Essay: Executive compensation

I. Introduction (1.5)
Executive compensation is a frequently mentioned topic nowadays. Is the remuneration of a CEO actually in accordance with his performance? A worker in the United States receives on average a yearly salary of $34.645 where a CEO receives on average $12.259.894, which means a ratio of 1 to 350 (Sorapop Kiatpongsan, Michael Norton, 2014). According to Cooper, Gulen and Raghavendra Rau the more CEOs get paid, the worse their companies perform over the next three years. This effect is even stronger for CEOs who receive higher incentive pay and stronger for CEOs with greater tenure (Cooper, Gulen, Raghavendra Rau, 2014). It is therefore important to research how executive compensation can be restricted. However executives are being paid by stock options and restricted shares and both shareholders as well as executives are affected by firm performance through the share prices. Both thus have a common interest in maximizing firm performance.
There have already been many studies about the relationship between CEO compensation and firm performance, which is stated as the pay-performance sensitivity (e.g. Jensen & Murphy, 1990; Hall & Liebman, 1998; Duffhues & Kabir, 2008). However, there is a possible effect of ownership concentration on this relation, which only has little research up till now. This study is about the moderating effect of ownership concentration on the relation between executive compensation and firm performance.
David F. Larcker found that on average, only 17.5 percent of the compensation a CEO receives comes from salary, where 16.6 percent comes from bonuses, 27.9 percent from stock options, 21.1 percent from restricted shares, 12.1 percent from performance plans and 4.7 percent consists of other compensation (Larcker, 2011). This means that compensation on average consists largely of incentive based compensation rather than base salary. Sanders and Donald state that ‘The argument for paying a CEO with stock options is that it gives the executive an incentive to increase value for shareholders. If the CEO drives up the underlying stock price, the options award will be worth more.'(Gerard Sanders, Donald C. Hambrick, 2007). Since shareholders are affected by the choices the CEO makes, they will also have the incentive to influence those choices. The concentration of ownership (due to beneficial ownership) is thus expected to affect the relation between firm performance and executive compensation, mainly because of the incentive based compensation.
Combining both the already known relationship between firm performance and CEO compensation and the expected effect of ownership concentration on this relation, due to the incentive based compensation, results in the following research question:
Does ownership concentration have a negative effect on the pay-performance sensitivity of the CEO compensation contract?
This means that when ownership is more concentrated, the pay-performance sensitivity will be weaker, so that firm performance has less influence on total compensation?!?! The concentration of ownership can thus affect the level of compensation a CEO receives, since it will mainly affect incentive based compensation. It is expected to have a negative effect because shareholders will want to have less fluctuating firm performance caused by the level of compensation?!?!
The sample used to study this research question consists of 60 listed companies in the United States, on the New York Stock Exchange or the Nasday. The data on these companies is collected from ExecuComp and the Proxy statements. This study uses univariate and multivariate analyses to test the hypothesis. A negative and significant relationship between the interactionterm (DblockhROA) and CEO compensation has been found. The Beta of the interactionterm on total compensation was -0.297, which means that a one percent change in the interactionterm results in a change of -0.297 per cent in total compensation. The insignificance of the control variables and the models using sharepercentage as a measure of ownership concentration could be driven by the small sample size. Future research should therefore increase the number of observations, to see if the results are more significant.
Understanding what the effect of ownership concentration on CEO compensation is, has relevance for several parties. It has social relevance because it shows firms and shareholders whether or not they should stimulate beneficial ownership. Concerning this study, it is indeed relevant that concentrated ownership should not be discouraged in order to reduce the pay-performance sensitivity. Furthermore it has academic relevance because it can provide more detailed future research on what further effects beneficial ownership could have,
This research paper is structured as follows. The second chapter discusses existing literature on the topic of this study, and describes the hypothesis used. The third chapter describes the methodology and the fourth chapter describes the sample selection. The fifth chapter will discuss the results of the regression, and the sixth chapter will show the conclusions.
II. Literature review and hypothesis development (3 G)
Agency theory.
The objective of shareholders is to maximize wealth therefore, agency theory predicts that CEO compensation policies will depend on changes in shareholder wealth (Michael C. Jensen, Kevin J. Murphy, 1990). The relation between pay and performance is derived from the agency theory (Holmstr??m, 1979, or Grossman and Hart, 1983). According to these models, compensation plans should be designed to align the interests of risk-averse self-interested executives with those of shareholders. Ex-post payouts depend on the likelihood that the desired actions were in fact taken. Agency theory also suggests that higher Pay-Performance Sensitivity (PPS) motivates top executives to put in more inputs and improves corporate performance (Yaron Amzaleg1, Ofer H. Azar , Uri Ben-Zion and Ahron Rosenfeld, n.d.).
