CEO compensation and firm performance
Evidence for 2004-2010 from the Netherlands
MSc BA Organization & Management Control
Thesis
M.A.J. Munsterhuis
Student at the University of Groningen, Faculty of Economics and Business
March 2012
CEO compensation and firm performance in the Netherlands 2
CEO compensation and firm performance
Evidence for 2004-2010 for the Netherlands
Abstract
This study examines the two-sided relationship between CEO compensation and firm performance, with the agency theory as theoretical framework. In order to analyze the alignment of the incentives of CEOs with their shareholders, data on CEO compensation practices and firm performance is collected from Dutch listed firms from 2004 to 2010. Control variables such as size, leverage and industry are taken into account in this analysis. Including a lag of 2 years, this study finds that firm performance is positively determined by long term equity-based incentives in the executive contract. Furthermore, it is shown that CEO payments are determined by performance as well. Both accounting and market returns are positively related to the annual bonus granted to the CEO. These results are in line with previous studies. Finally, the current Dutch public debate about executive pay in the both the financial and the public sector is discussed.
Keywords: CEO compensation, Corporate Governance, Agency Theory, Firm performance
CEO compensation and firm performance in the Netherlands 3
TABLE OF CONTENTS
1
INTRODUCTION
4
2
THEORY AND BACKGROUND
5
2.1
Agency theory
5
2.2
The executive contract
6
2.2.1 Fixed pay
7
2.2.2 Variable pay
8
2.3
CEO compensation and firm performance
10
2.4
Dutch remuneration practices
13
3
DATA AND METHODOLOGY
14
3.1
Data
14
3.1.1 CEO compensation variables
15
3.1.2 Firm performance variables
17
3.1.2 Control variables
19
3.2
Methodology
20
4
EMPIRICAL RESULTS
22
4.1
Firm performance as explanation for CEO compensation
22
4.1.1 Lagged relationship
24
4.2
CEO compensation and future firm performance
25
5
REFLECTION ON THE DUTCH SITUATION
27
5.1
Bonuses in the financial sector
27
5.2
Compensation in the Dutch public sector
30
6
CONCLUSION
31
REFERENCES
35
CEO compensation and firm performance in the Netherlands 4
1
INTRODUCTION
In recent years, CEO compensation and corporate governance practices in general have been the subject of a heated public debate in the Netherlands. Total compensation granted to CEOs has risen substantially over the last years, which is primarily driven by an increase in equity-based compensation (Swagerman and Terpstra, 2009). The relationship between CEO compensation and firm performance is of great public concern and receives a lot of media attention, which especially holds when the link between compensation and firm performance is absent. Recently, in the aftermath of the credit crisis, the proposed bonus for 2010 of CEO Jan Hommen of ING was highly criticized by the Dutch minister of finance, other politicians and also a large part of the Dutch society. The reason for this was the fact that ING obtained government support in preventing bankruptcy during the crisis.1
As a result of the accounting scandals at the beginning of the 21st century, the Dutch corporate governance code has been introduced in 2003. Initiated by the former minister of Finance a commission has been established in order to come up with a code to improve the Dutch corporate governance. The purpose of this code is to increase the transparency of annual reports, better justification to the supervisory board and improve the control and protection of (minority) shareholders for all public listed companies. Since that moment all Dutch listed firms are obliged to disclose the compensation practices to their CEO and the board of directors into their annual reports. However, compliance to the governance code happens on a ‘comply-or-explain’ basis. The combination of a unique governance system and a rapidly changing economic environment makes Dutch executive compensation an interesting topic for research.
Although the relationship between firm performance and CEO compensation has been extensively investigated by various prominent authors, for example Jensen and Murphy (1990) with their study about performance compensation and top-management incentives, the evidence found is still not persuasive. The agency problem is the most prominent theory in this field of research. According to this theory there should exist is a positive relationship between firm performance and CEO pay. Nevertheless, the results in this field of research contradict each other. Several studies including the study of Jensen and Murphy (1990), found a positive association between CEO pay and firm performance. Though, a substantial number of studies found no relationship, or even a negative relationship between CEO compensation and firm
1
CEO compensation and firm performance in the Netherlands 5 performance. Most of the studies about this topic have been conducted for countries with the Anglo-American corporate governance model with a one-tiered board of directors and in a much lesser extent for countries with a two-tiered board of directors like the Netherlands.
This study aims to extent the literature and to fill the gap for non-Anglo-American countries, by investigating how firm performance influences CEO remuneration and vice versa in Dutch listed firms. Unique panel data on individual CEO compensation of firms listed on the Dutch Stock Exchange from 2004 until 2010 is used for this research. Duffhues and Kabir (2007) where the only authors that examined the pay for performance relationship in the Netherlands. Their period of research stretched from 1998 until 2001, and failed to detect a positive pay for performance relationship for Dutch listed companies. However, after their period of research some important regulations were established in the Netherlands. One of the main objectives of these regulations was to counteract unjustified excessive CEO compensations, and make CEO compensation more dependent on firm performance. Since their study only considers cash payments to CEOs, this study contributes to the existing literature by taking equity-based compensation like stocks and options into account as well.
This study proceeds with giving a theoretical background and formulating hypotheses in the next section. The data and methodology will be explained in section 3. After that, the empirical results will be presented and discussed in section 4. In section 5 the public debate in The Netherlands will be reviewed and the conclusion is presented in section 6.
2
THEORETICAL BACKGROUND AND HYPOTHESES
CEO compensation and firm performance in the Netherlands 6 2.1 Agency theory
Research on CEO compensation started in the 1980s together with the emergence and wider acceptance of the agency theory which was formulated by Jensen and Meckling (1976). However, Berle and Means (1932) were the first to examine the consequences of separation of ownership and control within a firm. They claimed that the separation of ownership and control may lead to a principal-agent problem as a result of asymmetric information problems between a firm’s executives and the shareholders. This problem arises since the CEO (agent) has better or more firm specific information than the shareholder (principal). Nevertheless, Jensen and Meckling (1976) were the first in formalizing the agency problem in a paper. According to them agency problems causes managers to prefer their own interests above the interests of the shareholders. As an example, risk-neutral shareholders are considered to be interested in getting a return on their investments, while the risk-averse CEO may also be interested in other objectives such as rapidly growing and large firm, or he uses the firm’s assets to meet his own private needs. In the end, by separating ownership and control, the CEO’s marginal benefit of his work effort does not truly reflect the marginal contribution to firm performance. As a result, a CEO could choose to neglect her duties by performing work effort which is sub-optimal to shareholder value. According to Ferrani and Molloney (2010), shareholders have to ensure themselves that the CEO will be prevented from opportunistic and inefficient behavior which could result in a large decrease in shareholder value. Tirole (2006) argues that a CEO, who puts insufficient effort on his work, engages in outrageous investments, takes benefits by self-dealing or uses an entrenchment strategy to keep his position safe could all result in decreasing shareholder value.
