentives and risk-taking. Journal of Financial Economics 79 (2): 431-468.
Coombes, A. 2005. Excess compensation: Institutional investors say U.S. execs are overpaid. MarketWatch (December. 13, 2005).
Core, J., and W. Guay. 1999. The use of equity grants to manage optimal equity incentive levels. Journal of Accounting and Economics 28 (2): 151-184.
Core, J., and W. Guay. 2003. When efficient contracts require risk-averse executives to hold equity: Implications for option valuation, for relative performance evaluation and for the corporate governance debate. Working Paper, University of Pennsylvania.
Core, J., W. Guay, and D. Larcker. 2003. Executive equity compensation and incentives: A survey. Economic Policy Review 9 (1): 191-212.
Core, J. W, Guay, and T. Rusticus. Does weak governance cause weak stock returns? An examination of firm operating performance and investors’ expectations. The Journal of Finance 61(2): 655-681.
Core, J., R. Holthausen, and D. Larcker. 1999. Corporate governance, chief executive officer compensation, and firm performance. Journal of Financial Economics 51(3): 371-406.
Dechow, P., A. Hutton, and R. Sloan. 1996. Economic consequences of accounting for stock-based compensation. Journal of Accounting Research 34 (1996): 1-20.
Demsetz, H., and K. Lehn. 1985. The structure of corporate ownership: Causes and consequences. Journal of Political Economy 93 (6): 1155-1177.
Dittmar, A. and J. Mahrt-Smith. 2007. Corporate governance and the value of cash holdings Journal of Financial Economics 83 (3): 599-634.
Fama, E. and K. French. 1997. Industry cost of equity. Journal of Financial Economics 43 (2): 153-193.
Fama, E., and M. Jensen. 1983. Separation of ownership and control. Journal of Law and Economics 26 (2): 301-325.
Financial Accounting Standards Board. 2004. Statement of Financial Accounting Standards No. 123R: Share-Based Payment. Norwalk, CT: FASB.
Gillan, S. 2005. Has pay for performance gone awry? Views from a corporate governance forum. Research Dialogue 68 (July).
Gillan, S., J. Hartzell, and L. Starks. 2006. Tradeoffs in corporate governance: Evidence from board structures and charter provisions.Working paper, Texas Tech University and University of Texas at Austin.
Gompers, P., J. Ishii, and A. Metrick. 2003. Corporate governance and equity prices, Quarterly Journal of Economics 118 (1): 107-155.
Guay, W., S. Kothari, and R. Sloan. 2003. Accounting for employee stock options. American Economic Review 93 (2): 405-409.
Hall, B., and J. Liebman. 1998. “Are CEOs really paid like bureaucrats?” Quarterly Journal of Economics 103 (3):653-91.
Hall, J., and K. Murphy (2002). Stock options for undiversified executives. Journal of Accounting and Economics 33: 3-42.
Hanlon, M., S. Rajgopal, and T. Shevlin. 2003. Are executive stock options associated with future earnings? Journal of Accounting and Economics 36 (1-3): 3-43.
Himmelberg, C., R. Hubbard, and D. Palia. 1999. Understanding the determinants of managerial ownership and the link between ownership and performance. Journal of Financial Economics 53 (3): 353-384.
Ingersoll, J. 2006. The subjective and objective evaluation of incentive stock options. Journal of Business 79(2): 453-487.
Ittner, C., R. Lambert, and D. Larcker. 2003. The structure and performance consequences of equity grants to employees of new economy firms. Journal of Accounting and Economics 34 (1-3): 89-127.
Jensen, M. 1986. Agency costs of free cash flow, corporate finance, and takeovers. American Economic Review 76: 323-329.
Jensen, M., and W. Meckling. 1976. Theory of the firm: managerial behavior, agency costs and ownership structure. Journal of Financial Economics 3 (4): 305-360.
Lambert, R., D. Larcker, and R. Verrecchia. 1991. Portfolio considerations in valuing executive compensation. Journal of Accounting Research 29 (1): 129-149.
Lie, E. 2005. On the timing of CEO stock option awards. Management Science 51(5): 802-812.
Meulbroek, L. 2001. The efficiency of equity-linked compensation: understanding the full cost of awarding executive stock options. Financial Management 30 (2): 5-44.
Milgrom, P., and J. Roberts (1992), Economics, Organization and Management, Englewood Cliffs: Prentice Hall.