Pay-performance sensitivity.
Jensen and Murphy (1990) define pay-performance sensitivity as the dollar change in CEO wealth associated with a dollar change in shareholder wealth and interpret higher sensitivities as indicating a closer alignment between the CEO and his shareholders. Taking into account cash compensation, stock options, and probability of dismissal, they find that a CEOs wealth changes $3.25 per $1,000 change in shareholder wealth (Michael C. Jensen and Kevin J. Murphy, 1990). Pay-performance sensitivity thus measures how managerial compensation changes with firm performance and shareholder wealth (Jensen and Murphy 1990; Garen 1994; Hall and Liebman 1998; Aggarwal and Samwick 1999).
Ownership concentration.
Ownership concentration refers to the amount of stock owned by individual investors and large-block shareholders. Block holders are investors that hold at least five per cent of equity ownership within the firm. Ownership concentration can be an internal governance mechanism that helps reduce the likelihood of managerial opportunism because managers and boards of directors are more likely to take into account the preferences and interests of large shareholders (Financial times, Murdoch sidesteps investor pressure).
Ownership concentration and firm performance.
In the study of Mohammed Soliman on Financial Firm Performance in Saudi Arabia, a significant relationship between ownership concentration and firm performance has been found. According to Mohammed Soliman firm financial performance measured by the accounting rate of return on assets and rate of return on equity generally improves as ownership concentration increases (Soliman, 2013). In Pakistan a study on the effect of ownership structure on firm performance has taken place. They have found evidence that the identity of the owners matters more than the concentration of ownership, but the concentration of ownership still seems to have a positive effect on firms’ profitability and performance measures, but only when it is family, foreign or director ownership (Attiya Y. Javid and Robina Iqbal, 2009). In the study of Irena Grosfeld on the Warsaw Stock Exchange, they found that ownership concentration does improve corporate value, but only in the firms belonging to more traditional industries. But in firms belonging to the segment of high technology they could not reject the hypothesis that ownership concentration does not affect corporate value. (Irena Grosfeld, 2006)
Level of ownership concentration.
A higher level of ownership concentration suggests a stronger monitoring power from investors over a firm’s managerial decisions, due to the incentives of these owners to actively safeguard their investment. Owners with a significant amount of shares may take aggressive actions, directly or indirectly, over firm decisions such as the replacement of CEO or poor management with their voting power (Murdoch sidesteps investor pressure, Financial Times). Also in firms with stronger ownership concentration, large shareholders have more incentives to directly monitor management and therefore explicit incentive pay contracts may become less important (Bin Kea, Kathy Petronib, Assem Safieddineb, 1999). In the study on Institutional ownership and CEO compensation is found that the largest owner’s concentration is associated with lower levels of compensation, as well as with higher ratios of salary to total compensation and lower ratios of options to total compensation, but that the number of block holders does not predict any aspects of CEO compensation. In addition, institutional ownership dispersion is associated with increased levels of compensation and greater use of incentive compensation (Raihan Khan, Ravi Dharwadker, 2004).
Firms with a low level of ownership concentration might indicate weaker governance power because investors with less ownership interests have little incentive to pay attention to the strategic decisions of the firm and thus, are less motivated to closely monitor and discipline top executive behaviors. Compared to large block holders, small investors are more likely to ‘vote with their feet’ in cases of poor firm performance (Murdoch sidesteps investor pressure, Financial Times).
CEO compensation and ownership concentration.
Most CEO compensation packages consist of four components, namely a base salary, an annual bonus based on performance, stock options and long-term incentive plans (Kim Vos, 2014). The bonus, stock options and long-term incentive plans are all three forms of incentive-based compensation. David F. Larcker found that on average, only 17.5 percent of the compensation a CEO receives comes from salary, where 16.6 percent comes from bonuses, 27.9 percent from stock options, 21.1 percent from restricted shares, 12.1 percent from performance plans and 4.7 percent consists of other compensation (Larcker, 2011). This means that compensation on average consists largely of incentive based compensation rather than base salary. The traditional view is that an optimal executive compensation contract can provide managers with efficient incentives to maximize shareholder wealth and mitigate associated agency costs (Bebchuk and Fried, 2003).
A study on ownership concentration and CEO compensation in New Zealand finds a non’linear effect of ownership concentration on CEO compensation’firm performance relationship, that is CEO compensation is negatively related to firm performance in firms with high concentrated ownership structure, and positively related to firm performance in firms with low concentrated ownership (Haiyan Jiang, Ahsan Habib, Clive Smallman, 2009).