2.2 The executive contract
CEO compensation and firm performance in the Netherlands 7 First, Van Ees & Van der Laan (2007) claim that the capacities of a CEO differ from those of the general labor force. Though the question remains what these capacities are or should be, firms are looking for persons with a combination of characteristics which is more difficult to find than other employees they would hire. Second, only rarely there are vacancies for CEOs and because the supply side is small only a very small group of candidates is reviewed (Khurana, 2002). Consequently, it is hard to match supply and demand in this case. Furthermore, a managerial labor contract is not only an affair between the employer and the employee, since their compensation and delivered performance is to certain extent public available information. Finally, strong political powers operating in both the private sector (board meetings, shareholder meetings, and internal corporate governance) and the public sector, affect executive compensation practices.
Murphy (1999) assumes the existence of substantial heterogeneity within compensation practices between firms and industries. The different contract components can be divided in fixed and variable part, which will also be discussed in this section. Next to that, variable pay can be divided in a short term incentive pay and long term incentive pay. In general an executive pay contract most of the time contains the following basic components: a base salary, pension payments, annual cash bonuses, share options and/or share based bonuses. An overview of the executive contract components is presented in Figure 1.
Figure 1: Executive contract components
2.2.1 Fixed pay
CEO compensation and firm performance in the Netherlands 8 that base salaries are needed because in general executives are risk averse. However, this does not imply that they prefer the complete contract to consist of only fixed pay, but that they prefer a one dollar increase in their base salary to a one dollar increase in their target bonus or variable pay (Murphy, 1999). The base salary plays an important role in a determining firm’s compensation policy, since many other compensation components are derived from it. The short term incentive, the annual cash bonus, is mostly expressed as a percentage of the fixed salary while options and shares could be expressed as a multiple of this base salary. Next to that, also pensions and severance payments often depend on the base salary (Murphy, 1999). A CEO’s fixed pay is typically determined by competitive benchmarking. The most important determinant of the total level of compensation is the size of the firm. According to Tosi, et al. (2000) this accounts for more than 40% of the variance in total CEO compensation.
2.2.2 Variable pay
Besides fixed base salary, virtually all listed firms offer variable pay to their executives. Variable payments are based on short and/or long term performance. They are several ways to pay out the variable compensation. Annual bonuses, stock options and share based plans are the most common examples of variable pay. In setting the targets for variable compensation there are two important issues. First, the extent of influence a manager has on the outcomes of the performance measurements. For instance, market related measures are harder to influence than accounting performance measures. Second, it is important to congruence measures with a firm’s objectives since not all measures are in line with achieving the goals. In general terms, an increase in one aspect might lead to a decrease of the second aspect, whereas it is ideal that these measurements score high on both aspects (Van Ees, Van der Laan et al., 2007).
Annual bonus
CEO compensation and firm performance in the Netherlands 9 years. In the Netherlands there typically are both minimum and maximum levels of performance. Under the minimum level and above the maximum level no bonus is paid out. Between those levels, typically a linear pay-performance relationship is part of the executive contract. However, this pay-performance structure in the contract might lead to problems in the case of managers who already performed far above their capped target and attempt to save some of this performance for the next year. This could also be the case when they performed far below target and again try to save earnings for next year, resulting in an even worse performance for that particular year.
Share-based compensation
Shared-based compensation is the second main component of variable compensation. From an agency perspective, the objective of the executive contract is to align the interests of shareholders and managers in order to maximize the value of the firm (Jensen and Meckling, 1976). This can be done by creating managerial incentives that make executives act in the interests of shareholders. By providing a direct link between realized compensation and company share price performance (with equity-based compensation) this can be done effectively.
Equity-based compensation can be defined as a fixed value or a fixed number of shares. In general, the rewarded shares have some restrictions for the CEO. The shares are awarded under conditions and will be granted after meeting several year performance criteria. In the case of not meeting these criteria or in the case of over performing them, less or more shares than initially planned might be given to the executive. These restricted shares cannot be sold in the market, and typically the corresponding rights (voting, dividends, etc.) are granted to the owners. Bebchuk and Fried (2006) explain that a restricted stock is in fact an option on the firm’s assets with an exercise price of $0. Though, this explanation only holds for increasing share prices. With declining stock prices these shares still do have value while options do not.
Stock options
CEO compensation and firm performance in the Netherlands 10 the absence of a charge against accounting income, favorable tax treatment and the positive incentive effects, which arise from the close link between the level of compensation and share price development (Hall and Murphy, 2003). Furthermore, stock options provide an incentive to act more in the long term interest of the firm, because they usually can only be exercised after the expiration of a so-called vesting period. Stock options can also help to attract highly motivated and entrepreneurial employees, since payout is based on future performance and consequently the rewards can be much higher than with regular cash compensation (Hall and Murphy, 2003). The vesting of an option means that the option can be exercised from that moment on. Options are most of the time forfeited when an executive leaves the firm before vesting. Despite their popularity in the last decades, in recent years new forms of equity-based compensation have slowly replaced the stock option in the long term pay package (Swagerman & Terpstra, 2009).
2.3 CEO compensation and firm performance
When examining the relationship between CEO compensation and firm performance it is relevant to differentiate the impact of firm performance on compensation (ex-post compensation) from the impact of compensation on (future) firm performance and CEO behavior (ex-ante compensation). The first one defines the pay-performance relationship which is based upon real firm performance, while the second relationship examines the incentives of the CEO whether he will behave in the interests of the shareholders (van der Laan, 2010). This study focuses both on ex-post compensation and ex-ante compensation with the purpose to determine whether CEO compensation is determined by realized firm performance and whether CEO compensation influences firm performance in the future.