Murphy, K.1999. Executive compensation. In: Orley, A., David, C. (Eds.), Handbook of Labor Economics. North-Holland, Amsterdam, pp. 2485–2563.
Petersen, M. 2009. Estimating standard errors in finance panel data sets: Comparing Approaches. Review of Financial Studies 22(1):435-480
Shleifer, A., and R. Vishny.1997. A survey of corporate governance, Journal of Finance 52 (2): 737-783.
Smith, C., and R. Watts, 1992. The investment opportunity set and corporate financing, dividend, and financing policies. Journal of Financial Economics 32 (3): 262-292.
Sundaramurthy, C., J.M. Mahoney and J. T. Mahoney. 1997. Board structure, antitakeover provisions, and stockholder wealth. Strategic Management Journal 18(3): 231–245.
Yermack, D. 1997. Good timing: CEO stock option awards and company news announcements. Journal of Finance 52 (2): 449-476.
Table 1
Sample Selection
| |# of firm-years |
|All firm-years reporting under the Execucomp 1992 reporting format between 1998 and 2006 with non-missing CEO |14,056 |
|compensation data | |
|Less: firm-years with missing information about the grant day fair value of stock options or restricted stock | |
|Less: firm-years without sufficient data on Compustat necessary to estimate the equity grants model |(107) |
|Less: firm-years without sufficient data on CRSP necessary to estimate the equity grant model | |
|Less: firm-years without sufficient data on RiskMetrics to construct governance variables |(3,451) |
| | |
| |(94) |
| | |
| |(2,320) |
|Sample firm-years |8,084 |
Table 2
Descriptive Statistics on Selected Variables
(N=8,084)
Penal A: CEO Compensation (Thousands of Dollars)
| |Mean |Std Dev |Median |Q1 |Q3 |
|(ESOt+RESTRICTEDt) |3,544.81 |13,158.6 |1,254.97 |271.209 |3,451.94 |
|ESOt |2,824.36 |1,0551.5 |915.450 |80.7060 |2,773.63 |
|RESTRICTEDt |720.449 |7,630.89 |0.0000 |0.0000 |203.000 |
|CASHt |1,579.86 |1,773.59 |1,112.51 |669.787 |1,906.55 |
|TOTCOMPt |5,649.69 |13,906.9 |2,916.24 |1,373.48 |6,003.77 |
|(ESOt+RESTRICTEDt)/TOTCOMPt |0.4419 |0.2896 |0.4721 |0.2080 |0.6741 |
|(for all sample firm-years) | | | | | |
|CASHt/TOTCOMPt |0.4752 |0.2737 |0.4246 |0.2623 |0.6546 |
|(for all sample firm-years) | | | | | |
|(ESOt+RESTRICTEDt)/TOTCOMPt |0.5332 |0.2284 |0.5437 |0.3642 |0.7108 |
|(for 6,686 firm-years that grant equity comp) | | | | | |
|CASHt/TOTCOMPt |0.3940 |0.2064 |0.3734 |0.2384 |0.5330 |
|(for 6,686 firm-years that grant equity comp) | | | | | |
This panel describes components of the CEO compensation. The sample covers 8,084 firm-years from 1998 to 2006 in which equity grants were provided in 6,686 firm-years. ESO is the grant day Black-Scholes value for stock option awards as computed by Execucomp. RESTRICTED is the grant day fair value of restricted stock awards. TOTCOMP is the total compensation, defined as the sum of base salary, bonus, grant-date Black-Scholes value of stock option awards, grant-date fair value of restricted stock awards, payouts under long-term incentive plans, other annual compensation, and all other compensation. CASH is cash compensation, calculated as the sum of salary and bonus.
