Executive Compensation and Firm Performance: Evidence from the Netherlands

(1)

Executive Compensation and Firm

Performance: Evidence from the

Netherlands

MSc in Business Administration - Organizational and Management

Control

Master Thesis

Student: Janine Legtenberg, s2554747

Thesis Supervisor: Prof. Dr. Ir. P.M.G. van Veen-Dirks Co-assessor: N.J.B. Mangin

June 21, 2018

Word count: 8980 (including references and appendices) UNIVERSITY OF GRONINGEN

(2)

Abstract:

This study examines the relationship between executive compensation and firm performance as a solution to the principal-agency problem. It also incorporates the strength of corporate governance and its effect on firm performance. Data concerning the executive remuneration and firm performance consists of Dutch listed companies for the period of 2004 till 2016, and is analysed for whether executive compensation is affected by firm performance. Executive compensation is measured using total compensation and the bonuses separately, and firm performance is measured using the accounting measures return on assets and return on equity. The analysis also encompasses lagged relationships and the financial crisis of 2008. The results show a positive relation between the compensation and the firm’s performance. The findings support the previous research by suggesting that executive compensation can resolve the principal-agency problem.

(3)

1. Introduction

Executive compensation has already been a topic of discussion for many years among researchers, employees, regulators, and the press. The number of establishments using some form of performance related pay have increased considerably (Bayo-Moriones et al., 2017). The executive compensation has been transformed to compensation packages which incorporate some form of variable pay, that are mostly dependent on the performance of the firm. Performance related pay is said to be the solution to the principal-agency problem (Bebchuk and Fried, 2003). The principal-agency problem is considered to be one of the biggest issues in management accounting (Jensen and Murphy, 1990). By aligning the interests of the agent, the manager, to the interests of the principal, the owners, firms can minimize or even erase this problem. Apart from this, Kaydos (1999) provides five reasons why companies should actually measure the performance of their employees, namely to improve control, to set clear

responsibilities and objectives, to ensure strategic alignment of objectives, to understand the business processes, and finally to determine process capability. Although much research has already been performed, the relationship between executive compensation and firm performance still remains unclear (Duffhues and Kabir, 2008).

Previous research found a, although relatively small, significant positive relation between executive compensation and firm performance (Jensen and Murphy, 1990; Mehran, 1995; Conyon and Schwalbach, 2000). Others were unable to find any significant relation (Izan et al. 1998) or found a negative relation (Duffhues and Kabir, 2008). Thus so far, the results of the research on effects of executive compensation and firm performance are giving mixed results. Also, even though there is considerable literature already in existence, most of the research is performed on US or UK data (Izan et al., 1998; Braam and Poutsma, 2015). Important to consider is that this data is collected in countries in which the corporate governance systems are stated to be Type I, a shareholder perspective, while in Europe, the corporate governance systems are stated to be Type II, a stakeholder perspective (Conyon and Schwalbach, 2000). According to the literature, the executive compensation packages may differ between countries as a result of different corporate governance systems (Conyon and Schwalbach, 2000). This might be the reason why Duffhues and Kabir (2008), who gathered their data from Dutch companies, found a negative relation between executive compensation and firm performance, while Jensen and Murphy (1990) and Mehran (1995) found a positive relation. Both countries are subjected to a different corporate governance system, with different strengths of corporate governance, which leads to differences in the effects of corporate governance on firm performance. Researchers have found a significant positive relation between the strength of corporate governance and firm

(5)

5 last couple of years. According to the newly adopted Tabaksblat Code of 2004, all listed companies in the Netherlands are obliged to report the remuneration of the Executive Board, including the

performance criteria (Dutch Corporate Governance Code, 2016). Due to this Code, corporate

governance was strengthened in the Netherlands (Swagerman and Terpstra, 2007). Previous research in the Netherlands concerned the period before this adoption, thus before 2004, at which time the corporate governance in the Netherlands was relatively weak. According to Core, Holthausen and Larcker (1999) there is a significant relation between the strength of corporate governance and the amount of compensation received by the executive. Stronger corporate governance indicates a better control of the executive board, in which case the shareholders or supervisory board should be better able to control the actions of the executives. As the compensation packages of the executives are a form of corporate governance, it would mean that as a result of stronger corporate governance, a better alignment of the interests between the principal and the agent should exist, leading to a positive relation between executive compensation and firm performance.

Therefore, it is of interest to investigate whether, since the adoption of the Tabaksblat Code, a positive relation can be found between executive compensation and firm performance in the Netherlands. This leads to the following research question:

What is the relation between executive compensation and firm performance in the Netherlands after the adoption of the Tabaksblat Code?

To investigate this relation, data will be gathered concerning Dutch companies listed on the AEX from the period 2004 till 2016. The year 2004 is the first year since the adoption of the Tabaksblat Code and 2016 the last year of available data. Previous research only collected data from relatively short time-periods (Swagerman and Terpstra, 2007; Duffhues and Kabir, 2008), meaning that this study

differentiates itself by collecting data over a longer time-period. In addition to the analysis of the effect of this year’s firm performance on this year’s executive compensation, also the effect of last year’s firm performance on this year’s executive compensation is measured. Jensen and Murphy (1990) suggest that as this year’s firm performance is usually still unknown at the time when compensation is set, but last year’s performance is given, compensation is set using last year’s performance results. Also, further analysis is performed considering the effects of the financial crisis of 2008, which might have affected the firm’s performance. As far as I know, this is the first study to investigate the

relationship between executive compensation and firm performance since the adoption of the Tabaksblat Code. Therefore, this study contributes to the understanding of the role of executive remuneration, as it provides new insights into the relation with firm performance in a strong, Type II, corporate governance setting.

(6)

6 explanation of the Tabaksblat Code and corporate governance in the Netherlands. Section 3 describes the methodology, followed by the data analysis. The results of the study are presented in section 4, after which the final section will provide the conclusion and limitations of this study, together with some implications for future research.