CEO compensation and firm performance in the Netherlands 11 Hypothesis 1: Total compensation depends positively on firm performance
A large body of corporate governance studies claims that conflicting interests between management and the shareholders could be solved through an appropriate CEO compensation package (Randøy and Nielsen, 2002 and Dalton et al, 2008). As it is obscure what an appropriate compensation package is, the relationships between the different parts of the compensation contract and firm performance will be investigated in this study. Although the existence of large heterogeneity in compensation practices across firms and industries, most CEO pay contracts are constituted of a fixed and a variable part as can be seen in figure 1. Several authors like Kato and Kubo (2003) and Murphy (1999) differentiate cash compensation (salary and bonus) and total compensation in their studies. However, with the objective to examine the impact of the different pay components this study follows the approach of van der Laan et. al (2010) to divide CEO compensation into fixed compensation, short term variable compensation (bonus), long term variable compensation (stocks and options) and total compensation. The part of the compensation contract that is fixed at the beginning of each year is known as the fixed salary (Core et al., 1999). Most studies which investigate the relationship between pay and performance do not expect any sensitivity in the salary as a result of firm performance. Van der Laan et. al (2010) argue that only indirect effects of performance on fixed compensation exist. In the case of a well performing CEO his variable pay component increases and therefore he faces more risk. That higher risk could be compensated with a risk premium which could be an increase in a CEO’s salary. Consequently, in this study no direct relationship between fixed pay and firm performance on the other hand is expected.
Hypothesis 2: Fixed compensation does not depend on firm performance
CEO compensation and firm performance in the Netherlands 12 Sloan, 1991). Performance measures which are based on accounting income (ROA) are backward looking whereas market based performance measures (stock return) also reflect future firm performance. The third hypothesis will test whether the bonus depends positively on firm performance.
Hypothesis 3: Short term variable compensation (bonus) depends positively on firm performance
Swagerman and Terpstra (2009) argue that long term variable compensation (stocks and options) has gained popularity in the last years in the Netherlands. This form of equity based compensation, particularly stock options, became important because of its ability to align CEO and shareholders’ interests. Authors like Core et. al (2003) support the advantages of this form of compensation in their studies. Furthermore, a well-known and widely cited paper Jensen and Murphy (1990) claims that a large part of the pay for performance relationship results out of equity-based CEO pay. Also according to Hall and Liebman (1998) a substantial part of the pay for performance relationship is related to changing values of a CEO’s stock and stock options holdings. As a result, the fourth hypothesis expects that long term variable CEO compensation (stocks and options) is positively determined by firm performance.
Hypothesis 4: Long term variable compensation (stocks and options) depends positively on firm performance
CEO compensation and firm performance in the Netherlands 13 is performing well (or under compensation in a poor market performance). Since a substantial part of a stock price movement is the consequence of market movements, the firm specific part of such a movement which reflects CEO performance can be doubtful.
Prominent studies like Hall and Liebman (1998) and Jensen and Murphy (1990) argue that the largest part of the pay for performance relationship is the result of this form of variable compensation and therefore also this study expects there exists positive relationship between long term variable CEO compensation and future firm performance.
Hypothesis 5: Future firm performance depends positively on long-term variable compensation (stocks and options)
2.4 Dutch remuneration practices
According to Roberts, Ceron et al. (2007), the Netherlands is known for its high level of disclosure compared to other European countries. Most firms in the Netherlands have a two‐tier board system, in which the board consists of the executive board and the supervisory board. The supervisory board of the firm has the task to act as sort of internal corporate governance mechanism. It is also their responsibility to set up management compensation practices and to monitor managers (van Ees, Postma et al., 2003).
CEO compensation and firm performance in the Netherlands 14 According to Swagerman and Terpstra (2009), executive pay levels in the Netherlands increased substantially in the beginning of this century. As a consequence of the popular equity bonuses, which aimed to align CEO incentives, CEO compensation rose strongly. Bebchuk and Friend (2004) claim that this trend is explained by the behavior of the consulting firms. They argue that these external firms have incentives to behave in favor of a CEO. Those firms are trying to ensure themselves with projects in the future in order to make money. To do so, it is of their interest to maintain a strong relationship with the principal (CEO) of the firm. Although supervisory boards expect these consulting firms to behave independent, in practice they do not. In addition, there is another trend which causes an upward shift of executive salaries. During the middle 1990s and the first years of the 21st century, firms reported huge profits and a common belief was that the only way to follow was up. To maintain or improve their competitive position it was of great importance to keep the best employees and next to that attract talented new employees at the firm. This belief causes firms to fight for the best people. This war of talent is associated with higher salaries for these people. Van Uffelen (2008) finds that executive compensation increased between 1988 and 2008 with an average of 7% per year, compared with a mere 2% increase for the average Dutch employee.
Examining the effects of higher CEO compensation, Duffhues and Kabir (2008) find no pay for performance relation in the Netherlands. They claim that CEO pay levels do not determine two year lagged firm performances. Furthermore, they also conclude that CEOs are able to influence their own compensation levels. However, their study only examines cash payments to CEOs and neglect equity-based compensation (stocks and options). Since it is expected that especially this equity-based compensation causes the pay for performance relationship, their study might be regarded as incomplete.
3
DATA AND METHODOLOGY
3.1 Data
CEO compensation and firm performance in the Netherlands 15 initial sample of 142 firms has been reduced after removing financial and utility firms, based on the Industry Classification Benchmark (ICB), and firms without any available data on CEO compensation. Furthermore, several firms were excluded because their primary listing was outside the Netherlands and thus CEOs of these firms would probably not conform to Dutch remuneration practices, since there was no material business activity in the Netherlands.
Eventually the dataset contains 105 firms listed on the Amsterdam Stock Exchange. However, the results should be interpreted with a sample bias in mind, since in total 39 firms have been removed and there is no data available for non-listed companies. The primary sources for the dataset are Thomson Reuters’ DataStream and Bureau van Dijk’s Amadeus. Data for most of the dependent variables and control variables are derived from these databases. Furthermore a website of the Dutch Investor Association (Vereniging van Effectenbezitters) has been used to obtain data on the variables concerning CEO compensation. This website does not only contain the total amount of compensation a CEO receives during a fiscal year, but also data on the different components of the compensation package. Finally, also annual reports are examined in order to complete the dataset.
The variables used to specify the model of the empirical analyses include several measures of CEO compensation, three firm performance measures and a set of control variables, which will be defined below. Table A in the appendix shows the definition and source(s) of all independent, dependent and control variables.
3.1.1 CEO compensation variables
CEO compensation and firm performance in the Netherlands 16 In some cases two or more CEOs received compensation during one year at a particular firm. In order to deal with this, the number of months the CEO worked for the firm has been investigated, and the compensation is annualized accordingly.
CEO compensation and firm performance in the Netherlands 17 Table 1
Descriptive statistics of CEO compensation variables CEO compensation in €1000
Year N Mean Median Min Max Std. Dev.