Penal B: Governance Variables
| |Mean |Std Dev |Median |Q1 |Q3 |
|CEOCHAIR t-1 |0.6813 |0.4660 |1.0000 |0.0000 |1.0000 |
|BOARDSIZE t-1 |9.3340 |2.4881 |9.0000 |7.0000 |11.0000 |
|INSIDEDIR t-1 |0.2576 |0.1313 |0.2308 |0.1429 |0.3333 |
|HIREDBYCEO t-1 |0.3695 |0.2492 |0.3636 |0.1667 |0.5714 |
|GRAYDIR t-1 |0.0730 |0.1021 |0.0000 |0.0000 |0.1250 |
|INTERLOCKDIR t-1 |0.0100 |0.0276 |0.0000 |0.0000 |0.0000 |
|OLDDIR t-1 |0.1187 |0.1541 |0.0000 |0.0000 |0.2000 |
|BUSYDIR t-1 |0.1202 |0.1510 |0.0000 |0.0000 |0.2000 |
|CEOHOLDING t-1 |0.0335 |0.0623 |0.0116 |0.0045 |0.0292 |
|NONCEOOWN5 t-1 |0.1241 |0.3297 |0.0000 |0.0000 |0.0000 |
|OUTDIROWNt-1 |0.0163 |0.0276 |0.0054 |0.0017 |0.0171 |
|BLOCKHOLDER t-1 |0.8607 |0.3463 |1.0000 |1.0000 |1.0000 |
|VULNERABILITY |14.7636 |2.5781 |15.0000 |17.0000 |13.0000 |
|GOVSCORE t-1 |-0.1971 |3.9499 |-0.1903 |-2.6891 |2.1828 |
|GOVFACTORt-1 |-0.0097 |1.2986 |-0.3691 |-0.8786 |0.3944 |
This panel provides descriptive statistics of governance variables. The sample covers 8,084 firm-years from 1998 to 2006. CEOCHAIR: indicator variable set equal to one if the CEO is also chairman of the board, and zero otherwise. BOARDSIZE: number of directors on the board. INSIDEDIR: percent of the board who are managers, retired managers, or relatives of current managers. HIREDBYCEO: number of outside directors on the board appointed by the CEO as a percent of the board size. GRAYDIR: number of outside directors whose employers have a financial relationship with the company or who are former employees of the company, as a percent of board size. INTERLOCKDIR: number of outside directors who are interlocked (a director is interlocked if an inside officer of the firm serves on the board of an outside director’s company), as a percent of board size. OLDDIR: percent of outside directors who are over age 69. BUSYDIR: percent of outside directors who serve on three or more other boards (six or more for retired outside directors). CEOHOLDING: percent of outstanding shares owned by the CEO.
NONCEOOWN5: indicator variable set equal to one if the firm has an internal board member other than the CEO who owns at least 5% of the firm’s outstanding shares, and zero otherwise. OUTDIROWN: total percentage of outstanding shares owned by outside directors divided by the number of outside directors. BLOCKHOLDER: indicator variable set equal to one if the firm has an external blockholder who owns at least 5% of the outstanding shares, and zero otherwise. GOVSCORE: an equally weighted sum of the standardized internal and external governance variables. Since higher values of the first eight internal governance variables indicate a board structure that provides more power to the CEO in its relationship with the board, these governance variables are multiplied by negative one before entering into the calculation of GOVSCORE. GOVFACTOR is the sum of the two standardized factor scores resulting naturally from a factor analysis on all thirteen internal and external governance variables.
Penal C: Selected Regression Variables
| |Mean |Std Dev |Median |Q1 |Q3 |
|EQUITYt |6.1513 |3.0823 |7.1357 |5.6066 |8.1470 |
|NONEQCOMPt |7.1848 |1.0250 |7.1644 |6.6184 |7.7655 |
|INCENRESIDt-1 |0.1081 |0.9773 |0.0773 |-0.4889 |0.6707 |
|SALE t-1 |7.3360 |1.5027 |7.2308 |6.3163 |8.3392 |
|BM t-1 |0.6422 |0.2721 |0.6505 |0.4321 |0.8476 |
|NOL t-1 |0.3289 |0.4699 |0.0000 |0.0000 |1.0000 |
|SHORTFALL t-1 |0.0099 |0.0806 |0.0027 |-0.0339 |0.0434 |
|DIV_CONSTRAIN t-1 |0.3564 |0.4790 |0.0000 |0.0000 |1.0000 |
|RET t |0.1367 |0.4968 |0.0775 |-0.1594 |0.3312 |
|RET t-1 |0.1678 |0.5204 |0.1046 |-0.1452 |0.3663 |
|INVOPP t-1 |-0.6217 |0.2455 |-0.6280 |-0.4320 |-0.8159 |
|ROA t-1 |0.0387 |0.1592 |0.0450 |0.0123 |0.0873 |
|STDROA t-1 |0.0566 |0.0848 |0.0308 |0.0150 |0.0640 |
|STDRET t-1 |0.1277 |0.0584 |0.1128 |0.0860 |0.1561 |
|NONEQCOMP t-1 |7.1848 |1.0250 |7.1644 |6.6184 |7.7655 |
|FUTABROAt |0.1358 |0.3230 |0.1405 |0.0334 |0.2822 |
|FUTABRETt |0.0028 |0.0291 |0.0027 |-0.0092 |0.0145 |
This panel provides descriptive statistics on selected regression variables. The sample covers 8,084 firm-years from 1998 to 2006. To mitigate undue influence from outliers, all continuous variables are winsorized at the 1% and 99% levels. EQUITY is the logarithm of (1+new equity grants to the CEO for the year). New equity grants to the CEO is calculated as the sum of the grant day fair value of new ESOs for the CEO as computed by Execucomp plus the grant day fair value of restricted stock. INCENRESID is the residual from the following regression model.