2. Theory

This section will provide a short overview of the literature and theory so far. First, the solution to the agency problem will be explained, including the motivation behind the executive pay for performance. This is followed by a short overview of the literature on executive compensation in different countries. Thereafter, an indication is given as to why the Netherlands is different from countries researched so far, for example the US and the UK. Finally, the hypotheses will be presented about the expected effects of pay for performance for executives in the Netherlands.

2.1 Agency problem and executive compensation

So far, literature mostly focused on the relation between CEO compensation, most importantly pay for performance and incentive contracts, and firm performance within the US and the UK (Core et al., 1999; Jensen and Murphy, 1990; McConaughy, 2000; Carpenter et al., 2001). The conflict between the shareholders of a publicly owned organisation and its Chief Executive Officer (CEO) is considered to be one of the most important principal-agent problems (Jensen and Murphy, 1990). Given that the shareholder cannot observe the CEO’s activities and the firm’s investment opportunities, agency theory suggests a compensation scheme which will align the incentives on the CEO with those of the shareholders. Financial economists label this the ‘optimal contracting approach’ (Bebchuk and Fried, 2003; Ntim et al., 2015). By designing these compensation schemes, boards assume that they can provide managers with the efficient incentives to maximize shareholder value (Bebchuk and Fried, 2003). According to the optimal contracting theory, there are several components in the compensation packages used to align the interests of the executives to those of the shareholders. These are the following:

● Stock ownership: By increasing the stock ownership of the executives, the compensation of these executives is closely related to the stock returns of the company. This way, the conflict of interest between the executives and stockholders can be diminished.

(7)

7 ● Long Term Incentive Plans (LTIP): These plans are usually based on stocks and options, with

the condition of reaching some predetermined target performance measures. Thus, the executive can only exercise the options or is only granted the stocks in case the executive performs according to the performance measures.

● Bonus plans: Executives can earn a cash bonus, which is determined using accounting performance measures and often some non-accounting performance measures. The bonuses are rewarded when the executive reaches the performance targets and are usually expressed as a percentage of their base salary.

Additionally, the compensation packages usually also consists of other components, like a base salary, pension rights, and/or maybe social and healthcare contributions. These compensation packages will result in increases in shareholders’ wealth, as executives will focus on improving firm performance, which will result in higher returns for the executives personally, but will also result in higher

shareholder returns (Jensen and Murphy, 1990). Furthermore, following the expectancy value theory, it is believed that employees who receive these packages will make extra efforts, since they will expect extra rewards from these efforts (Braam and Poutsma, 2015).

According to Antle and Smith (1986) and Jensen and Murphy (1990), there are three primary

mechanisms included in variable compensation packages of executives, namely 1) flow compensation, which are the CEO’s annual salary, bonus, new equity grants, and other compensation; 2) the changes in the value of the CEO’s stocks and options; and 3) the possibility that the CEO’s human capital will decrease according to the market’s assessment after termination due to poor performance (Guay et al. 2002).

Thus all in all, aligning the interests of the managers with the interest of the shareholders by means of the performance related compensation packages, should result in a solution to the agency problem. However, how does this lead to a higher firm performance?

2.2 Corporate governance and firm performance

(8)

8

2.3 Executive compensation and firm performance

While much research has been performed concerning the relation between corporate governance and firm performance, the evidence concerning a relation between executive compensation and firm performance has been relatively limited. Based on US data, Mehran (1995) found a positive relation between the firm’s performance, measured using the Tobin’s Q and return on assets, and the

percentage of equity-based executive compensation, but also between firm performance and the percentage of equity held by managers. Furthermore, Jensen and Murphy (1990) found a, although relatively small, positive relationship between pay-performance and shareholders’ wealth. They suggest that the compensation policy set, which relates the CEO’s wealth and welfare to the shareholders’ wealth, aligns the private and social benefits and costs, resulting in incentives for the CEO to make the appropriate decisions concerning the actions to be taken. Core et al. (1999) found that in firms where the corporate governance is relatively weaker, the executives receive a higher compensation. Additionally, they found that this result leads to a lower firm operation and stock return performance.

Apart from literature based on US and UK data, the effects of executive compensation on firm

performance in other parts of the world is also tested, although the evidence is relatively less compared to the US and UK. Hill et al. (2011) compared executive employment contracts between the US and Australia and found that the base salaries from the Australian executives are significantly greater compared to their US counterparts. Also, the US CEOs are more commonly compensated with stock options and restricted stock compared to the Australia CEOs. Additionally, Ntim et al. (2015) found a small, significant, positive relation between executive compensation and firm performance for South African listed firms. They investigated the relationship after the implementation of King’s report (King II), an expansion of the corporate governance best practices of King I, which includes details concerning executive remuneration. King I served as the first step to improving corporate governance practices in South Africa, but failed to include the composition and independence of the Remuneration Committee, whose job is to recommend executive’s compensation, and also did not include the structure and possible involvement of the shareholders in the modification of the executive’s pay. King II included the missing factors of King I by expanding with a more detailed section concerning executive’s remuneration (Ntim et al., 2015).

Furthermore, Izan et al. (1998) stated that they could not find any significant relation between CEO pay and a wide variety of share price and accounting performance measures. This is consistent with the existing Australian literature. However, the existing literature in Australia very limited, just like the literature concerning executive compensation in Europe. Conyon and Schwalbach (2000)

(9)

9 who are not much involved with the company, and Type II, where the shareholders/equity holders are involved in the company at the long-term. The UK and US are economies typically stated as examples of Type I governance systems, while Germany and Japan are examples of Type II governance systems. Conyon and Schwalbach (2000) found a significant positive relationship between the company’s performance and executive pay. They also stated that chief executives in the UK seem to have a higher compensation compared to the German chief executives.