Fixed pay ( Salary)
2004 86 403 312 63 1,502 314 2005 84 407 320 49 1,928 345 2006 92 425 335 81 1,666 306 2007 94 448 367 96 1,785 323 2008 98 452 382 104 1,925 342 2009 102 450 394 122 1,612 318 2010 99 438 391 118 1,648 325 Pooled 655 433 359 49 1,928 325
Short term variable pay (Bonus)
2004 66 325 122 4 3,677 318 2005 72 311 152 6 1,938 455 2006 74 347 138 11 2,060 412 2007 78 411 162 14 3,600 609 2008 82 386 188 10 3,752 523 2009 81 362 192 12 3,042 468 2010 83 321 168 10 3,018 488 Pooled 536 353 162 4 3,752 434
Long term variable pay (Stocks & Options)
2004 38 488 221 2 6,165 1,012 2005 45 622 248 8 3,720 925 2006 49 667 317 8 4,468 1,013 2007 42 810 333 12 5,614 1,248 2008 46 790 406 10 5,012 1,109 2009 44 645 328 10 4,783 1,347 2010 50 632 340 14 4,826 1,167 Pooled 314 667 316 2 6,165 1,121
Total pay (Salary, Bonus, Stocks & Options)
CEO compensation and firm performance in the Netherlands 18 3.1.2 Firm performance variables
This study uses both accounting-based and market/equity-based proxies for measuring firm performance. The choice of these firm performance measures is not unimportant in this kind of research and there exists some guidance to which measures are used frequently and are appropriate. In here, the approach by Core et al (1999) to use return on assets (ROA), stock return and the Tobin’s Q as firm performance proxies is followed.
Return on assets is a widely used performance measure in corporate governance research (Brick et al, 2006). It is an accounting-based measure and is calculated as net income divided by average total assets during that particular year. Devers (2007) finds that accounting-based measures have a backward-looking character. Therefore, the second performance measure which is used equals the firm’s stock return. This is a market-based measure and is calculated as the natural logarithm of the total return index of a particular year divided by the natural logarithm of the total return index of the previous year. The third proxy for firm performance is known as the Tobin’s Q ratio. This one is a both an accounting-based and market-based performance measure. The numerator of this ratio equals the market value of assets and therefore also reflects the value which investors assign to the firm. The denominator of the ratio equals the book value of the firm’s assets and thus does not include any investor perception on firm value.
In summary, this Tobin’s Q ratio recognizes and reflects the growth opportunities a firm has. Consequently this measure is forward-looking A relatively high Tobin's Q suggests good performance in the sense that the firm has made investments and these are valued higher than the initial book value (Demsetz and Villalonga, 2001).
Table 2
Descriptive statistics of firm performance - and control variables
N Mean Median Min Max Std. Dev.
Firm performance variables
CEO compensation and firm performance in the Netherlands 19 3.1.3 Control variables
With regard to the control variables, the approach of several authors is used (Duffhues and Kabir, 2007; Core et al, 1999; Brick et al, 2006). These variables are included in the regression analysis in order to control for changes in firm characteristics and industry specific factors that might influence the pay-for-performance relationship. In this study, the control variables used are; firm size, leverage and time effects.
According to Tosi et al (2000) the largest part of CEO compensation is determined by the size of the firm. Also Murphy (1999) aruges that the pay-performance relationships differ as a result of varying firm sizes. Oxelheim and Randoy (2006) argue for example that this is not surprising, since larger firms, might be able to employ better qualified and thus better paid CEOs. Because firm size is such an important explanatory variable of firm performance it is necessary to use this control variable. It is measured as the natural logarithm of the average total assets of a firm at year t. The second and other proxy used is calculated as the natural logarithm the number of employees at year t.
According to the agency theory a high amount of debt within a firm can serve as a disciplining function for CEOs. More leverage is associated with more monitoring of the CEO by debt holders. Therefore, CEO compensation is expected to be lower at these firms (Jensen, 1989). In addition he argues that as a consequence of this debt disciplining function, firms with a higher leverage ratio are assumed to have less free operational cash flow. Therefore they should be less able to pay high compensations to a CEO. Leverage is measured as the book value of debt divided by the book value of total firm assets and reflects the capital structure of the firm. Next to that, time dummies are used to control for time specific shocks. Industry dummies are not included. Since this study uses a cross-sectional fixed model, which will be explained later, these are not necessary (Brooks, 2008). The sample firms and their Industry Classification Benchmark (ICB) classifications can be found in table B in the appendix.
CEO compensation and firm performance in the Netherlands 20 small (< 0.50) that possible multicollinearity problems can reasonably be ignored (Brooks, 2008). Finally, in this study each of the firm performance variable measures a different aspect of firm performance. As it is shown in the table, the highest correlation coefficient among the performance indicators is only 0.33
3.2 Methodology
To test CEO compensation as a function of firm performance the following OLS regression equation (1) is estimated.
CEO compensationi ,t = +
1 ROAi ,t +
2Stock returni ,t +
3Tobin’s Qi ,t +
4TAi ,t + 5
Employeesi ,t +
6Leveragei ,t + θ Time effectst
i ,t , (1) where α denotes the constant, the β’s denote the coefficients to be estimated and
it denotes the residual error term. ROA equals the return on assets of firm i in year t, Stock return is the average stock return of firm i during year t and Tobin’s Q is the Tobin’s Q ratio of firm i in year t. TA is the natural logarithm of the balance sheet value of total assets of firm i at the end of year t, Employees is the natural logarithm of the number of employees working in firm i in year t and leverage is the ratio of debt over total assets of firm i in year t, Time dummies (θ) are used to control for time specific shocks. In order to avoid a dummy variable trap, one time dummy will be removed from the equation (Brooks, 2008).Equation (1) captures the first four hypotheses to be tested. Next to that, lagged performance measures are used to account for the hypothesis that CEO compensation paid in year t might be determined by previous year’s firm performance (Duffhues and Kabir, 2007). According to Brick et al (2006), when the performance measures in the regressions are lagged with one year also a potential endogeneity problem will be reduced. This problem when measuring the contemporaneous relationships might occur since independent variables could be correlated with the error term. This could lead to biased results of the regression analysis. Though, this way of testing is often done in this field of research (Randøy and Nielsen, 2002; Duffhues and Kabir, 2007).
CEO compensation and firm performance in the Netherlands 21 shareholders. Equation (2) is estimated to determine the relationship between long term variable CEO pay (stocks and options) and (future) firm performance.