Hypothesis Development.
From the study in New Zealand we find that CEO compensation differs between strong and diffuse ownership concentration. Furthermore the research of Jensen and Murphy finds that there is a relationship between firm performance and CEO compensation. This pay-performance sensitivity measures how managerial compensation changes with firm performance and shareholder wealth (Jensen, Murphy, 1990). Kea et al state that in firms with stronger ownership concentration, large shareholders have more incentives to directly monitor management and therefore explicit incentive pay contracts may become less important (Bin Kea, Kathy Petronib, Assem Safieddineb, 1999). Combining these things results in the following hypothesis:
H1: Ownership concentration has a negative effect on the pay-performance sensitivity of the CEO compensation contract.
III. Method (1.5)
Model
To study the hypothesis a regression model is needed. The model used is a combination of ownership concentration, pay-performance sensitivity, an interaction term and five control variables. Ownership concentration will be measured in two different ways, namely by the number of block holders the firm has and by the total share percentage those block holders hold together. Total compensation is used since both equity-based compensation as well as cash-based compensation have incentive based parts. To check whether using extra control variables helps getting the model more significant, three different models per ownership concentration measure are used. The following models will empirically test the hypothesis through regression analysis:
Model 1: Ln(TOTCOMP)=f(?? – ??1Dshaper + ??2ROA + ??3DshaperROA + ??)
Model 2: Ln(TOTCOMP)=f(?? – ??1Dshaper + ??2ROA + ??3DshaperROA + ??4ExecAge + ??5FirmSize + ??)
Model 3: Ln(TOTCOMP)=f(?? – ??1Dshaper + ??2ROA + ??3DshaperROA + ??4ExecAge + ??5FirmSize +??6TEN + ??7LEV + ??8MTB+ ??)
Model 4: Ln(TOTCOMP)=f(?? – ??1Dblockh + ??2ROA + ??3DblockhROA + ??)
Model 5: Ln(TOTCOMP)=f(?? – ??1Dblockh + ??2ROA + ??3DblockhROA + ??4ExecAge + ??5FirmSize + ??)
Model 6: Ln(TOTCOMP)=f(?? – ??1Dblockh + ??2ROA + ??3DblockhROA + ??4ExecAge + ??5FirmSize +??6TEN + ??7LEV + ??8MTB+ ??)
Test variables
The computations of the variables used in this study can be found in Appendix 1, Table 1.
TOTCOMP – Total Executive Compensation. Total CEO Compensation consists of the total of base salary and bonus (the cash-based part) as well as the total of stock options exercised and long-term incentive plans (the equity-based part).
Dshaper, Dblockh – Ownership Concentration. The concentration of corporate ownership has been found to systematically influence the performance of firms, and thus the CEO compensation (Holderness, 2003 and Denis and McConnell, 2003). Ownership concentration refers to the amount of stock owned by individual investors and large-block shareholders. It refers to the beneficial ownership of stock, which means that one investor has at least five percent of total outstanding stock. A block holder (block shareholder) is defined as an outside investor, no member of either management or supervisory board nor an anti-takeover entity, who owns at least 5% of the cash flow rights or 5% of the voting power (Nicolai Knop and Gerard Mertens, 2010). Ownership concentration can be measured in many ways. But in this study two types of measures with regard to block ownership will be used, Dblockh and Dshaper. Both measures of ownership concentration are dummy’s. We expect ownership concentration to have a negative effect on CEO compensation, because firms which have more control by shareholders will have less possibility to pay the CEO high amounts.
ROA ‘ Firm Performance. The firm performance will be measured by Return on Assets, which is a measure of firm performance. A high Beta of the ROA means a high pay-performance sensitivity. We expect firm performance to have a positive relationship with CEO compensation, because more positive return will result in a higher compensation for the CEO.
DshaperROA, DblockhROA – Interaction term/ Moderator. The interaction term is a combination of the dummy for ownership concentration (both Dblockholders and Dsharepercentage) and the pay-performance sensitivity (return on assets). This results in the following interactions terms: DblockhROA and DshaperROA, where the first is used in combination with Dblockholders and the second with Dshareperc. A negative coefficient of the interactionterm means a negative effect of ownership concentration on the relation between firm performance and total CEO compensation. This negative effect is also expected, since beneficial shareholders have more incentive to control because it will affect their own benefits more than it will for a small share owner. The hypothesis can be accepted if the interactionterm indeed has a negative and significant coefficient.