Thus, while considerable research has already been performed on the relation between executive compensation and firm performance, the concrete evidence of the relationship is still very mixed.

2.4 Additional characteristics concerning the relation

While much literature exists which considers the overall effect of executive pay for performance, also many focused on specific characteristics, like the industry, economic factors or the knowledge of the manager compared to his or her supervisor. Anderson et al. (2000) compared the executive

compensation in the IT industry to other industries, and found that in the IT industry more use is made of options in the compensation packages. According to them, the differences are partly due to the economic factors they identify, but also result from some unidentified factors. The first economic factor they consider is moral hazard, which is higher in situations where the manager’s actions have a higher impact on firm performance, but also in case the actions are harder to observe. The second is the adverse selection problem; by incorporating stock options in the executive compensation packages, firms try to lower the executive turnover, given that the value of the options increases with tenure of the executives (Salop and Salop 1976). Furthermore, they also included precision, risk, and innovation as economic factors. Additionally, the size of the organization is set to influence executive

compensation; larger and more complex firms tend to hire better managers (Rosen, 1982). Finally, income taxes, executive ownership, debt, and interest are also considered as economic factors which influence executive compensation. Moreover, it is stated that the less precise the performance measures, the less the performance sensitivity of compensation, given that the agent, in this case the executive, will have to bear more risk due to the variance in the performance measure (Banker and Datar 1989).

(10)

10 So far, it is clear that considerable research has been performed concerning the relation between pay for performance, executive compensation, and firm performance. However, as mentioned, most of the research has been performed using US and UK data, with some exceptions, which concerns data from Australia, South Africa, and Germany. Yet, data from the Netherlands seems to still be missing from the literature.

2.5 Dutch executive compensation

So far, only a limited amount of researchers focused on the relation between executive pay and firm performance in the Netherlands. For the period of 1998 till 2001, Duffhues and Kabir (2008) investigated the relation of executive compensation and corporate performance in the Netherlands. They found a significantly negative relation and suggest that firms offer compensation for reasons which are unrelated to firm performance, like attracting, retaining, and motivating executives. These results are consistent with the ‘managerial power view’, which regards executive compensation packages as a product of the relationships and the negotiations between the executive and the

corporate board (Bebchuk and Fried, 2003). A weak corporate board leads to inefficient compensation contracts, which actually exacerbates the agency problem (Ntim et al., 2015). However, this

managerial power view is partially contradictory to the optimal contracting theory; the managerial power view states that compensation is unrelated to performance, while the optimal contracting theory does indicate that there is a significant relation between the two. It can be argued that the managerial power view is wrong, in the sense that higher motivation can lead to higher performance, thus indirectly linking compensation to higher performance.

Swagerman and Terpstra (2007) also conducted a research investigating the executive compensation in the Netherlands. Their study resulted in the finding of a trend over the period of 2002 till 2004, which shows that the structure of incentive plans in the Dutch companies have been revised and attention was given to the discretionary powers of the supervisory board. Also, due to the Tabaksblat Code and pressure from outside investors, performance conditions were added to option and share plans. Another trend they found was that in the studied period, there was an increase in the use of equity-based pay and that new forms of long-term incentive arrangements had been set up. The last trend concerns the change from options to performance shares.

(11)

11 system, used by the managers; while the US firms follow a shareholder perspective, the Dutch firms apply the stakeholder perspective. Their results show that the choice of which corporate governance system is applied, results in a different effect of the incentive compensation. Given that in the US, the systems and perspectives are focussed on the shareholders, who usually desire a high return, incentive compensation, which leads to high performance and thus high returns, has a significant effect. In contrast, in the Netherlands systems and perspectives are focused on all stakeholders, not only shareholders, meaning that more factors than just the returns need to be considered in order to please all stakeholders. An example of these factors is the employees of the firm; due to the mortgage system applied in the Netherlands, fixed pay is more attractive than variable pay (Jansen et al., 2009). The results from Jansen et al. (2009) show that the corporate governance system adopted in the country is of high importance in order to receive effective results from the pay for performance compensation. However, the Dutch corporate governance system is rather different compared to other corporate governance systems (Ntim et al., 2015), which will be explained in the next section.

2.6 Dutch corporate governance

As mentioned by Conyon and Schwalbach (2000) and Jansen et al. (2009), differences in executive compensation can exist due to differences in corporate governance. There are many different corporate governance systems, like the Anglo-Saxon Model (US and UK), the German Model (Germany), the Latin model (Spain and Italy), and the structure model (the Netherlands) (Bolt and Peeters, 1998). Companies in the Netherlands usually have incorporated a two tier board; the Executive Management Board carries out overall management and the Supervisory Board supervises and advises (when necessary) the Executive Board (Corporate Governance Report: Corporate Governance in the Netherlands, 1997).

Over the last decade, the Dutch corporate governance environment has been subject to significant changes (Van der Laan, van Ees, and van Witteloostuijn, 2010). The main reason for this change is the implementation of the Dutch Corporate Governance Code in 2003, adopted since 1 January 2004, also known as the Tabaksblat Code or simply the Code. The Code has incorporated a couple of principles and best practice provisions in order to regulate the relations between the two-tier board and the shareholders, which, amongst others, concerns the remuneration policy set by the firms. The Code is applied to all companies listed in the Netherlands, provided that their balance sheet value exceeds €500 million (Dutch Corporate Governance Code, 2016). The Code states that, in case a management board member receives variable compensation, it is determined by predetermined measurable

performance criteria. It also states that the compensation policy should not include incentives which lead to self-optimizing behaviour of the executives, nor to taking risks which do not fit within the accepted strategy.

