«Published Annually Vol. 6, No. 1 ISBN 978-0-979-7593-3-8 CONFERENCE PROCEEDINGS Sawyer School of Business, Suffolk University, Boston, Massachusetts ...»
Many researchers criticize the use of a cash bonus as an incentive component in executive compensation contracts (e.g., Sigler, 2011; Gaver and Gaver, 1993). They argue that since a cash bonus is linked to current performance, it usually encourages the decisions which are reflected in current year measures. Since current earnings is not a good measure of future period effects, this type of incentive may not be appropriate for managers whose decisions have significant effects on future periods. Accounting earnings are also subject to management manipulation.
The problem of emphasizing short-term performance can be reduced by using stock-based incentive schemes. Jensen and Murphy (2010, 1990) argue that executive inside stock ownership and stock options can be used to link executives’ wealth to the value of the firm. As noted by Abdel-khalik (1998), when executives receive their contingent pay in a form of stock or stock options grants, they face a payoff that is related to their firm performance and to the changes in share prices. These incentive plans are assumed to encourage executives to emphasize both current and future performance. Abdel-khalik (1998) examined the effect of different stock-based incentive schemes. He posits that “CEOs who own more options than current equity shares will increase their wealth only when prices rise and, therefore, do not share in the downside risk as much as those CEOs who own more equity shares” (Abdel-khalik, 1998, p. 9). His results suggest that although risk sharing is related to stock performance, different types of equity instruments induce different incentives. His study, like other limited research examining risk-sharing effect on corporate performance, may be criticized on the ground that it employs measures that are not refined enough to explain the relationship between executive compensation and corporate performance. Most of that research, however, investigates the relationship between CEO compensation and corporate performance, ignoring the effect of risk sharing
Conference papers © Knowledge Globalization Institute, Pune, India, 2012
embodied in the compensation contracts of other top executives. More research is needed to provide more evidence as to the relationship between CEO and other executives risk sharing and stock performance, and to examine whether different incentive instruments have different effects on current and future stock performance. This study addresses these issues by examining two propositions: The first is that there is a positive relationship between the level of risk sharing in executive compensation contracts and stock performance of the firms they manage. The second proposition is that, contrary to a cash bonus, stock-based incentives enhance future, rather than current, stock performance. Formally stated, the first two
hypotheses of this study are:
H1: Stock of firms managed by relatively high risk sharing executives performs better than those of firms managed by relatively low risk sharing executives.
H2: Future stock performance of firms with more intensive stock-based incentives will be better than those of firms with less intensive stock-based contracts.
The term “new economy firms” is new in the literature. As defined by Murphy (2003), new economy firms are “companies competing in computer, software, internet, telecommunications, or networking fields.” (p. 131). Recent literature provide strong evidence that new economy firms rely more on stock-based compensation than do old economy firms (Ittner et al, 2003; Murphy, 2003; Anderson et al., 2000). Hall and Liebman (1998) have noted that stock options have become a common component of executive compensation contract in publicly traded corporations. The importance of options and other forms of stock-based compensation is more evident in the new economy firms (Murphy, 2003). This might be due to the distinct characteristics of those firms. As noted by Ittner et al. (2003), new economy firms are relatively smaller (in terms of sales and employees but not in market value) than old economy firms. They are growing more rapidly and investing more in R&D activities. Another distinguishing characteristic of new economy firms is “the extensive use of stock options as compensation for top level executives. Stock options were embraced as a critical component contributing to the success of these companies.” (Murphy, 2003, p. 146). This is consistent with the conclusions of Ittner et al. (2003) and Anderson et al. (2000).
Both studies show that stock and options grants in new economy firms remain higher than grants in old economy firms after controlling for all other distinguishing variables. Combining these results with those provided by Abdel-khalik (1998), we can
derive the third hypothesis as follow:
H3: Stock performance is more positively correlated to executives’ risk sharing in new economy firms than in old economy firms.
The next section describes the methodology to be employed for testing the hypotheses of the study.
The data used in this study is obtained from Compustat’s ExecuComp database. Aggarwal and Samwick (1999) provide several reasons for using the ExecuComp database in executive compensation studies. First, it contains data on total compensation for the top five executives, not only the CEO, at each of the firms included in the dataset. Second, it contains data on the executives’ holding of stock in their own companies and existing stick options on their own company’s stock. Third, it provides data for a very wide cross-section of firms. Fourth, it provides data on the total return to shareholders in each year, specified in percent returns and the market value of the firm at the beginning of each year. An initial large sample was selected for the period 1996 to 2006. It was then screened for missing data and a final usable sample of 6,084 firm-year observations was obtained.
The current study is concerned with the effect of executive risk sharing on stock performance. Two dependent variables are used to represent stock performance; the market price per share (MVPS), and the total return to common shareholders (TRCS). As in Murphy (2003), the total return to shareholders is measured by the market value at the beginning of the year multiplied by the percent of return to common shareholders. MVPS was chosen for two reasons: First, prior research shows that share price is a powerful inventive device as it aggregates the information of many investors about the value of the firm, and that stock market
Conference papers © Knowledge Globalization Institute, Pune, India, 2012
reacts to the announcement of stock-based incentive schemes (Bhagat et al, 1985). Second, share price may provide information about executive activities which cannot be inferred by analyzing long-term value of the firm (Grant et al., 1995). TRCS was also chosen for three reasons: First, it is not subject to manipulation by management. Second, prior research shows that it is a good measure of the change in the firm’s financial condition. Third, Executives’ wealth is usually tied directly to stock return when stock options or other share ownership plans are used in the incentive scheme.
