University of Amsterdam
Faculty of Economics and Business
MSc Business Economics - Finance
The relative difference in compensation between corporate ranks within
public corporations and its effect on company performance
Master thesis – Final version
Student: Alina Gatea
Student number: 6122280
Supervisor: Dr. Z. Sautner
Table of contents
I. Company performance, the relative difference in compensation between corporate ranks
and long-term career incentives...6
A. Executive board level...6
A.1. The relative difference in remuneration between the CEO and other executives …….………7
A.2. The age difference between the CEO and other executives... 8
A.3. Executive board size... 8
A.4. Outcome of internal promotion to CEO position from within the executive board...9
A.5. Executive board monitoring mechanisms...10
B. Non-executive level...11
B.1. The relative difference in remuneration between executive and non-executive employees..…11
B.2. Outcome of internal promotion to CEO position from the non-executive rank …..………...12
B.3. Inside executives: outcome of internal promotion to executive manager level..………..12
II. Data collection and definitions………..12
A. Sample selection………...………..12
B. Variable definitions and computations………..……….13
B.1. Measures of company performance………..………..13
B.2. Measures of explanatory variables………13
B.3. Measures of governance control variables………17
B.4. Measures of financial control variables……….………18
B.5. Descriptive statistics………...18
III. Methodology...20
A. Measuring the effect of compensation levels on company performance………..…………..20
B. Measuring the effect of the relative compensation difference and long-term career horizons on company performance…...21
IV. Empirical results and analysis...22
A. Analysis of the effect of compensation levels on company performance………..….23
B. Analysis of the effect of the relative difference in compensation between corporate ranks on company performance………….………...25
V. Drawbacks and suggestions for further research...27
1
The relative difference in compensation between corporate ranks within
public corporations and its effect on company performance
ABSTRACT
This paper hypothesizes that the relative difference in compensation between corporate ranks in public corporations has a positive effect on company performance due to the long-term career incentives of employees. Tests are conducted on the effect of the ratio of CEO compensation to average compensation of other executive board members and between average executive and non-executive compensation on company performance, using a sample of 29 companies listed in the Netherlands. The results show a positive relation between company performance and the relative short-term cash compensation difference and absolute age difference between the CEO and other members of the executive board.
Studies on the motivation of employees to maximize company value in public corporations has mostly centered on the roles of CEOs, on how their personal characteristics and expertise influence their performance, and on optimal remuneration packages used to motivate them to undertake company value-enhancing projects and to limit agency issues (e.g., Brookman and Thistle, 2009; Jensen and Murphy, 1990; Hall and Murphy, 2003; Bebchuk and Fried, 2003).
Research on the roles of managers other than the CEO has only recently begun to draw more academic attention. In their 2011 paper, Acharya, Myers and Rajan develop a theoretical model of the roles of subordinate managers in monitoring the CEO and in increasing long-term company performance. Graham, Harvey and Puri (2010) find that key officers other than the CEO hold substantial decision-making authority with respect to the direction of corporate strategy. Employees other than the CEO clearly play an important role in driving company value and thus further research focusing on the motivation of these employees is required.
This paper explores the distribution of compensation across corporate levels and investigates the effect of the relative difference in compensation between corporate ranks within public corporations on company performance. Specifically, this paper hypothesizes that as employee motivation to be promoted to a higher level within the organization increases, employees become more likely to invest effort to produce value-adding work in order to receive the promotion. By doing
2
so, they increase company value. Employee motivation to seek promotion depends on the additional payoff of being employed in the higher position. This payoff is proxied in this thesis by the remuneration gap between corporate ranks. Furthermore, since the probability of promotion is necessarily linked to the extent of an employee’s career horizons, this paper tests for the effect on company performance of the difference between CEO age and the average age of other executives within the board, as younger employees are expected to be more invested career-wise in the company and therefore more motivated to produce superior results in terms of performance. Finally, this paper explores the effect on company performance of internal promotion to key positions within the corporation, specifically to executive board member and CEO positions.