(12)

12 significantly. Stock options provide the owner the right to buy or sell an asset on, or prior to a fixed date, at a fixed price (Van der Laan, van Ees, and van Witteloostuijn, 2010). These options have become increasingly popular in the executive compensation packages, mostly in the US, but have also become increasingly popular in the Dutch executive compensation packages. The Code introduced a new condition when one wants to exercise his or her option, namely that in order to vest, or exercise, an option, certain performance targets or conditions must be met (Van der Laan, van Ees, and van Witteloostuijn, 2010). This condition can increase the motivation of executives to reach the

performance targets, given that it will result in a higher pay-out for them personally. In other words, due to the fact that the performance conditions must be met in order to receive the higher

compensation, executives will (try to) increase performance, resulting in higher firm performance. Considering the empirical evidence found in the US, the UK, and Germany of a positive relation between executive pay and firm performance (Mehran, 1995; Jensen and Murphy, 1990; Conyon and Schwalbach, 2000) and the condition set by the Dutch Corporate Governance Code about the

predetermined performance targets, the following hypothesis can be set:

Performance related pay for Dutch executives has a positive effect on firm performance.

The next section describes the data collection and the variables used in order to investigate the relation between executive compensation and firm performance in the Dutch market.

3. Methodology

In order to answer the research question with regard to the relation between executive compensation and firm performance in the Netherlands after the adoption of the Tabaksblat Code, the standard empirical literature on executive compensation is adopted (Duffhues and Kabir, 2008) and the following equation will be used:

Pay(i,t) = x + x1Performance(i,t) + x2Sales(i,t) + x3TotalEquity(i,t) + x4Age(i,t) + x5Tenure(i,t) + ԑ

3.1 Dependent variable

This study will focus on the flow compensation, which includes the bonus, as these are obliged to be performance related (Dutch Corporate Governance Code, 2016). Two dependent variables will be tested. The first dependent variable (Pay(i, t)) is the total compensation earned by the executive at firm ‘i’ at time ‘t’. This includes the base salary, the bonus, the pension, and other forms of cash

compensation. The variable is expressed as a natural logarithm, in order to adjust for the non-normality of compensation distribution (Duffhues and Kabir, 2008). The part of total compensation which is of most interest, is the bonus earned, given that these bonuses are dependent on the

(13)

13 also be tested, namely the bonus earned by the executive at firm ‘i’ at time ‘t’. The reason why the two variables are both tested, is to check whether there is only an effect on the bonus, or whether the effect is only measurable in total compensation, or in both. Additionally, by testing both variables the influence of other factors is controlled for and immediately serves as a check for the robustness of the variables.

3.2 Independent variable

In order to measure the effects of firm performance on CEO compensation, several measures of firm performance have been used in the literature. In previous research, firm performance is measured using stock price increases (Rosenstein and Wyatt, 1990), rate of return (Byrd and Hickman, 1992), Tobin’s Q, and return on assets (Mehran, 1995). For this analysis, the explanatory variable used as firm performance is measured in two ways, namely return on assets (ROA) and return on equity (ROE). First, ROA is defined as the net income earned over the book value of total assets (Mehran, 1995). Second, the ROE is defined as net income earned in the period, over total shareholders’ equity (Duffhues and Kabir, 2008). Several prior studies also make use of these estimates to measure firm performance (Byrd and Hickman, 1992; Mehran, 1995; Izan et al, 1998; Duffhues and Kabir, 2008). The reason to use accounting based measures is that these measures are more directly influenced by top management (Braam and Poutsma, 2015). Given that the data was rather limited, as mentioned in section 3.4, only the ROA and ROE are used as performance measures. Nonetheless, since both measures are used in different models, a check for robustness is also created. If the aforementioned hypothesis is correct, the estimated regressions coefficient x1 should be positive. This would indicate a positive relationship between firm performance and executive compensation.

3.3 Control variables

Executive compensation is not only affected by the performance of the firm, but also by many other factors. In order to control for these factors, they will be included in the regression. The first control variable concerns firm size. According to the literature, a significant relation between firm size and executive compensation is found (Rosen, 1982; Conyon, 1997); in case the company is larger and more complex, there is a tendency to hire better managers, which results in higher compensation. The size of the company is measured using annual sales and total equity.

The second control variable concerns the age of the executive (Mehran, 1995). It is said that executives with a higher age have a relatively higher compensation compared to their younger counterparts. The age of the executive is documented for each year in the sample period, using the annual reports of the companies.

(14)

14 Other factors which might affect the executive compensation, like taxes or interest rates (Banker and Datar, 1989), are controlled for by limiting the sample to only Dutch listed companies, which apply the Code.

3.4 Sample

In order to answer to the research question on the relationship between pay for performance and firm performance, data is collected using Boardex. Data is collected from the companies listed on the AEX in the Netherlands. The AEX, formerly known as the Amsterdam Stock Exchange, is composed of Dutch companies which are the most frequently traded securities on exchange. The reason why to use the companies listed at the AEX, is to make sure that the companies are obliged to report according to the Tabaksblat Code. In other words, it will make sure that CEO remuneration and the performance measurements applied to the bonuses, are reported in the annual report.

Furthermore, as a sample period, the period 2004 till 2016 is chosen. From the year 2004 onwards, companies at the AEX were obligated to apply the Code, therefore this year is chosen as a starting point. The reason why the year 2016 is chosen as final year, is that this is currently the last period for which the data is available. The other literature published so far mostly focused on the period before 2004 (Swagerman and Terpstra, 2007; Duffhues and Kabir, 2008), thus before the Code was applied. Therefore, it is interesting to see whether there has been an increase in firm performance, since the obliged publication of the executive remuneration and performance measures.

As mentioned, the data was extracted from Boardex for the years 2004 till 2016. The data was filtered on the country the Netherlands, after which only the information concerning the Executive Board chairman, the CEO, and the president was selected. Also, to make sure that the selected executives were not on the Supervisory Board, only the executives with the label of Executive Director were used in the sample. Finally, only the companies which were listed on the AEX were selected. Also,

following previous research, ING, as a company in the financial sector, was excluded from the sample, given that the financial sector has to comply with different legislation concerning executive

compensation (Ntim et al., 2015). This resulted in a sample of 127 CEO observations.