[pic](10)
where EQPORT is the CEO's total portfolio incentives from stock and option holdings at the end of the year. The incentives provided by the CEO’s restricted stock holdings are estimated by multiplying year-end stock value by 1%. We use the procedures outlined in Core and Guay (1999) to estimate the incentives provided by the CEO’s option holding. Core and Guay (1999) break the total option holding into those options granted in the current year and those options received from the previous years. Incentives from newly granted options are estimated as the sensitivity of the Black-Scholes value of the new option grant to a 1% change in the year-end stock price. All the inputs required to calculate the Black-Scholes value - stock price, exercise price, time-to-maturity, stock return volatility, dividend yield, and the risk-free rate - are available on Execucomp. Incentives from options granted in previous years are measured using the weighted average exercise price and weighted average time-to-maturity as described in Core and Guay (1999). MVE is logarithm of the market value of equity. IDRISK is logarithm of the idiosyncratic risk measured as the standard deviation of the residual from a 36-month market model regression. BM is book value of assets divided by the market value of assets, where market value of assets is the sum of book value of debt and market value of common equity. TENURE is CEO tenure, measured as the number of years between the year that the CEO first took office and the current year-end. FCF is the degree of the free cash flow problem measured as the three-year average (operating cash flow - common and preferred dividends) / total assets if the firm has a book-to-market ratio greater than one (a proxy for low growth opportunities), and zero otherwise. SALE is the logarithm of sales. BM is the book to market ratio. NOL is an indicator variable set equal to one if the firm has net operating loss carry-forwards in any of the previous three years, and zero otherwise. SHORTFALL is cash flow shortfall, measured as the three-year average of cash flow used in investing activities plus common and preferred dividends minus cash flow from operations, all deflated by total assets. DIV_CONSTRAIN is an indicator variable set equal to one if [(retained earnings + cash dividends + stock repurchases)/the prior year’s cash dividends and stock repurchases] is less than 2.0 in any of the previous three years, and zero otherwise. DIV_CONSTRAIN is also set equal to one if the denominator is zero for all three years. RET is the current stock return. INVOPP is the firm’s investment opportunities, calculated as negative one multiplied by the firm’s year-end book-to-market ratio averaged over the five-year period ending the current year, where the market-to-book ratio is defined as the market value of equity plus the book value of debt divided by total assets. ROA is income before extraordinary items after ESO expense divided by lagged total assets. STDROA is the standard deviation of ROA for the five-year period ending the current year. STDRET is the standard deviation of RET for the five-year period ending the current year. FUTABROA is industry-adjusted operating performance, calculated as income before extraordinary items after ESO expense summed over the subsequent three years scaled by total assets at the end of the current year, denoted FUTROA, minus the industry median FUTROA, where the industry assignments are based on the 48 Fama-French industry classifications. FUTABRET is the intercept coefficient, or alpha, from the firm-specific Carhart’s (1997) four factor model estimated over the three-year period following the end of the current year:
[pic] (3)
where R is firm return minus risk-free rate, RMRF is the value-weighted market return minus the risk-free rate, and the terms SMB (small minus big), HML (high minus low), and Momentum are the monthly returns on zero-investment factor-mimicking portfolios capturing size, book-to-market, and momentum effects, respectively.