This thesis therefore explores the effect on company performance of the relative compensation gap between employees and higher management. That is, the compensation an employee receives relative to that of a higher-ranking officer within the company. This paper hypothesizes that the greater this remuneration gap is, the higher the effort employees supply will be in order to increase their productivity and be considered for promotion to the higher-ranking position. As company performance is driven by employee productivity and expertise, this paper proposes that the remuneration gap between corporate ranks is positively related to company performance.
Moreover, the effort supplied by employees, and therefore company performance, will also be positively related to the perceived probability of their promotion and to the extent of their career horizons. The model used in this paper is therefore based on the possibility of employee promotion and long-term career opportunities within the company they are currently working for. Career concerns of employees are also key in the model of company value creation developed by Acharya, Myers and Rajan (2011). Likewise, Prendergast (1993) focuses on the role of promotion as a mechanism to induce employees to develop company-specific skills. This paper also relies on the existence of internal promotion to hypothetically stimulate employee productivity and value creation, and indeed, empirical data gathered as part of this thesis supports this assumption: a preliminary analysis shows that it is more likely for higher corporate ranks to be filled by promotion from within the company, as opposed to external hires (Table I, Appendix B). In this sense, employees are, career-wise, invested in the company they work for.
3
The analysis of this paper focuses on two levels within the organization: the level of the executive board, otherwise known as the board of management, and subsequently on the level of employees below the rank of the executive board, which will be referred to as the non-executive level. The first level focuses on the potential of value creation by key officers which are members of the executive board, motivated by the possibility of future promotion to CEO position. A preliminary analysis of promotion patterns within the sample data used in this paper shows that it is frequently the case that when a CEO steps down, a member of the executive board is promoted to fill the new position. Specifically, Table I shows that 69.05% of CEOs included in this thesis were appointed from among the existing members of the management board. This paper hypothesizes that as the relative difference in remuneration between the CEO and that of other members of the board increases, so too does the motivation of these other executives to be promoted to CEO position, as the payoff of this outcome would be greater. This thesis therefore tests the effect of this relative remuneration gap on company performance. Additionally, the model tests for the effect of the difference in age between the CEO and other members of the board, as the manager chosen to replace the retiring CEO is likely to be younger. A greater age difference implies longer career horizons and more possibilities for future promotion to CEO position for the younger manager. It therefore serves as a proxy for higher motivation of younger executive board members to undertake value-adding projects. Whether the current CEO has been promoted from among the members of the executive board also makes another internal promotion likelier. Company performance is therefore expected to be greater in companies where the current CEO was appointed from inside the board. Additionally, the number of executives on the management board could increase the level of competition among board members striving to be promoted to CEO position, thereby increasing their productivity, and consequently company value. However, it is also possible for the size of the executive board to have an effect analogous to that of the negative relation found between the size of the supervisory board and that of company performance, due to less effective communication and decision-making present in larger boards (Yermack, 1996). Furthermore, to allow for the effect of other parties which are in a position to monitor the executive board as a whole, the model includes the effect of company supervisory board
4
size, number of significant shareholders and proportion of independent executives on the management board.
The second level on which the effect of relative remuneration on firm performance is tested focuses on employees outside the executive board, that is, non-executive employees, and their remuneration relative to that of executive board members. To this end, the second model in the analysis includes the relative difference between the average compensation paid to an executive board member and the average of that of a non-executive employee. As before, the higher relative difference in compensation is hypothesized to increase non-executive employee motivation to supply value-adding effort. The effect of this remuneration gap on company performance is an especially important research direction due to the on-going debate on the fairness and ethical implications of the size of remuneration packages of executives compared to those of non-executive employees. Furthermore, whether the current CEO has previously worked within the company at the non-executive level can reflect a company-specific culture for internal promotion, and is likely to imply a higher probability that the next CEO will eventually be promoted from among current company non-executive employees. This is another driver of employee motivation hypothesized to be positively related to performance. Finally, this paper also tests for the effect of the fraction of inside executives, that is, the proportion of members on the management board (excluding the CEO) which were promoted from inside the company. As this is also likely to increase non-executive employee motivation to supply effort in order achieve promotion, it is hypothesized to be positively related to company performance.