(15)

15 age and tenure of the executive was gathered using the annual statements and information gathered from the Internet.

3.5 Models

In order to test the equation presented at the beginning of this section, several separate models were used. Given that two different dependent variables are used, and two different performance measures, the equation is split into four models. The first two models, model 1.1 and model 1.2, use total compensation as the dependent variable. The last two models, model 2.1 and model 2.2, use the bonuses rewarded to the executives as the dependent variable. Also, models 1.1 and 2.1 use ROA as the independent variable, while models 1.2 and 2.2 use ROE as the independent variable. This results in the following four models:

● Model 1.1: Total(i,t) = x1ROA(i,t) + x2Age(i,t) + x3Tenure(i,t) + x4Sales(i,t) + x5Totalequity(i,t) ● Model 1.2: Total(i,t) = x1ROE(i,t) + x2Age(i,t) + x3Tenure(i,t) + x4Sales(i,t) + x5Totalequity(i,t) ● Model 2.1: Bonus(i,t) = x1ROA(i,t) + x2Age + x3Tenure(i,t) + x4Sales(i,t) + x5Totalequity(i,t) ● Model 2.2: Bonus(i,t) = x1ROE(i,t) + x2Age(i,t) + x3Tenure(i,t) + x4Sales(i,t) + x5Totalequity(i,t)

4. Results

4.1 Descriptive statistics

(16)

16

Table 1 Descriptive statistics.

The next table, table 2, shows the correlation matrix including significance levels concerning all the variables. As can be seen from the table, the correlation between return on assets and return on equity is high, which is not surprising, as they both measure the net income towards some balance sheet output.

Table 2 Correlation matrix including significance levels.

ROElag 100 .2024632 .1619252 -.1398901 .5899777 ROAlag 100 .0604914 .0469255 -.0557276 .1976452 Totalequity 117 5827.803 5652.697 228 23185 Sales 117 12205.62 11856.92 263 52713 Tenure 117 4.897436 3.489849 1 22 Age 117 55.19658 5.95564 43 67 ROE 117 .1919068 .1455055 -.1398901 .5899777 ROA 117 .0611898 .0450325 -.0557276 .1976452 Totallag 100 1862.83 1165.042 35 6824 Total 117 1870.393 1184.909 35 6824 Bonuslag 100 734.51 761.1918 0 4292 Bonus 117 729.0598 777.2324 0 4292 Variable Obs Mean Std. Dev. Min Max

(17)

17

4.2 Regression analysis

Given the four separate models, all using a different dependent variable for compensation and a different variable for performance measurement, the models can be said to present robust results. The following two tables present the results using total compensation (Total) as the dependent variable. The multivariate regression results from model 1.1, which are to be found in table 3, show that there is a positive relation between the return on assets (ROA) and total executive compensation. Also, the results are marginally significant, at a 10% level. Therefore, we can conclude that there is evidence regarding a positive relation between executive compensation and firm performance. Another conclusion which can be made from the results, concerns the relationship between firm size and executive compensation. Just like the literature suggests, the results show a significant positive

relation between both Sales and Totalequity to total compensation (Rosen, 1982; Conyon, 1997). Also, the coefficient related to Tenure shows a significant positive relation between the tenure of the

executive and the total compensation. The only control variable which is not significant, but does give a positive coefficient, is Age. However, considering the relatively low level of the R-squared, it can be suggested that variables are missing from the presented model.

Table 3 Regression results model 1.1.

Table 4 shows the results when using return on equity (ROE) as a measure of firm performance. In this model we, again, find a significant positive relation between firm performance and total executive compensation. This supports the aforementioned hypothesis of a positive relation between executive compensation and firm performance. Therefore, it can be suggested that an increase in firm

performance, measured by the return on equity or return on assets, leads to an increase in the total compensation distributed to the executive. This also supports the solution to the agency problem, as the ROE shows the net income returned to the firm’s shareholders. Thus it can be suggested that an increase in shareholder’s value leads to an increase in executive compensation, therefore, aligning the

(18)

18 interests of the executive to those of the shareholders. Furthermore, the results in table 4 again show a significant positive relationship between the control variables and total compensation, except for the executives’ age. Also, the R-squared is slightly higher compared to model 1.1, suggesting that measuring the firm performance using ROE better fits the model. However, the R-squared is still relatively low, indicating that there are still variables missing to the model.

The next two tables present the regression results of models 2.1 and 2.2, which measure the effect of firm performance on the bonus earned by the executives.

_cons -11.0224 677.6833 -0.02 0.987 -1353.897 1331.852 Totalequity .073116 .021725 3.37 0.001 .0300664 .1161656 Sales .0468526 .0100839 4.65 0.000 .0268706 .0668346 Tenure 40.75093 21.76987 1.87 0.064 -2.387517 83.88937 Age 6.332413 12.65354 0.50 0.618 -18.74143 31.40625 ROE 1742.211 512.2379 3.40 0.001 727.1774 2757.244 Total Coef. Std. Err. t P>|t| [95% Conf. Interval] Total 162865214 116 1404010.46 Root MSE = 764.15 Adj R-squared = 0.5841 Residual 64816140.9 111 583929.197 R-squared = 0.6020 Model 98049073 5 19609814.6 Prob > F = 0.0000 F(5, 111) = 33.58 Source SS df MS Number of obs = 117

(19)

19

The results of models 2.1 and 2.2 are very similar to the results of models 1.1 and 1.2. The only differences are the lower coefficients of the firm performance in the models using the bonuses as the dependent variable and the lower R-squared. This suggests that the effect of firm performance is higher on total compensation than it is on the bonuses earned by the executives. Also, due to the higher R-squared, the models 1.1 and 1.2 seems to be a better fit for the investigated issue.