Firm performance = +
2 0 k k , 1
Long term variable payi,tk +
2TAi ,t +
3Employeesi ,t +
4Leveragei ,t + θ Time effectst
i ,t , (2)in which long term variable pay is the amount of stocks and options a CEO of firm i received for year t–k. Thus coefficient
1,k estimates the influence of stocks and options in year t–k on subsequent firm performance measures. Thus, besides measuring the contemporaneous relationship between CEO compensation and firm performance, the independent variables are lagged one and two years in order to account for the hypothesis that firm performance in year t is usually determined by CEO compensation paid in the previous years (Duffhues and Kabir, 2007). Furthermore, a potential endogeneity problem will be reduced when the performance measures in the regressions are lagged one or more years (Brick et al, 2006).CEO compensation and firm performance in the Netherlands 22
4
EMPIRICAL RESULTS
4.1 Firm performance as explanation for CEO compensation
Table 3 presents the results of the fixed effects model of equation (1) to gain insight into the first four hypotheses, using total compensation, fixed compensation and both short term and long term variable compensation for the full sample of firms. The regression specifications are similar except that one firm performance measure is replaced with another for each hypothesis. The number of firm year observation varies between 652 and 307. For the sake of brevity the regression estimates for the dummy variables that capture time effects are not reported.
Regarding table 3, the first outcome to notice is that except for the relationship of stock return on long term variable pay, all coefficients show positive signs. These positive signs are in line with the stated hypothesis for the pay-performance relationship. Moreover, they can be seen as evidence in favor of the agency theory and are in line with the results of Jensen and Murphy (1990). Randoy and Nielsen (2002) also found positive results. However, these results were mostly insignificant. Therefore, my outcomes are more in line with their work.
The results in table 3 show that the bonus of a CEO (short term variable compensation) is significantly positively related to the ROA of the firm in a particular year. The p-value (0.000) indicates a strong pay-performance relationship in this case and suggests that the bonus of a CEO is strongly based on the accounting income (ROA) of the firm. Using stock return as a market performance measure, no significant relationships are found. So based on these outcomes, none of the components of CEO compensation, is related to a firm’s stock performance on the financial markets. The third performance measure, defined as Tobin’s Q ratio, is as well an accounting - as a market performance measure. It can be seen that this ratio has a positive relationship with total compensation and all the specific components of CEO compensation. This relationship is significant for the variable compensation part with a p-value of 0.058 for the bonus part and a p-value of 0.042 for the value of a CEO’s stocks and/or options.
CEO compensation and firm performance in the Netherlands 23 for firm size, total assets, shows also significant positive relationships with salary and total compensation. Next to that, the value of a firm’s total assets is also positively associated with the short term variable CEO pay in the form of a bonus.
Finally, the leverage control variable shows mixed results with respect to the compensation components. Moreover, none of these results is significant. The arguments of Duffhues and Kabir (2007) that associate an increase in leverage with more risk and as a result of that higher compensation are not confirmed in this study. Also the disciplining function of debt, stated by Jensen (1989), is not confirmed here.
Table 3
Contemporaneous relationship between firm performance and CEO compensation
Fixed Short term variable Long term variable Total Salary Bonus Stocks & Options Compensation
Variable Coefficient Coefficient Coefficient Coefficient
(Prob.) (Prob.) (Prob.) (Prob.)
ROA 0.156 1.818 0.844 0.254 (0.651) (0.000)*** (0.232) (0.499) Stock return 0.031 0.105 -0.049 0.082 (0.897) (0.457) (0.720) (0.910) Tobin's Q 0.084 0.642 1.355 0.452 (0.776) (0.058)* (0.042)** (0.083)* Employees 0.946 0.055 -0.048 0.141 (0.002)*** (0.458) (0.667) (0.004)*** Total Assets 0.461 0.836 0.173 0.097 (0.044)** (0.087)* (0.742) (0.041)** Leverage -0.064 -0.005 0.451 -0.054 (0.389) (0.810) (0.292) (0.586) Constant 9.515 11.237 8.648 10.022 (0.000)*** (0.000)*** (0.000)*** (0.000)*** Adj. R-squared 0.72 0.84 0.66 0.69 F-statistic 10.581*** 18.468*** 12.159*** 21.263*** N 651 528 307 652
CEO compensation and firm performance in the Netherlands 24 4.1.1 Lagged relationship
Lagged performance measures to account for the hypothesis that CEO compensation paid in year t is usually determined by previous year’s firm performance (Duffhues and Kabir, 2007) are also used (Table 4). According to Brick et al (2006), when the performance measures in the regressions are lagged with one year also a potential endogeneity problem will be reduced.
Table 4
Lagged relationship between firm performance and CEO compensation
Fixed Short term variable Long term variable Total Salary Bonus Stocks & Options Compensation
Variable Coefficient Coefficient Coefficient Coefficient
(Prob.) (Prob.) (Prob.) (Prob.)
ROA 0.056 0.652 0.998 0.094 (0.771) (0.271) (0.074)* (0.755) Stock return -0.092 -0.091 0.617 -0.302 (0.547) (0.639) (0.293) (0.638) Tobin's Q -0.043 0.092 1.206 -0.004 (0.878) (0.668) (0.062)* (0.879) Employees 0.756 0.421 0.067 0.155 (0.041)** (0.188) (0.805) (0.000)*** Total Assets 0.771 0.937 0.129 0.081 (0.000)*** (0.094)* (0.852) (0.077)* Leverage 0.024 -0.082 -0.002 -0.008 (0.727) (0.664) (0.919) (0.603) Constant 10.286 13.855 11.088 11.835 (0.000)*** (0.000)*** (0.000)*** (0.000)*** Adj. R-squared 0.69 0.68 0.71 0.72 F-statistic 13.587*** 21.929*** 19.005*** 21.447*** N 574 442 239 576
Note: The table presents cross-sectional fixed effects regression results. Time dummies are included but these are not reported for the sake of brevity. ***, **, * significant at respectively the 1%, 5% and the10% level (2-tailed).
CEO compensation and firm performance in the Netherlands 25 insignificant. Therefore, the argument of Duffhues and Kabir (2007) that CEO compensation is determined by the firm’s performance in the previous year is not completely proven with these data. The negative coefficients of the stock return in previous years are contradicting my hypothesis that compensation is partly determined by a firm’s stock return. However, these results are far from significant. The third performance measure, Tobin’s Q ratio, shows a significant (p-value of 0.062) relationship with long term variable compensation. This result is not surprising since this compensation part equals the value of the shares and options granted to a CEO.