TABLE 3
Correlation Matrix on Selected Regression Variables
| |EQUITYt |GOVSCOREt-1 |GOVFACTORt-1 |INCEN_ |
| | | | |RESID t--1 |
| |Efficient |Managerial |Coefficient |Marginal |Coefficient |Marginal |
| |contracting |Power |(Standard Error) |Effect |(Standard Error) |Effect |
|Intercept |? |? |-0.016 | | |0.860 | | |
| | | |(0.985) | | |(0.799) | | |
|GOVSCOREt-1 |0 |- |-0.094 |*** |-0.398 | | | |
| | | |(0.012) | | | | | |
|GOVFACTORt-1 |0 |- | | | |-0.499 |*** |-0.544 |
| | | | | | |(0.038) | | |
|INCENRESIDt-1 |- |- |-0.242 |*** |-0.251 |-0.184 |*** |-0.191 |
| | | |(0.054) | | |(0.052) | | |
|SALEt-1 |+ |+ |0.416 |*** |0.727 |0.403 |*** |0.703 |
| | | |(0.044) | | |(0.041) | | |
|BMt-1 |- |- |-1.110 |*** |-0.390 |-1.130 |*** |-0.405 |
| | | |(0.136) | | |(0.143) | | |
|NOLt-1 |+ |+ |0.164 |** |0.133 |0.087 | |0.072 |
| | | |(0.076) | | |(0.072) | | |
|SHORTFALLt-1 |+ |+ |0.934 |** |0.062 |1.171 |*** |0.078 |
| | | |(0.458) | | |(0.446) | | |
|DIV_CONSTRAINt-1 |+ |+ |0.127 |* |0.104 |0.124 |* |0.095 |
| | | |(0.076) | | |(0.070) | | |
|RETt-1 |+ |+ |0.027 | |0.011 |0.044 | |0.018 |
| | | |(0.104) | | |(0.104) | | |
|RETt |+ |+ |0.320 |*** |0.134 |0.349 |*** |0.143 |
| | | |(0.031) | | |(0.030) | | |
|INVOPP t-1 |+ |+ |0.733 |*** |-0.251 |0.808 |*** |-0.268 |
| | | |(0.211) | | |(0.201) | | |
|ROA t-1 |+ |+ |0.020 | |0.002 |0.094 | |0.006 |
| | | |(0.113) | | |(0.102) | | |
|STDROA t-1 |? |? |1.421 |*** |0.059 |1.168 |*** |0.049 |
| | | |(0.405) | | |(0.400) | | |
|STDRETt-1 |? |? |1.912 | |0.116 |1.806 | |0.106 |
| | | |(1.340) | | |(1.780) | | |
|NONEQCOMPt |? |? |0.573 |*** |0.537 |0.528 |*** |0.496 |
| | | |(0.079) | | |(0.075) | | |
|Industry dummies | | |Yes | | |Yes | | |
|Year Dummies | | |Yes | | |Yes | | |
|N | | |8,084 | | |8,084 | | |
|Pseudo R2 | | |0.0334 | | |0.0375 | | |
*,**,*** significant at the 0.1, 0.05 and 0.01 levels (two-tailed).
This panel provides Tobit results from estimating equation (1). Standard errors are calculated based on the Petersen’s (2009) double cluster procedure to allow inter-correlations of residuals across firms or across time. All variables are as defined in Table 2. Coefficients on the industry and year dummies are suppressed for expositional convenience. Industry dummies are created based on the Fama-French 48 industry classifications and the year dummies are created for the years included in our sample period.