Data used in this study derives from a sample of 29 companies listed in the Netherlands during the period 2008 to 2012. Annual data over these 5 years is extracted, resulting in a sample of 145 observations. Due to the common two-tier board structure of Dutch companies, where the executive board and the supervisory board are two separate entities, this sample allows for research on the roles of key executive managers other than the CEO, and their impact on company performance.
The results indicate that the relative difference between CEO compensation and the average of that of other executives on the board is a significant driver of company profitability. Moreover, the component of the relative difference in remuneration classified as “Other compensation” seems to be driving this effect. This implies that executives on the board are particularly motivated to achieve
5
promotion to CEO rank due to the fringe benefits associated with this position. Furthermore, companies awarding lower company stock options value to executives relative to the CEO are less profitable. This supports the idea that executives other than the CEO have long-term interests in the company, and therefore awarding them a comparatively reduced value in terms of long-term components of remuneration is detrimental to their efforts to improve company performance. What is more, there is evidence to indicate that companies with relatively younger executives in the CEO’s entourage exhibit higher productivity levels, as well as improved performance on the stock market. This suggests that key officers surrounding the CEO who are younger, and therefore facing longer-term career horizons, may be comparatively more motivated to drive company performance. However, performance on the stock market is poorer for companies whose CEOs are appointed from within the executive board, implying that investors are less confident in insiders as choices for CEO position. The size of the executive board is not found to have a statistically significant effect on company performance.
The analysis carried out at the non-executive level does not find statistically significant results between the compensation of non-executive relative to executive employees and company performance. A comparison with results found at the executive level would suggest that any promotion incentives created by the relative difference in compensation between corporate ranks are stronger at higher levels of management.
The analysis of this paper offers a fresh perspective on the drivers of employee motivation within public corporations. The focus on executives other than the CEO, as well as non-executive employees sheds light on their roles in company value creation and performance. In particular, the findings of this paper are valuable to boards of directors as well as human resources departments charged with the design of employee compensation and promotion policies. What is more, the conclusions of this paper with respect to relative executive age and its relation to company performance should encourage corporate initiatives of young talent acquisition, professional development and promotion.
The rest of this paper proceeds as follows. Section I provides an overview of existent literature on the research topic and outlines the hypotheses tested. Section II provides a presentation of the data
6
collection and variable definitions used, while section III describes the methodology used in this paper. Section IV discusses the results of this thesis while Section V provides some limitations, issues and suggestions for future research. Finally, Section VI concludes the findings of this paper.
I. Company performance, the relative difference in compensation between corporate ranks and long-term career incentives
This paper researches the effect on company performance of the relative difference in employee compensation between corporate ranks. This relative difference in remuneration is hypothesized to have a positive impact on company performance through the incentives it provides for employees to supply effort in order to achieve a promotion to a higher rank within the organization. The likelihood of employee promotion to a higher level within the company is therefore also hypothesized to have a positive effect on company performance. For this reason, this paper also explores the effect on company performance of the age difference between corporate ranks at the executive level, and the impact of assigning company employees through promotion to key executive positions, as opposed to hiring external candidates.
The next subsections explore the two corporate levels on which relative differences in employee remuneration and their relation to company performance are researched. The first level, investigated in section A, focuses on the executive board, and in particular on executive managers subordinate to the CEO. Section B subsequently explores the hypotheses tested at the non-executive level.
A. Executive board level
The important roles of key managers other than the CEO are increasingly being researched by academic literature. In addition to the theoretical model of Acharya, Myers and Rajan (2011), Duchin and Sosyura (2013) emphasize the roles of divisional managers and how their connections to the CEO affect capital allocation decisions within the company. Graham, Harvey and Puri (2010) further explore the circumstances affecting the degree to which CEOs delegate strategic tasks and decisions to subordinate managers.
7
Within a two-tier board structure, the executive board and supervisory board are two separate entities. As such, executive boards are solely in charge of the management of the company and, in addition to the CEO, are frequently composed of other key officers such as the CFO (Chief Financial Officer) or the COO (Chief Operating Officer).