The control variables show almost the same results as the contemporaneous model. The number of employees is positively associated with fixed pay (salary) and total compensation. These positive relationships are also found for the other size proxy which is determined by a firm’s total assets of previous year. Likewise for leverage, the same results are found with the lagged model. Positive and negative insignificant results are shown for the leverage relationship.
The results of both the contemporaneous and the lagged model show a high explanatory power. This is expressed in the relative high R² values (0.66 – 0.84) and the significant F-statistics. However, the R² values should be interpreted with care as a result of the use of cross-sectional fixed effects estimations. In this case the model derives a constant for every firm in the regression and therefore the R² is artificially high.
4.2 CEO compensation and future firm performance
CEO compensation and firm performance in the Netherlands 27 With regard to the control variables, significant positive relationships between firm size and the contemporaneous ROA and stock return of a firm are found. Both proxies for size (employees and total assets) are positively associated with ROA and stock return in the same year. Also in this model the coefficients for leverage show mixed results and are insignificant. Therefore, the agency cost hypothesis of Jensen (1976) is not supported in this study, since he states that higher leverage is associated with better performance.
Finally, the inclusion of lags might cause a sample bias, because the sample size has been reduced in this model. The relative high R² values (0.64 – 0.78) and the significant F-statistics should be interpreted with care as a result of the use of cross-sectional fixed effects estimations. In this case the model derives a constant for every firm in the regression and therefore the R² is artificially high.
5
REFLECTION ON THE DUTCH SITUATION
It seems to be the case that public firms have a hard time in explaining why bonuses are paid. The underlying targets and measures are more or less concealed so that the public does know that a bonus has been paid, while they cannot gauge what the underlying achievement of each particular managers or executive is. Firms argue that they cannot disclose the information to such a level of detail as the disclosure of measures and targets would give away too much information to the firm’s competitors (Indjejikian and Nanda, 2002). In this section the public debate about remuneration practices in The Netherlands will be reflected. Fist, by focusing on the discussion about bonuses in the financial sector and thereafter the situation in the public sector will be discussed.
5.1 Bonuses in the financial sector
CEO compensation and firm performance in the Netherlands 28 the bonus payments to the executives. So, why do we have this debate about the bonuses of firms like ING?
The social and public debate about bonuses at banks strongly deteriorated this year. The smallest form of government aid is a nowadays a reason to refuse bonuses. All the extras that bankers become are examined carefully. The Dutch House of Representatives was furious about the proposed bonus to the executives of ING, while the bank did nothing else than following the rules which had been established after obtaining government support. Total variable pay should not exceed one year salary. CEO Hommen of ING, received in 2010 € 1.25 million variable compensation where his fixed salary equaled € 1.35 million. In the business ING operates in, this is not an exorbitant amount. ING was in 2010 the third firm of the Dutch AEX index based on market capitalization after Royal Dutch Shell and Arcelor Mittal. CEOs of smaller firms like Philips, AkzoNobel, Heineken and KPN all earned more than Hommen.2 Moreover, Hommen did not receive any variable pay in 2009 and finally he voluntary refused his bonus over 2010.3 For comparison, Lloyd Blankfein of Goldman Sachs and Jamie Dimon of JP Morgan received respectively 9 and € 12.8 million in 2010.4 Not unimportant; they both received government aid.
The ING bonus became controversial because the society en politicians changed the playing rules. No variable compensation as long as there is government support is the new overall opinion. Moreover, this thought is expanding at this moment. Initially, the Dutch Minister of Finance and other politicians only considered the capital injections at ING, Aegon and SNS as government support, whereas a few months later also the guarantees for ING’s Alt-A-mortgages felt within that definition. Since those guarantees will continue for years, this implicates that top management of ING is forbidden to receive variable pay for this period. The definition of government aid extended even further by including bonds with government guarantee. Remarkably, these bonds were issued at the end of 2008 because there was much financial distress in financial markets. Encouraging banks to help firms in financial trouble was underlying this emission. At that time this action was not considered as government aid.
CEO compensation and firm performance in the Netherlands 29 aftermath of the financial crisis. That’s why the Code Banken starts with the role of these supervisors. However, in the public debate about the bonuses within ING, the supervisory board kept quiet. One may regard this as cowardly, since especially Hommen was attacked in the media while he was expected to lead the company through socially and financially though times. The ING supervisory board could be blamed for the fact that they did not foresee that both politicians and citizens would not accept those bonuses at that moment. They should have explained their bonus decision in a press conference to play the public opinion. Since they did not, the whole Dutch society felt over Hommen. After a tumultuous debate in the House of Representatives he handed in his bonus accompanied with an open letter in a Dutch paper. Head of the supervisory board Jeroen van der Veer (former CEO of Royal Dutch Shell), Lodewijk de Waal (former chairman of the Dutch FNV) and other members kept silent again. As they were afraid of being attacked by the press, this was a conscious choice. As a result, the important role of these cowardly behaving supervisors in defining the new banking might be questionable.
With regard to variable compensation the Code Banken contains a rule that the bonus should not exceed the fixed salary. However, about 600 employees of large international oriented Dutch banks do not meet that rule. Often this is the case for specialized positions in trading and mergers/acquisitions practices. 70 percent of these 600 people are working outside The Netherlands.5 Especially in Asia bankers prefer a lower base salary in combination with higher variable pay. So, if the Code is complied strictly, there is a risk that these jobs disappear to foreign countries. Therefore, we have to offer these specialized financial talents a competitive remuneration package to attain and retain them, while we have to keep an eye on bankers’ pay in general.
Finally, a more general longer term solution which is mentioned is a separation between banks which manage private saving money (retail banks) en banks which do not (investment banks). The retail banks should be monitored more accurately with stricter rules about for example compensation than the investment banks. The retail bank with the private saving money is then able to obtain government support whereas investment banks never can claim support. Though, this separation encounters much resistance of the Dutch Central Bank (DNB) and other banks, but further elaborating on this would be behind the scope of this study.
5
CEO compensation and firm performance in the Netherlands 30 5.2 Compensation in the Dutch public sector
The level of top salaries in the Dutch public sector is the subject of a returning public discussion. According to the de Balkende-norm (a voluntary norm prepared in 2006 and named after Prime Minister Balkenende) public executives are not allowed to earn more than 130 percent of a minister salary. In 2011 this norm equaled 187.340 Euro and a maximum expenses allowance of 7.560 Euro. However, this Balkenende-norm is not that hard and is characterized by a ‘comply or explain’ principle. The only requirement of the government is to publish salaries above the norm. The idea behind this principle is known as self-regulation. Meanwhile it is clear that this does not work. A lot of cases where executives were paid (far) above the norm passed by in the Dutch media and the public debate heated again.