TABLE 4
Tobit Analysis of the Relation between Equity Grants and Corporate Governance
[pic] (1)’
Panel B: The relation between equity grant values and corporate governance with GOVINDEX replaced by five internal governance variables
|Variable |Expected Sign |Expected Sign |Model (1) |Model (2) |Model (3) |
|Control Variables |Efficient |Managerial Power|Coefficient |Coefficient |Coefficient |
| |contracting | |(Standard Error) |(Standard Error) |(Standard Error) |
|Intercept |? |? |1.079 | |-1.325 | |2.467 |*** |
| | | |(1.517) | |(1.152) | |(1.212) | |
|CEOCHAIRt-1 |0 |+ |0.062 | | | |0.062 | |
| | | |(0.073) | | | |(0.075) | |
|BOARDSIZEt-1 |0 |+ |0.009 | | | |0.011 | |
| | | |(0.018) | | | |(0.018) | |
|INSIDEDIRt-1 |0 |+ |-1.441 |*** | | |-1.471 |*** |
| | | |(0.355) | | | |(0.357) | |
|HIREDBYCEOt-1 |0 |+ |0.035 | | | |0.011 | |
| | | |(0.110) | | | |(0.111) | |
|GRAYDIRt-1 |0 |+ |-0.187 | | | |-0.263 | |
| | | |(0.521) | | | |(0.524) | |
|INTERLOCKDIRt-1 |0 |+ |0.239 | | | |0.216 | |
| | | |(1.635) | | | |(1.618) | |
|OLDDIRt-1 |0 |+ |-0.834 |*** | | |-0.834 |*** |
| | | |(0.247) | | | |(0.246) | |
|BUSYDIRt-1 |0 |+ |0.832 |*** | | |0.851 |*** |
| | | |(0.234) | | | |(0.236) | |
|CEOHOLDINGt-1 |0 |- |-11.532 |*** | | |-11.756 |*** |
| | | |(1.373) | | | |(1.406) | |
|NONCEOOWN5t-1 |0 |- |-0.402 |** | | |-0.405 |*** |
| | | |(0.154) | | | |(0.158) | |
|OUTDIROWNt-1 |0 |- |-9.109 |*** | | |-9.268 |** |
| | | |(3.936) | | | |(3.965) | |
|BLOCKHOLDERt-1 |0 |- |0.117 | | | |0.112 | |
| | | |(0.086) | | | |(0.087) | |
|VULNERABILITYt-1 |0 |- | | |-0.082 |*** |0.002 | |
| | | | | |(0.018) | |(0.016) | |
|INCENRESIDt-1 |- |- |0.186 |*** |-0.282 | |0.198 |*** |
| | | |(0.046) | |(0.056) |*** |(0.047) | |
|SALEt-1 |+ |+ |0.353 |*** |0.485 | |0.336 |*** |
| | | |(0.042) | |(0.043) |*** |(0.042) | |
|BMt-1 |- |- |-1.280 |*** |-1.093 | |-1.175 |*** |
| | | |(0.127) | |(0.138) |*** |(0.125) | |
|NOLt-1 |+ |+ |0.047 | |0.140 |* |0.073 | |
| | | |(0.065) | |(0.078) | |(0.065) | |
|SHORTFALLt-1 |+ |+ |0.515 | |1.243 |*** |0.934 |** |
| | | |(0.415) | |(0.467) | |(0.407) | |
|DIV_CONSTRAINt-1 |+ |+ |0.132 |** |0.145 |* |0.136 |** |
| | | |(0.062) | |(0.079) | |(0.064) | |
|RETt-1 |+ |+ |0.169 | | 0.020 | |0.169 | |
| | | |(0.101) | |(0.105) | |(0.101) | |
|RETt |+ |+ |0.408 |*** |0.315 |*** |0.389 |*** |
| | | |(0.024) | |(0.031) | |(0.025) | |
|INVOPP t-1 |+ |+ |0.576 |*** |0.838 |*** |0.692 |*** |
| | | |(0.176) | |(0.222) | |(0.176) | |
|ROA t-1 |+ |+ |0.118 |*** |-0.017 | |0.249 |** |
| | | |(0.039) | |(0.117) | |(0.098) | |
|STDROA t-1 |? |? |1.048 |*** |1.460 |*** |0.993 |*** |
| | | |(0.284) | |(0.424) | |(0.297) | |
|STDRETt-1 |? |? |2.763 |*** |1.801 | |2.597 |*** |
| | | |(0.760) | |(1.390) | |(0.782) | |
|NONEQCOMPt |? |? |0.439 |*** |0.604 |*** |0.457 |*** |
| | | |(0.061) | |(0.081) | |(0.063) | |
|Industry dummies | | |Yes | |Yes | |Yes | |
|Year Dummies | | |Yes | |Yes | |Yes | |
|N | | |8,084 | |8,084 | |8,084 | |
|Pseudo R2 | | |0.0446 | |0.0324 | |0.0452 | |
*,**,*** significant at the 0.1, 0.05 and 0.01 levels (two-tailed).