In order to examine the effect of risk sharing compensation on each of the dependent variables, three independent variables were identified: The first is the percentage of the number of shares currently owned by top executives to the total number of voting shares outstanding at the beginning of the year (SHOS). The second is the stock-based compensation (SBC) measured, as in Murphy (2003), by the sum of restricted stock grants (valued at the grant-date market price) and stock options (valued at the grant-date using ExecuComp’s modified Black-Scholes methodology). The third independent variable is the percentage of the bonus to the total variable compensation granted in a year (CBC), where total variable compensation is measured by total compensation minus the salary. Due to the nature of the empirical data used in this study and to reduce the effect of extreme values, the natural logarithm of each of the variables is used in the regression analysis. The reason for choosing these independent variables is to examine the effect of different incentive instruments on the dependent variables.
In order to determine if the market value per share and total return to shareholders are related to the level of risk sharing of the executives, and whether different incentive instruments have different effects on current and future stock performance, four multiple regression models were used. The first two models are to test the effect of risk sharing on the current
year stock performance. They are presented as follows:
ln(MVBS)t = a0 + a1 ln(SHOS)t + a2 ln(SBC)t + a3 ln(CBC)t + ε ln(TRCS)t = b0 + b1 ln(SHOS)t + b2 ln(SBC)t + b3 ln(CBC)t + ε Where, ln(MVBS)t = the natural logarithm of the market value per share in year t.
ln(TRCS)t = the natural log of the total return to shareholder in year t.
ln(SHOS)t = the natural log of the percentage of currently owned shares in year t.
ln(SBC)t = the natural logarithm of stock-based compensation.
ln(CBC)t = the natural logarithm of non-stock-based compensation.
These two models will be used to test for the relationship between each of the independent variables in year t, and each of the dependent variables in the same year. The other two models are used to test for the effect of risk sharing on future
performance. They are as follows:
ln(MVBS)t+1 = the natural logarithm of the market value per share in year t+1.
ln (TRCS)t +1= the natural logarithm of the total return to shareholder in year t+1.
The other variables are as defined in the first two models. These two models will be used to test for the effect of incentives in year t on future performance in year t+1.
Conference papers © Knowledge Globalization Institute, Pune, India, 2012 To test for the first two hypotheses, the final sample was partitioned based on the percentage of variable compensation to total compensation into two sub-samples: Highrisk sharing firms (N=3,183), and low-risk sharing firms (N=2,901). To test for the third hypothesis, the final sample was partitioned into two sub-samples based on the primary SIC code. The first sub-sample includes the new economy firms (N=749) as defined by Murphy (2003), while the other sub-sample includes the old economy firms (N=2,530). Table 1 presents the descriptive statistics of the samples.
------------------------------------------------Insert Table 1 about here
To test for the first two hypotheses, the four models were used for the whole sample and for each of the high-risk sharing and lowrisk sharing sub-samples. Table 2 and Table 3 show the results of the regression analyses and the correlation results respectively.
The results indicate that the level of risk sharing does not influence the future market value of shares in both high-risk sharing and low-risk sharing firms. However, it does negatively influence the current and future return to shareholders in both types of firms.
The results also indicate that more intensive stock-based contacts positively influence current and future stock performance. This indicates that although stock based compensation includes Additional tests of multicollinearity showed a VIF between 1 and 1.15, with a tolerance value between 0.82 and 0.99, indicating no multicollinearity problem with the data.
a level of risk sharing, the level of risk sharing is not what drives stock performance. Rather, it is the value of the stock based compensation that drives performance leading to a higher market value per share and higher return to shareholders.
------------------------------------------------Insert Table 2 about here
------------------------------------------------Insert Table 3 about here
------------------------------------------------To test for the third hypothesis, the four models were used with each of the new economy firms and the old economy firms subsamples. The results show that the level of risk sharing does not influence the future market value of shares in both new economy and old economy firms. However, it does negatively influence the current and future return to shareholders in old economy firms, but not in new economy firms. The results also indicate that more intensive stock-based contacts positively influence current and future stock performance in both types of firms. As with the previous two cases, these results indicate that although stock based compensation includes a level of risk sharing, the level of risk sharing is not what drives stock performance. Alternatively, it is the value of the stock based compensation that drives performance leading to a higher market value per share and higher return to shareholders. This result appears to be much more pronounced in new economy firms due to their compensation structure.
Much of the research on the relationship between executive compensation and firm performance in publicly owned corporations employs the agency theory to explain how incentives in compensation contracts affect performance (e.g., Indjejikian, 1999;Bushman and Indjejikian, 1993; Lambert and Larcker, 1987). A common proposition underlying this line of research is that in order to motivate executives to work for the best interest of the shareholders, compensation contracts should include some form of incentive component (Hays and Schaefer, 2000; Indjejikian, 1999; Baber et al., 1999; Sloan, 1993). Despite the differences in compensation contracts used in different firms, they embody both the incentive and risk sharing components.