Since self-regulation does not work, the government intervenes by proposing a new act (WNT). With this new act the Balkenende-norm gets a more compulsory character, since it simply will be forbidden to pay executives more than the norm. Next to that, severance payments are kept in bounds with a maximum of 75.000 Euro. Organizations which do not stick to these rules are expected to pay back the difference within three weeks and will be fined. The act distinguishes three regimes: (1) semi-public organizations like housing associations and educational institutions, (2) hospitals and (3) insurers. For the first and second regimes maximum salaries are determined. The third regime (insurers) is free of any norm and is only obliged to publish about their compensation practices. The WNT is already approved by the House of Representatives and it is expected that the Upper House will agree as well.
The distinction between the three regimes is meaningful, since the exact difference between the public, semi-public and private sector is often used inappropriate in the discussion about compensation practices within these sectors. Former chairman of the Dutch financial markets supervisor (AFM), Hoogervorst, defended his compensation of 420.000 Euro by explicitly stating that he operated in the semi-public sector. This reference to the market is misguiding for some reasons. Supervisors like the AFM are, although at a distance, part of the Dutch government. Consequently, they are fully paid off tax money. Moreover, they are little or not exposed to revenues and triggers out of the market and so the comparison with commercial organizations is not valid. The main challenge in The Netherlands is to stop these public and political debates about compensation practices in the public sector. We have to find out whether the Balkenende-norm is still realistic.
CEO compensation and firm performance in the Netherlands 31 professionalize a sector, experienced top managers are needed who might be judged on their results. With a maximum salary that is far below salaries in the private sector it will be difficult to attract well-grounded people. Moreover, the norm is contra productive since talented officials move to commercial organizations or rent themselves against market prices. In the end this is more expensive than a well-paid experienced top official who knows the ins and outs for his job. Next to that, the Balkenende-norm could provoke corruption. Officials who are dissatisfied with their salary are more vulnerable to gifts and corruption. Finally, for jobs in the public sector there exists a market. It is foolish to impose the salary of the only function for which no market exists to the complete public sector.
Increasing the Balkenende-norm could be a first step in the right the direction. In a constitutional monarchy the function of Prime Minister is the most important. No other officer is carrying such broad responsibility. And since it is rewarded with a relative low salary, every manager in the public sector who earns more is publicly attacked. Therefore, the discussion should focus more on the level of the Balkenende-norm than on the remuneration of well-performing managers in the public sector.
6
CONCLUSION
This study examines the two-sided relationship between CEO compensation and firm performance. In order to do so, it uses unique panel data of firms listed on the Dutch stock exchange from 2004 to 2010. The analysis is based upon individual CEO compensation components which include fixed salary, short term variable pay (bonus), long term variable pay (equity-based compensation) and total compensation. First, it is examined whether the compensation which a CEO receives depends on firm performance (ex-post compensation) and next to that the influence of CEO pay practices on firm performance in the future is investigated (ex-ante compensation). The agency theory, formalized by Jensen and Meckling (1976) is the point of departure for this study since they argue that CEOs prefer their own interest above the interest of the shareholders as a result of agency problems. Interest alignment of CEOs and shareholders is a solution to this problem. Linking compensation to firm performance is therefore a way to do so.
CEO compensation and firm performance in the Netherlands 32 However, no significant relationship between accounting income (ROA) and total compensation is found. These results partly contradict the findings of Duffhues and Kabir (2007) who found an even negative relationship between accounting income and CEO pay. An explanation could be the fact that they conducted their research in the period before the introduction of the Dutch Corporate Governance Code in 2003. One of the purposes of this code was to align interests by for example a pay for performance structure for CEOs. The second hypothesis which expects no direct relationship between firm performance and fixed CEO pay is confirmed since no significant positive or negative relationships are found. This result is in line with various studies in which also no relationship between accounting and/or market returns and fixed salary is shown.
With respect to the relationships between short - and long term variable pay and firm performance some significant results are found. First, the third hypothesis is confirmed since it is shown that ROA is a positive determinant of short term variable pay (bonus). This is in line with findings of Core et al (2003) and Jensen and Murphy (1999). Furthermore, this is evidence for the prediction that a bonus is mainly affected by accounting based performance measures such as ROA which have a backward-looking character. Next to that, it is found that a firm’s Tobin’s Q ratio is a significant positive determinant of the long term variable compensation component (stocks and options) of CEOs. This result is not surprising, since the market value of equity is part of the Tobin’s Q ratio and stock and option prices are also based upon this market value of equity. According to Demsetz and Villalonga (2001), this measure indicates the growth opportunities of a firm and is based on the beliefs of investors regarding this future profitability. Therefore, the Tobin’s Q ratio can be assumed as an appropriate determinant of CEO pay since it can be used to align interests between shareholders and CEOs.
CEO compensation and firm performance in the Netherlands 33 As argued by Jensen and Meckling (1976) a way to mitigate agency problems between shareholders and management is the alignment of interests between them by linking compensation to future firm performance. The results in this study partly support this theory. The two year lagged relationship between long term variable pay and stock return is significantly positive. So, a firm’s stock return in year t depends positively on the value of stocks and options granted to a CEO in year t-2. This finding confirms the arguments of Bebchuk and Fried (2004) who also claim that long term variable pay is meant to align interest between the shareholders and the CEO. Next to that, the two year lagged relationship between long term variable pay and the Tobin’s Q ratio is also positive and significant, which is in line with the above results. As regards the one year lagged and the contemporaneous relationships, no significant results are found. This suggests that long term variable pay (stocks and options) determines future performance with a lag of at least two years.
Since this study only partly shows that future firm performance positively depends on long term variable pay components of the executive contract, it is hard to form an opinion about the level of CEO compensation in the Netherlands. With regard to the bonuses in the financial sector, the public debate should be more nuanced. For example, CEOs of foreign banks or even Dutch multinationals of the same size receive extremely higher bonuses. Focusing on bonuses for the more aggressive bankers who sell those highly complicated financial products to consumers would be more justified. With respect to the situation on executive compensation in the public sector, increasing the Balkenende-norm could be a first step in the right the direction. In the Dutch constitutional monarchy the function of Prime Minister is the most important. No other officer is carrying such a broad set responsibilities. Since it is rewarded with a relative low salary, every manager in the public sector who earns more is publicly attacked. Therefore, the discussion should focus more on the level of the Balkenende-norm than on the remuneration of well-performing managers in the public sector.