Control variables.
FirmSize ‘ Firm Size. Several studies find that firm size has a strong effect on executive compensation, for example Tosi et al (2002) find that firm size accounts for more than 40% of the variance in CEO pay. Firm size will be measured by the market value of the firm. Generally CEOs of larger firms receive higher compensation. This means that firm size has a positive effect on CEO compensation (Zhou, 2000) (Kim Vos, 2014).
ExecAge ‘ Executive’s Age. This is an executive specific variable, since it consists of the age of the executive. Age is measured in years, and will be used because the older executives are, the higher their compensation is expected to be because of knowledge and experience (Twan Janssen, 2010). The executive’s age is thus expected to have a positive effect on CEO compensation.
TEN ‘ CEO tenure. CEO tenure means how long the CEO has been at this position, which is expected to have a positive effect on total compensation.
LEV ‘ Leverage. Leverage consists of the debt to equity ratio, which is expected to have a negative effect.
MTB – Market-to-Book ratio. This holds the market value of the share against the book value of those shares, which is expected to have a positive effect.
IV. Sample selection (samen met 3 1.5)
The Sample of this study consists of 60 listed firms in the United States. These firms are on the New York Stock Exchange or the NASDAQ. All the companies used in this study have information on their CEO compensation, firm performance, firm size, age of executive and an overview of beneficial ownership. The information needed for all these listed companies is found in the proxy statements and ExecuComp. From ExecuComp the data on CEO total compensation is derived, also net income, total assets and the data for the control variables are found there. The proxy statements provided the data on block holders (beneficial ownership) and share percentage, since the data on beneficial ownership from ExecuComp was outdated (1996-2001). All the firms are randomly selected and there was not made a distinction between different industries.
V. Results (2.5 G)
Descriptive statistics
Table 2 represents the descriptive statistics of the variables used in this study. It shows that the mean of the total compensation is $16036.446, where the maximum is almost $140000 Dollars. Furthermore, the mean of MTB (1639.5628) shows that the market value of the shares is way higher than the book value. The mean of ROA is positive (0.06163) which means that firms have on average a positive return on their assets.
Univariate analysis.
The Independent-Samples T-test is used to compare means between groups. The sample of this study is split into two groups based on the dummy variable for the Blockholders (Dblockh, Appendix 1, table 3) and also split in two groups based on the dummy variable for the Sharepercentage (Dshaper, Appendix 1, table 4). For this study it is relevant to investigate whether the means of the variables are different with diffuse or strong ownership concentration.
There indeed is a significant difference in Table 3 for total compensation between diffuse or strong ownership (t=2.461), this difference is a little less significant in Table 4 (t=1.776). Therefore, it could be stated that ownership concentration has a negative effect on total CEO compensation. Furthermore, there is a significant difference for FirmSize (t=2.239; t=2.124). This shows that when the size of a firm is bigger, there is more diffuse ownership concentration (less blockholders), since there are expected to be more shares outstanding so it is more difficult to gain a beneficial ownership. The ROA is also significantly higher for diffuse ownership concentration than for strong ownership concentration (t=1.936; t=3.693). The Market-to-Book ratio and Leverage are not significant in both tables.
Correlation matrix.
The Pearson correlation matrix can be found in appendix 1, Table 5 and shows many significant correlations. The correlation matrix shows that there is a significant negative correlation between Dblockh and total compensation, and Dshaper and total compensation, with correlations of -0.293 and -0.218 respectively. This shows that there indeed is a significant relationship between ownership concentration and the level of total compensation and can complement the results found in the univariate analysis. The matrix also shows that there is no significant relationship between TOTCOMP and FirmSize, ExecAge, MTB and LEV, only TOTCOMP and TEN have a significant negative relationship, with a correlation of -0.238. FirmSize has a significant negative relationship with both Dblockh and Dshaper, thus there is diffuser ownership concentration when a firm is of a bigger size. There are also significant negative relationships between ROA and both Dshaper and Dblockh, as well as significant positive relationships between ROA and both DshaperROA and DblockhROA. Therefore, a significant relation between firm performance and ownership concentration can be concluded. Furthermore, there are no multicollinearity problems, since the correlation coefficients of the independent variables are not higher than 0.8. Because of that, no variables have to be excluded in the multivariate analysis.
Multivariate analysis.
The results of the multivariate analysis can be found in appendix 1; Table 6, 7, 8, 9, 10 and 11. Three models per different ownership concentration variable are used to see if the models are more significant with or without the control variables. Table 6 presents the first model; Table 7 the second model; Table 8 the third model; Table 9 presents the fourth model; Table 10 the fifth model; Table 11 the sixth model.