Nonetheless, all models present the same conclusion to the investigated problem, namely that higher firm performance leads to higher executive compensation, therefore confirming the studied hypothesis.

4.3 Last year’s performance

Apart from the models presented above, one can argue that it is not this year’s firm performance which influences total compensation, but that last year’s firm performance is more relevant for this year’s executive compensation (Duffhues and Kabir, 2008). After the year has ended, the performance of that year is measured. In order to reward the executive for the good, or punish for the bad performance of last year, the compensation of this year is adjusted (Jensen and Murphy, 1990). As this is known to the executive, he or she will be motivated to outperform the year before, in order to increase his or her compensation for the next year. Therefore, four more regressions are run, using a time lag of firm performance to measure the firm performance effect of last year on this year’s total executive compensation. The models can be represented as follows:

(20)

20 The tables with the regression results can be found in appendix B. Table 5 represents the results concerning model 3.1, table 6 concerning model 3.2, table 7 concerning model 4.1, and table 8 concerning model 4.2.

The tables do not show any surprising results compared to the results presented above of the models without the time lag. All the coefficients concerning firm performance remain positive, although they are no longer significant. The difference in the coefficients that can be found lies the fact that the coefficients with the time lag are lower compared to the coefficients without the time lag, indicating that the effect of this year’s performance is higher than last year’s performance. Also, the R-squared in the models in lower compared to the models without the time lag. This indicated that the models without the time lag show a better representation of the relationship.

4.4 The financial crisis

Next to the time lag, there is another important factor which needs to be considered when regressing the data of this time period, namely the financial crisis of 2008. Given that the crisis can have had a significant effect on the performance of the firm, it is important to control for this factor. Therefore, regressions were run using two different time periods, namely for the years 2004 till 2007, the period before the crisis, and for the years 2008-2016, the period during and after the crisis. The results can be found in appendix C, tables 11 till 18. Ascan be seen from both periods, the performance coefficients remain positive, but the significance levels change. The biggest difference compared with the models presented above concerns the level of the R-squared; the period 2004-2007 shows very high R-squared levels compared to the overall period of 2004-2016, and 2008-2016 relatively low R-squared levels. This indicated that for the period from 2008 till 2016 there is a significant variable, of maybe more than one variable, missing from the model, which could be the result of the financial crisis. These results suggest a subject for further investigation.

5. Conclusion

This study contributes to the growing understanding of the role of executive remuneration and firm performance in a strong, Type II, corporate governance setting. Performance related pay for company executives is said to serve as a solution to the principal-agency problem (Bebchuk and Fried, 2003), as it links the interests of the agent, in this case the executives, to the interests of the principals, the shareholders (Jensen and Murphy, 1990).

(21)

21 obliged to report the executive remuneration (Dutch Corporate Governance Code, 2016). This also includes the publication of the performance targets and measures which are applied to determine the executive’s remuneration. The adoption of the Tabaksblat Code has led to a stronger corporate governance system in the Netherlands (Swagerman and Terpstra, 2007), which, according to the literature, leads to higher firm performance (Drunev and Kim, 2005; Drobetz et al., 2004; Klapper and Love, 2004).

Furthermore, previous research concerning the relationship between executive compensation and firm performance in other countries with stronger corporate governance settings, show a positive relation between the compensation and firm performance (Jensen and Murphy, 1990; Mehran, 1995; Ntim et al., 2015).

The results of this study support the findings of previously performed research. The results show that an increase in firm performance, measured using the accounting measures return on assets (ROA) and return on equity (ROE), leads to an increase of the total compensation and the bonus distributed to the executive. Also, given that the data was gathered in a period of strong corporate governance as a result of the adoption of the Tabaksblat Code, it supports the theory that stronger corporate governance leads to more efficient compensation contracts. Additionally, the results show that last year’s firm

performance is also related to this year’s executive compensation, although this year’s firm

performance does seem to be a better fit to explain this year’s compensation. Moreover, the effect of this year’s performance on executive compensation is higher compared to last year’s performance. Finally, the effects of the financial crisis of 2008 are also taken into consideration. By splitting the total time period into two separate periods, one before the crisis (2004-2007) and one during and after the crisis (2008-2016), the differences in the effect of firm performance on executive compensation is investigated. The results do not show any surprising differences; all coefficients remain positive, only some are no longer significant. The only major difference concerns the fit of the models, which is measured using the R-squared. It can be seen that for the period of 2004 till 2007, the variables more extensively explain the level of executive compensation compared to the period of 2008 till 2016. For the latter period, there seems to be an indicator missing from the model, which can be the result of the crisis.

All in all, the results show a solution to the agency problem; by linking the firm’s performance to the executive’s compensation, the firm can align the interests of the executives to those of the

shareholders. The executives will have a personal gain in higher firm performance, as their personal return, their compensation, will be higher. Shareholders will also benefit from this situation, as their returns will also be higher with higher firm performance.

(22)

22 stocks can lead to more reliable and extensive results. Unfortunately, data limitations prevented this study from incorporating these variables.

(23)

23

6. References

Anderson, M.C., Banker, R.D., and Ravindran, S. (2000) ‘Executive compensation in the information technology industry’, Management Science, 46(4), 530-547

Antle, R., & Smith, A. (1986). An empirical investigation of the relative performance evaluation of corporate executives. Journal of Accounting Research, 1-39.

Banker, R. D., & Datar, S. M. (1989). Sensitivity, precision, and linear aggregation of signals for performance evaluation. Journal of Accounting Research, 21-39.

Bayo-Moriones, A., Galdon-Sanchez, J. E., & Martinez-de-Morentin, S. (2017). Performance Measurement and Incentive Intensity. Journal of Labor Research, 38(4), 496-546.

Bebchuk, L. A., & Fried, J. M. (2003). Executive compensation as an agency problem. Journal of economic perspectives, 17(3), 71-92.