CEO compensation and firm performance in the Netherlands 34 every firm in calculating the value of the stock options a CEO receives, since it would be complicated and time-consuming to do not. This means for example that for every firm in the sample the same stock volatility rate is used. As a result, differences between the real option values and the calculated option values come into existence. Consequently, these differences could reduce the validity of the results of the regression analysis. Finally, this study did not control for potential endogeneity problems which are assumed to be present in most of the corporate governance studies. In the case of such problems, independent variables are correlated with the error term. Further research could try to solve this endogeneity problem by including instrumental variables that have to be exogenous, and then use a two-stage-least-squares (2SLS) regression analysis.
Further research should try to solve these limitations and elaborate further on these results. An interesting topic could be the level of managerial entrenchment in corporate governance in the Netherlands. There is little evidence on this theme for the Netherlands with its unique corporate governance system. The question would be whether managerial entrenchment influences Dutch CEO compensation practices. It is interesting to examine whether CEOs are able to compose the benchmark group and what the consequences are for their own pay.
CEO compensation and firm performance in the Netherlands 35
REFERENCES
Baker, G. and B. Hall (2004), “CEO Incentives and Firm Size”, Journal of Labor Economics 22: 767-798
Bebchuk, L. and J. Fried (2003), “Executive Compensation as an Agency Problem”, Journal of Economic Perspectives 17: 71-92
Bebchuk, L. and J. Fried (2004), “Pay without Performance: The Unfulfilled Promise of Executive Compensation”, Harvard University Press, Cambridge
Berle, A. and G. Means (1932), “The Modern Corporation and Private Property”, Harcourt, Brace and World, New York
Brick, I., O. Palmon and J. Wald (2006), “CEO compensation, director compensation, and firm performance: Evidence of cronyism?”, Journal of Corporate Finance 12: 403-423
Brooks, C. (2008), “Introductory Econometrics for Finance”, University Press, Cambridge Brown, C. and J.L. Medoff (2003), “Firm Age and Wages”, Journal of Labor Economics 21: 667-697
Cools, K. (2006). “Controle Is Goed, Vertrouwen Nog Beter. Over Bestuurders en Corporate overnance”, Van Gorcum, Assen
Core, J., W. Guay and D. Larcker (2003), “Executive Equity Compensation and Incentives: A Survey”, Economic Policy Review 3: 27-50
Core, J., R. Holthausen and D. Larcker (1999), “Corporate governance, chief executive officer compensation, and firm performance”, Journal of Financial Economics 51: 371-406
Dechow, P. and R. Sloan (1991), “Executive Incentives and the Horizon Problem”, Journal of Accounting and Economics 14: 51-89
Demetz, H. and B. Villalonga (2001), “Ownership structure and corporate performance”, Journal of Corporate Finance 7: 209-233
Devers, C., G. Reilly and M. Yoder (2007), “Executive compensation: A multidisciplinary review of recent developments”, Journal of Management 33: 1016-1072
Duffhues, P. and R. Kabir (2007), “Is the pay-performance relationship always positive? Evidence from the Netherlands”, Journal of Multinational Financial Management 18: 45-60 Fama, E. and M. Jensen (1983), “Agency Problems and Residual Claims”, Journal of Law and Economics 26: 327-349
CEO compensation and firm performance in the Netherlands 36 Hall, B. and J. Liebman (1998), “Are CEOs really paid like bureaucrats?”, Quarterly Journal of Economics 113: 653-691
Hall, B. and K. Murphy (2003), “The Trouble with Stock Options”, Journal of Economic Perspectives 17: 49-70
Healy, P.M. (1985), “The Effect of Bonus Schemes on Accounting Decisions”, Journal of Accounting and Economics 7: 85-107
Indjejikian, R. and D. Nanda (2002), “Executive target bonuses and what they imply about performance standards”, Accounting Review 77: 793–819
Jensen, M. and W. Meckling (1976), “Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure”, Journal of Financial Economics 3: 305-360
Jensen, M. (1989), “Eclipse of the Public Corporation”, Harvard Business Review 67: 61-75 Jensen, M. and K. Murphy (1990), “Performance Pay and Top-Management Incentives”, Journal of Political Economy 98: 225-264
Kato, T. and K. Kubo (2003), “CEO compensation and firm performance in Japan: Evidence from new panel data on individual CEO pay”, Journal of Japanese International Economies 20: 1-19
May, D. (1995), “Do Managerial Motives Influence Firm Risk Reduction Strategies?” Journal of Finance 50: 1291-1308
Miller, D. (1995), “CEO salary increases may be rational after all: referents and contracts in CEO pay”, Academy of Management Journal 38: 1361-1385
Murphy, K., (1999), “Executive compensation”, Handbook of Labor Economics vol. 3, Amsterdam
Oxelheim, L. and T. Randøy (2006), “The Anglo-American financial influence on CEO compensation in non-Anglo-American firms”, Journal of International Business Studies 36: 470-483
Randøy, T. and J. Nielsen (2002), “Company Performance, Corporate Governance, and CEO Compensation in Norway and Sweden”, Journal of Management and Governance 6: 57-81 Roberts, V. and M. Ceron (2007), "Executive compensation disclosure in Europe", Mercer Limited, Londen
Shleifer, A. and R. Vishny (1997), “A survey of corporate governance”, Journal of Finance 52: 737- 783
Swagerman, D. and E. Terpstra, (2007), “The Effectiveness of Dutch Executive Pay Packages”, Compensation & Benefits Review 39, 47-57
CEO compensation and firm performance in the Netherlands 37 Tirole, J. (2006), “The Theory of Corporate Finance”, Princeton University Press, Princeton Tosi, H., S. Werner, J. Katz and L. Gomez-Mejia (2000), “How Much Does Performance Matter? A Meta-Analysis of CEO Pay Studies”, Journal of Management 26: 301-339
Van der Laan, G., H. van Ees and H. van Witteloostuijn (2010), “Is Pay Related to Performance in the Netherlands? An Analysis of Dutch Executive Compensation, 2002–2006”, The Economist 158: 123-149
Van Ees, H., M. Postma, et al. (2003), "Board characteristics and corporate performance in The Netherlands", Eastern Economic Journal 29: 41‐58