All the models using Dshaper are not significant (p=1.668;p=1.271;p=1.381), therefore they cannot be used to make conclusions on the hypothesis of this study. These models also only predict 8.2%, 10.5% and 17.8% of the variance in TOTCOMP respectively. The small sample size can explain why the models are not significant.
On the contrary, the models using Dblockh are all three significant, with p-values of 3.647, 2.315 and 1.838 respectively. Therefore we can make conclusions considering these models. The first model explains 16.3% of the variance in total CEO compensation, where the second model explains 17.6% and the third model 22.4%. Adding the control variables thus increases the variance in total compensation explained by the model. The first model which only consists of Dblockh, ROA and DblockhROA shows that there is a significant positive relationship between ROA and TOTCOMP (Beta=0.414; P-value=0.026) as well as significant negative relationship between DblockhROA and TOTCOMP (Beta=-0.297; P-value=0.101). Therefore, the hypothesis ‘Ownership concentration has a negative effect on the pay-performance sensitivity of the CEO compensation contract’ can be accepted. The Beta of DblockhROA of -0.297 states that an increase of 1 per cent in the interactionterm results in a decrease of 0.297% of the total compensation. The ROA shows an increase of 0.414% in TOTCOMP when it increases with 1%.
In the second and third model, where the control variables are added, DblockhROA is not significant anymore. The control variables themselves are also not significant, therefore we cannot conclude if they affect total CEO compensation. Although the results are not significant now, they could have been with a bigger sample size.
VI. Conclusion (1 BG)
This study aimed at answering the question whether or not ownership concentration has a negative effect on the pay-performance sensitivity of the CEO compensation contract. This means that ownership concentration should have a moderating effect on the relation between the firm performance and total CEO compensation. A negative and significant relationship between the interactionterm has been found, which consists of ROA and Dblockh, and total compensation (p=0.101). Therefore the hypothesis Ownership concentration has a negative effect on the pay-performance sensitivity of the CEO compensation contract can be accepted. The Beta of DblockhROA is -0.297 which states that an increase of 1% in the interactionterm results in a decrease of 0.297% of the total compensation. Both DblockhROA and DshaperROA were not significant at the correlation matrix, but did have a significant p-value at the univariate analysis. On the contrary, both Dblockh and Dshaper were significant at the correlation matrix, but did not show any significance in the multivariate analysis.
Furthermore, following table 9, 10 and 11, ROA has a strong, positive and significant effect on total compensation. This complements prior research that already stated that there is a positive relationship between firm performance and total compensation, and with that the existence of the Pay-Performance Sensitivity.
All the control variables (FirmSize, ExecAge, TEN, LEV, MTB) did not have a significant relationship with TOTCOMP, therefore it cannot be concluded if they have an effect on total compensation. It could be that some control variables have more effect on cash-based compensation, or have more effect on equity-based compensation, and therefore the significant effect is gone when measuring it against total compensation. For instance, prior results stated that FirmSize did have a significant effect on cash-based compensation (p=0.000; B=4.63) but did not have this effect on equity-based compensation. The insignificance of the variables is supposedly driven by the small sample size used in this study. Therefore, for future research it would be recommendable to retain a bigger sample size.
The univariate analysis gives strong evidence on the effect of the concentration of ownership. TOTCOMP shows both with using Dshaper as well as Dblockh a significant difference between diffuse ownership concentration and strong ownership concentration. The univariate analysis also shows that there is a significant difference in FirmSize between diffuse and strong ownership concentration. It is more likely that when a firm is of a bigger size, to have less beneficial ownership since the firm has more outstanding stock, and therefore less block holders. A bigger firm size comes with diffuser ownership concentration, and thus higher compensation, while a smaller firm comes with stronger ownership concentration and thus a lower compensation
There are also several limitations of this study, which could be changed for future research. First of all, the size of the sample is expected to be a big influence for the significance of the multivariate analysis. It would thus be very important to take into account more firms when researching this topic on further relationships. In this study only 16.3% of the variance in total compensation was explained by the model without the control variables and 22.4% in the model with the control variables, so it would be possible that there are other factors, besides the variables used in this study, that affect total compensation. For example it could be that industry differences also affect total compensation, for instance financial or technological firms, therefore it could be interesting to take this into account for future research, since that has not been done in this study.
Future research should focus on the effect of ownership concentration on the pay-performance sensitivity of the CEO compensation contract, to reduce the ridiculously high compensation CEOs receive nowadays, while they should not always really earn them.