The first part of the analysis of this paper takes a closer look at these key managers. As they form the executive suite, they are ideally positioned to monitor the CEO and to ensure that he or she does not take any actions which will be detrimental to company value. Moreover, Acharya, Myers and Rajan (2011) argue that the possibility of achieving promotion acts as an incentive for employees to monitor upper management so that at the point of their eventual promotion to said management position, they “inherit” a high-value company. Following this reasoning, the model developed in this paper predicts that employees care about the future of their company because they are invested in it due to internal career advancement possibilities. Consequently, they have important incentives to maximize company value beyond the short term. This is especially true in the case of officers surrounding the CEO as they are in a position to monitor the CEO, have power over the direction of company strategy and projects, and they face a high likelihood of achieving promotion to CEO position in the future. Indeed, as mentioned previously, a preliminary analysis of the data collected for this paper shows that almost 70% of the CEOs surveyed were promoted to this rank from among the members of the executive board at the time (Table I, Panel A). Moreover, as shown in Panel B of Table I, executives who reach the rank of CEO spend on average 2.73 years on the executive board prior to being promoted to CEO position.
A.1. The relative difference in remuneration between the CEO and other executives
The paper of Bognanno (2001) researches compensation and promotion within U.S. corporations and finds that remuneration rises sharply with hierarchical rank. At the level of the executive board, where it is likely that the next CEO will be promoted from among the current members of the management board, the relative difference in compensation between the CEO and other members of the board is likely to act as an incentive for members to improve their performance in order to be promoted to CEO position when the current CEO retires or resigns. Through their
8
enhanced performance, they act to improve company value. As the relative compensation gap between the CEO and other members increases, so too does the payoff of the potential promotion driving executive effort. This should ultimately lead to superior company performance. From this our first hypothesis follows:
Hypothesis 1: Company performance is positively related to the relative difference between CEO
remuneration and the average compensation awarded to other executives in the CEO’s entourage.
A.2. The age difference between the CEO and other executives
Younger executives, with longer-term horizons given by distance to retirement, have more time available to advance their careers, achieve promotion, and receive the higher compensation associated with a higher corporate rank. They should therefore be particularly invested in the company they are currently working for, especially if strong internal promotion patterns are present. In other words, as the age of a subordinate manager compared to that of the CEO is lower, there are more opportunities for internal promotion for the younger manager, and this is hypothesized to increases their value-creating incentives. Evidence from large corporations shows that subordinate managers are indeed, on average, younger than the current CEO and are therefore likely to eventually replace him (Acharya, Myers and Rajan, 2011, p. 694). In this sense they have, career-wise, longer-term perspectives than the CEO. They therefore have higher incentives to monitor the current CEO to ensure that he or she does not undertake investments which reduce long-term company value as they are likely to “inherit” the company in the future. It therefore follows that the age gap between the CEO and other executive managers should have a positive influence on company value. Our second hypothesis is therefore:
Hypothesis 2: Company performance is positively related to the age difference between the CEO and
other executives in the CEO’s entourage.
A.3. Executive board size
The model outlined so far suggests that the size of the executive board is positively related to company performance, as a larger board size would imply more executives in a position to monitor the
9
CEO in order to ensure value preservation for the time of their potential promotion to CEO position. Moreover, the larger number of executives contending for CEO position is likely to increase competition between executives and to therefore improve performance efforts. From this follows a third hypothesis follows:
Hypothesis 3: Company performance is positively related to the size of the executive board.
As an alternative hypothesis however, it is possible that as boards grow larger, communication and cooperation become less efficient to the point that the increased number of executives leads to lower performance, as found in the case of supervisory boards (Yermack, 1996).
A.4. Outcome of internal promotion to CEO position from within the executive board
As observed previously, it is often the case that when a CEO retires or resigns, the next CEO is chosen from among the current members of the executive board. If the current CEO has been promoted from inside the board, this is likely to signal a company policy of internal promotion to CEO position and implies a higher probability that the next CEO will be likewise chosen from among the current members of the executive board. This would increase executive motivation to ensure company value-maximization:
Hypothesis 4: Companies where the current CEO has been promoted from within the executive board
show an enhanced performance.