This panel provides Tobit results from estimating equation (1)’. Standard errors are calculated based on the Petersen’s (2009) double cluster procedure to allow inter-correlations of residuals across firms and across time. Coefficients on the industry and year dummies are suppressed for expositional convenience. GOV is one of the following thirteen internal or external governance mechanisms. CEOCHAIR: indicator variable set equal to one if the CEO is also chairman of the board, and zero otherwise. BOARDSIZE: number of directors on the board. INSIDEDIR: percent of the board who are managers, retired managers, or relatives of current managers. HIREDBYCEO: number of outside directors on the board appointed by the CEO as a percent of the board size. GRAYDIR: number of outside directors whose employers have a financial relationship with the company or who are former employees of the company, as a percent of board size. INTERLOCKDIR: number of outside directors who are interlocked (a director is interlocked if an inside officer of the firm serves on the board of an outside director’s company), as a percent of board size. OLDDIR: percent of outside directors who are over age 69. BUSYDIR: percent of outside directors who serve on three or more other boards (six or more for retired outside directors). CEOHOLDING: percent of outstanding shares owned by the CEO. NONCEOOWN5: indicator variable set equal to one if the firm has an internal board member other than the CEO who owns at least 5% of the firm’s outstanding shares, and zero otherwise. OUTSIDEROWN: total percentage of outstanding shares owned by outside directors divided by the number of outside directors. BLOCKHOLDER: indicator variable set equal to one if the firm has an external blockholder who owns at least 5% of the outstanding shares, and zero otherwise. VULNERABILITY is equal to 24 minus G-Index. All other variables are as defined in Table 2.
TABLE 5
The Relation between Future Performance and Components of Equity Grants
[pic]
(2)
Panel A: OLS regression of future performance on components of equity grants when governance strength is measured by GOVSCORE.
| Variable |Coefficient |Coefficient |
| |(Standard Error) |(Standard Error) |
| |Model (1) |Model (2) |
|Dependent Variable |FUTABROA |FUTABRET |
|Intercept ((0) |0.050 | |0.000 | |
| |(0.241) | |(0.004) | |
|EQUITY_GOVt |-0.041 |*** |-0.004 |*** |
| |(0.012) | |(0.001) | |
|EQUITY_ECOt |0.091 |*** |0.003 |*** |
| |(0.014) | |(0.001) | |
|EQUITY_RESIDt |-0.001 | |0.000 | |
| |(0.001) | |(0.000) | |
|ABROAt |1.153 |*** | | |
| |(0.065) | | | |
|ABRETt | | |0.006 | |
| | | |(0.015) | |
|SIZEt |-0.035 |*** |-0.001 |*** |
| |(0.005) | |(0.000) | |
|INVOPPt |0.211 |*** |0.001 | |
| |(0.027) | |(0.002) | |
|STDROAt |0.046 | |0.001 | |
| |(0.081) | |(0.007) | |
|STDRETt |-0.627 |*** |-0.011 | |
| |(0.141) | |(0.012) | |
|LEVERAGEt |-0.011 | |-0.003 | |
| |(0.027) | |(0.003) | |
|Industry dummies |Yes | |Yes | |
|Year Dummies |Yes | |Yes | |
|N |7,535 | |7,522 | |
|Adj. R2 |0.6266 | |0.0514 | |
*,**,*** significant at the 0.1, 0.05 and 0.01 levels (two-tailed).
This panel presents OLS results from estimating equation (2). Standard errors are calculated based on the Cameron’s (2008) bootstrapped clustered standard errors in both a firm and time dimension. Coefficients on the industry and year dummies are suppressed for expositional convenience. FUTABROA is industry-adjusted operating performance, calculated as income before extraordinary items after ESO expense summed over the subsequent three years scaled by total assets at the end of the current year, denoted FUTROA, minus the industry median FUTROA, where the industry assignments are based on the 48 Fama-French industry classifications. FUTABRET is the intercept coefficient, or alpha, from the firm-specific Carhart’s (1997) four factor model estimated over the three-year period following the end of the current year as follows:
[pic] (3)
where R is firm return minus risk-free rate, RMRF is the value-weighted market return minus the risk-free rate, and the terms SMB (small minus big), HML (high minus low), and Momentum are the monthly returns on zero-investment factor-mimicking portfolios capturing size, book-to-market, and momentum effects, respectively. EQUITY_GOV and EQUITY_ECO are the predicted components of equity grants arising from governance and economic factors, respectively. EQUITY_GOV is computed as follows:
[pic] (4)
where (1 is the estimated coefficient on GOVSCORE reported in model (1) of Panel A in Table 4.