Bolt, W., & Peeters, M. (1998). Corporate governance in the Netherlands. In Corporate Governance, Financial Markets and Global Convergence (pp. 89-112). Springer, Boston, MA.

Boyd, B. K. (1994). Board control and CEO compensation. Strategic management journal, 15(5), 335-344.Byrd, J. W., & Hickman, K. A. (1992). Do outside directors monitor managers?: Evidence from tender offer bids. Journal of financial economics, 32(2), 195-221.

Braam, G., & Poutsma, E. (2015). Broad-based financial participation plans and their impact on financial performance: Evidence from a Dutch Longitudinal Panel. De Economist, 163(2), 177-202. Byrd, J. W., & Hickman, K. A. (1992). Do outside directors monitor managers?: Evidence from tender offer bids. Journal of financial economics, 32(2), 195-221.

Carpenter, M. A., Sanders, W. G., & Gregersen, H. B. (2001). Bundling human capital with organizational context: The impact of international assignment experience on multinational firm performance and CEO pay. Academy of management journal, 44(3), 493-511.

Conyon, M. J. (1997). Corporate governance and executive compensation. International journal of industrial organization, 15(4), 493-509.

Conyon, M. J., & Schwalbach, J. (2000). Executive compensation: Evidence from the UK and Germany. Long Range Planning, 33(4), 504-526.

Core, J. E., Holthausen, R. W., & Larcker, D. F. (1999). Corporate governance, chief executive officer compensation, and firm performance. Journal of financial economics, 51(3), 371-406.

Corporate Governance Committee. (2003). The Dutch Corporate Governance Code: Principles of good corporate governance and best practice provisions. Corporate Governance Committee. Drobetz, W., Schillhofer, A., & Zimmermann, H. (2004). Corporate governance and expected stock returns: Evidence from Germany. European financial management, 10(2), 267-293.

Duffhues, P., & Kabir, R. (2008). Is the pay–performance relationship always positive?: Evidence from the Netherlands. Journal of multinational financial management, 18(1), 45-60.

Durnev, A., & Kim, E. H. A. N. (2005). To steal or not to steal: Firm attributes, legal environment, and valuation. The Journal of Finance, 60(3), 1461-1493.

(24)

24 Gompers, P., Ishii, J., & Metrick, A. (2003). Corporate governance and equity prices. The quarterly journal of economics, 118(1), 107-156.

Guay, W., Core, J., & Larcker, D. (2002). Executive equity compensation and incentives: A survey. Hill, J. G., Masulis, R. W., & Thomas, R. S. (2011). Comparing CEO employment contract

provisions: Differences between Australia and the United States. Vand. L. Rev., 64, 557.

Izan, H. Y., Sidhu, B., & Taylor, S. (1998). Does CEO pay reflect performance? Some Australian evidence. Corporate Governance: An International Review, 6(1), 39-47.

Jansen, E. P., K. A. Merchant and W. A. Van der Stede (2009), National differences in incentive compensation practices: The differing roles of financial performance measurement in the United States and the Netherlands. Accounting, Organizations and Society, 34(1), 58-84

Jensen, M.C. and Murphy, K.J. (1990) ‘Performance pay and top-management incentives’, Journal of Political Economy, 98(2), 225-264.

Kaydos, W. (1998). Operational performance measurement: increasing total productivity. CRC press. Klapper, L. F., & Love, I. (2004). Corporate governance, investor protection, and performance in emerging markets. Journal of corporate Finance, 10(5), 703-728.

Mathur, A., Dangayach, G. S., Mittal, M. L., & Sharma, M. K. (2011). Performance measurement in automated manufacturing. Measuring business excellence, 15(1), 77-91.

McConaughy, D. L. (2000). Family CEOs vs. nonfamily CEOs in the family-controlled firm: An examination of the level and sensitivity of pay to performance. Family Business Review, 13(2), 121-131.

Mehran, H. (1995). Executive compensation structure, ownership, and firm performance. Journal of financial economics, 38(2), 163-184.

Ntim, C. G., Lindop, S., Osei, K. A., & Thomas, D. A. (2015). Executive compensation, corporate governance and corporate performance: a simultaneous equation approach. Managerial and Decision Economics, 36(2), 67-96.

Rosen, S. (1982). Authority, control, and the distribution of earnings. The Bell Journal of Economics, 311-323.

Rosenstein, S., & Wyatt, J. G. (1990). Outside directors, board independence, and shareholder wealth. Journal of financial economics, 26(2), 175-191.

Salop, J., & Salop, S. (1976). Self-selection and turnover in the labor market. The Quarterly Journal of Economics, 619-627.

Swagerman, D., & Terpstra, E. (2007). The effectiveness of Dutch executive pay packages. Compensation & Benefits Review, 39(2), 47-57.

(25)

25

7. Appendices

Appendix A – List of AEX companies

Aegon NV Ahold (Koninklijke) NV Akzo Nobel NV ASML Holding NV Corio NV Fugro NV

Gemalto NV (Axalto Holdings prior to 06/2006) Heineken NV

Imtech NV

PostNL NV (TNT NV prior to 05/2011) (TPG prior to 05/2005) Randstad HLDGS NV

Royal DSM NV (DSM prior to 10/2005)

Royal Philips NV (Royal Philips Electronics NV prior to 05/2013) Royal KPN NV

SBM Offshore NV (IHC Caland prior to 05/2005) Unilever NV

(26)

26

Appendix B

_cons -170.6821 824.3088 -0.21 0.836 -1807.367 1466.002 Totalequity .0621821 .0252324 2.46 0.016 .0120826 .1122816 Sales .0506349 .0130112 3.89 0.000 .0248009 .0764689 Tenure 52.91465 27.88862 1.90 0.061 -2.458857 108.2882 Age 12.73817 14.95193 0.85 0.396 -16.94925 42.42559 ROAlag 1943.06 1878.783 1.03 0.304 -1787.307 5673.427 Total Coef. Std. Err. t P>|t| [95% Conf. Interval] Total 139733485 99 1411449.34 Root MSE = 831.79 Adj R-squared = 0.5098 Residual 65035867.6 94 691870.932 R-squared = 0.5346 Model 74697617.1 5 14939523.4 Prob > F = 0.0000 F(5, 94) = 21.59 Source SS df MS Number of obs = 100