[pic] (5)
where (j is the estimated coefficient on each economic determinant j reported in model (1) of Panel A in Table 4. The residual equity grant value explained by neither economic nor governance factors, denoted EQUITY_RESID, is calculated as EQUITY minus the sum of EQUITY_GOV and EQUITY_ECO. Size is the logarithm of total assets. Industry dummies are based on the 48 Fama-French industry classifications. All other variables are as defined in Table 2.
TABLE 5
The Relation between Future Performance and Components of Equity Grants
[pic]
(2)
Panel B: OLS regression of future performance on components of equity grants when governance strength is measured by GOVFACTOR.
|Variable |Coefficient |Coefficient |
| |(Standard Error) |(Standard Error) |
| |Model (1) |Model (2) |
|Dependent Variable |FUTABROA |FUTABRET |
|Intercept ((0) |0.047 |*** |0.000 | |
| |(0.239) | |(0.005) | |
|EQUITY_GOVt |-0.021 |*** |-0.002 |*** |
| |(0.005) | |(0.001) | |
|EQUITY_ECOt |0.098 |*** |0.0013 |*** |
| |(0.015) | |(0.0007) | |
|EQUITY_RESIDt |-0.001 | |0.0001 | |
| |(0.001) | |(0.0002) | |
|ABROAt |1.156 |*** | | |
| |(0.062) | | | |
|ABRETt | | |0.007 | |
| | | |(0.016) | |
|SIZEt |-0.038 |*** |-0.001 |*** |
| |(0.005) | |(0.000) | |
|INVOPPt |0.203 |*** |0.000 | |
| |(0.028) | |(0.002) | |
|STDROAt |0.047 | |0.001 | |
| |(0.085) | |(0.008) | |
|STDRETt |-0.619 |*** |-0.012 | |
| |(0.135) | |(0.012) | |
|LEVERAGEt |-0.008 | |-0.003 | |
| |(0.027) | |(0.003 ) | |
|Industry dummies |Yes | |Yes | |
|Year Dummies |Yes | |Yes | |
|N |7,535 | |7,522 | |
|Adj. R2 |0.6265 | |0.051 | |
*,**,*** significant at the 0.1, 0.05 and 0.01 levels (two-tailed). Standard errors are calculated based on Cameron’s (2008) bootstrapped clustered standard errors in both a firm and time dimension. All variables are as defined in Table 2. Coefficients on the industry and year dummies are suppressed for expositional convenience. FUTABROA is industry-adjusted operating performance, calculated as income before extraordinary items after ESO expense summed over the subsequent three years scaled by total assets at the end of the current year, denoted FUTROA minus the industry median FUTROA, where the industry assignments are based on the 48 Fama-French industry classifications.; FUTABRET is the intercept coefficient, or alpha, from the firm-specific Carhart’s (1997) four factor model estimated over the three-year period following the end of the current year:
[pic] (3)
where R is firm return minus risk-free rate, RMRF is the value-weighted market return minus the risk-free rate, and the terms SMB (small minus big), HML (high minus low), and Momentum are the monthly returns on zero-investment factor-mimicking portfolios capturing size, book-to-market, and momentum effects, respectively. EQUITY_GOV and EQUITY_ECO are the predicted components of equity grants arising from governance and economic factors, respectively. EQUITY_GOV is computed as follows:
[pic] (4)
where (1 is the estimated coefficient on GOVFACTOR reported in model (2) in Panel A of Table 4.
[pic] (5)
where (j is the estimated coefficient on each economic determinant j reported in model (2) of Panel A in Table 4. The residual equity grant value explained by neither economic nor governance factors, denoted EQUITY_RESID, is calculated as EQUITY minus the sum of EQUITY_GOV and EQUITY_ECO. Size is logarithm of total assets. Industry dummies are based on the 48 Fama-French industry classifications. All other variables are as defined in Table 2.
TABLE 6
Test of the Substitution Explanation
Panel A: Mean (median) future operating performance measured as income before extraordinary items after ESO expense cumulated over the three-year period subsequent to the grants of stock options and restricted stock (FUTABROA)
|Mean |EXCESS_EQUITY |t-test |
|[Median] | |Wilcoxon |
|FUTABROA | | |
|GOVSCORE |Low |High |p-value for |
| | | |High-Low |
|Weak |0.1828 |0.1243 | ................
................
In order to avoid copyright disputes, this page is only a partial summary.
To fulfill the demand for quickly locating and searching documents.
It is intelligent file search solution for home and business.