_cons -279.5919 784.656 -0.36 0.722 -1837.545 1278.361 Totalequity .0719807 .0250095 2.88 0.005 .0223238 .1216377 Sales .0449676 .0127191 3.54 0.001 .0197136 .0702216 Tenure 44.05536 27.45118 1.60 0.112 -10.44959 98.56032 Age 13.22012 14.48011 0.91 0.364 -15.53048 41.97072 ROElag 1256.413 521.9426 2.41 0.018 220.0836 2292.742 Total Coef. Std. Err. t P>|t| [95% Conf. Interval] Total 139733485 99 1411449.34 Root MSE = 811.86 Adj R-squared = 0.5330 Residual 61956629.4 94 659113.079 R-squared = 0.5566 Model 77776855.3 5 15555371.1 Prob > F = 0.0000 F(5, 94) = 23.60 Source SS df MS Number of obs = 100

(27)

27

(28)

28

Appendix C

Table 11 Regression results model 1.1, period 2004-2007.

_cons -784.9103 1135.909 -0.69 0.493 -3068.809 1498.988 Totalequity .0161887 .0312662 0.52 0.607 -.0466762 .0790536 Sales .072057 .0148339 4.86 0.000 .0422314 .1018825 Tenure 9.511311 28.25015 0.34 0.738 -47.28941 66.31204 Age 28.5373 18.79033 1.52 0.135 -9.243185 66.31778 ROA 970.9175 3286.838 0.30 0.769 -5637.714 7579.549 Total Coef. Std. Err. t P>|t| [95% Conf. Interval] Total 88579094.1 53 1671303.66 Root MSE = 739.39 Adj R-squared = 0.6729 Residual 26241641 48 546700.855 R-squared = 0.7037 Model 62337453 5 12467490.6 Prob > F = 0.0000 F(5, 48) = 22.80 Source SS df MS Number of obs = 54

_cons -1796.29 951.1528 -1.89 0.065 -3708.711 116.131 Totalequity .0232076 .0284865 0.81 0.419 -.0340683 .0804836 Sales .0729441 .0125738 5.80 0.000 .0476628 .0982254 Tenure .4359589 26.32889 0.02 0.987 -52.50183 53.37375 Age 38.48245 16.69379 2.31 0.026 4.917328 72.04757 ROE 2318.785 818.7695 2.83 0.007 672.539 3965.032 Total Coef. Std. Err. t P>|t| [95% Conf. Interval] Total 88579094.1 53 1671303.66 Root MSE = 685.04 Adj R-squared = 0.7192 Residual 22525505.8 48 469281.371 R-squared = 0.7457 Model 66053588.3 5 13210717.7 Prob > F = 0.0000 F(5, 48) = 28.15 Source SS df MS Number of obs = 54

(29)

29

_cons -877.5999 611.5654 -1.44 0.158 -2107.235 352.0346 Totalequity .0289749 .018316 1.58 0.120 -.007852 .0658018 Sales .0450558 .0080846 5.57 0.000 .0288006 .061311 Tenure -8.084983 16.92876 -0.48 0.635 -42.12254 25.95257 Age 14.97945 10.73366 1.40 0.169 -6.60201 36.56091 ROE 711.031 526.4465 1.35 0.183 -347.4606 1769.523 Bonus Coef. Std. Err. t P>|t| [95% Conf. Interval] Total 40327216.1 53 760890.87 Root MSE = 440.46 Adj R-squared = 0.7450 Residual 9312358.86 48 194007.476 R-squared = 0.7691 Model 31014857.2 5 6202971.45 Prob > F = 0.0000 F(5, 48) = 31.97 Source SS df MS Number of obs = 54

_cons 1292.046 1082.863 1.19 0.238 -876.3498 3460.442 Totalequity .0871371 .0312147 2.79 0.007 .0246307 .1496435 Sales .0331983 .0151377 2.19 0.032 .0028856 .0635109 Tenure 104.8004 36.16688 2.90 0.005 32.37748 177.2233 Age -21.53591 20.22369 -1.06 0.291 -62.03315 18.96133 ROA 4365.567 2252.135 1.94 0.058 -144.2563 8875.391 Total Coef. Std. Err. t P>|t| [95% Conf. Interval] Total 74220792.4 62 1197109.56 Root MSE = 790.34 Adj R-squared = 0.4782 Residual 35604618.7 57 624642.434 R-squared = 0.5203 Model 38616173.7 5 7723234.73 Prob > F = 0.0000 F(5, 57) = 12.36 Source SS df MS Number of obs = 63

(30)

30

_cons 191.4886 773.8219 0.25 0.805 -1358.063 1741.04 Totalequity .0432417 .0223063 1.94 0.058 -.0014258 .0879092 Sales .0164518 .0108175 1.52 0.134 -.0052098 .0381134 Tenure 71.60059 25.84512 2.77 0.008 19.84663 123.3545 Age -9.928092 14.452 -0.69 0.495 -38.86772 19.01153 ROA 3396.915 1609.393 2.11 0.039 174.1615 6619.668 Bonus Coef. Std. Err. t P>|t| [95% Conf. Interval] Total 29644105.4 62 478130.733 Root MSE = 564.79 Adj R-squared = 0.3329 Residual 18181987.9 57 318982.244 R-squared = 0.3867 Model 11462117.5 5 2292423.5 Prob > F = 0.0000 F(5, 57) = 7.19 Source SS df MS Number of obs = 63

Executive Compensation and Firm Performance: Evidence from the Netherlands