Ang and Nagel (2010) present a similar hypothesis, in that firms hiring CEOs internally experience a significant improvement in operating performance. They find that it is financially more profitable to assign an insider to the role of CEO rather than hiring externally. However, their model relies on insiders having a better knowledge of company operations, thereby requiring less time to learn on the job, which increases their performance, and on supervisory boards having more information on internal nominees, leading to a better selection of candidates from inside the company.
10
A.5. Executive board monitoring mechanisms
To allow for the effect of mechanisms with the authority to monitor the actions of executive boards, the analysis includes the effect on company performance of the number of large shareholders, the proportion of executives independent from the CEO, and the size of the supervisory board.
Shareholders owning a significant share of company stock have an incentive to monitor executives and to ensure that no actions are taken which would be detrimental to company performance. Research shows that indeed, activist shareholders, whether institutional or individual investors, play an important role in monitoring company activities and in working towards implementing change in company policy and activities when boards fail to properly monitor company management (Gillan and Starks, 2000). Importantly for the model used in this paper, previous research has uncovered evidence of a relationship between the presence of significant institutional shareholders and both CEO pay and company performance as measured by stock performance, return on assets and operating profit margin (Brav et al., 2008).
The proportion of independent executives is another variable with a potential monitoring role of the conduct of the executive board. A member of the executive board is defined as independent from the CEO if he or she joined the management board before the current CEO was appointed to their current position. This definition is adapted from a study by Landier, Sraer and Thesmar (2006) who find that companies with a lower proportion of independent managers exhibit a lower operating performance as measured by the net margin and return on equity. Their interpretation is that independent top-ranking managers act as a monitoring mechanism imposing CEO discipline.
Finally, supervisory boards have the task of monitoring management on behalf of shareholders. Boards must approve of, and may offer advice on, firm strategy and major business decisions. They are also in charge of executive compensation, monitoring of operational risk and internal audit. Studies on board characteristics which may influence the quality of their decision making efficiency highlight the relationship between the size of the supervisory board and company value. Along these lines, Yermack (1996) and Eisenberg, Sundgren and Wells (1998) provide empirical support in favour of smaller boards, showing that company performance is inversely related
11
to board size. The underlying explanation for these findings is that efficient communication within the supervisory board breaks down when there are numerous directors to monitor management and agree on company strategic decisions.
B. Non-executive level
The second level on which the relative remuneration difference between corporate ranks is analyzed is that between employees below the level of the board of management and the members of the executive board. Table I shows that approximately 64% of CEOs surveyed in this paper were previously non-executive employees within the company. Out of the sample of non-CEO executive members, 68.13% were appointed to the board from the non-executive rank. Future CEOs spend on average 7.70 years at the non-executive level within the company. Likewise, other board members spend 6.17 years. This promotion channel from the non-executive level to the board of management is therefore significant in practice, and this paper hypothesizes it is relevant in increasing non-executive employee productivity through the prospect of a higher compensation associated with promotion to a higher corporate rank, in this case possibility of promotion to the executive rank.
B.1. The relative difference in remuneration between executive and non-executive employees
As between the CEO and other executive managers, the remuneration gap between members of the board of management and employees below this level is likely to act as an incentive for employees to strive to be promoted to this rank. Empirical literature on the remuneration gap between employees and executive managers is scarce. Heyman (2005) analyzes the effect of wage dispersion, which is the variation in compensation unexplained by worker characteristics, such as gender, education or labor market experience. In this sense, it is a study of wage inequality. His paper shows that wage dispersion is positively related to company profitability as it increases worker effort and productivity. The main hypothesis to be tested in this paper at the non-executive level is:
Hypothesis 5: Company performance is positively related to the relative difference between the
12
B.2. Outcome of internal promotion to CEO position from the non-executive rank
Another relationship of interest is that between company performance and whether the current CEO has previously been employed as a non-executive employee of the company (aside from any positions on the executive board, as these are analyzed separately). If this condition holds, it may reflect a company policy of internal promotion to CEO position from among employees below the level of the executive board. As such, this would encourage employee motivation to supply effort to achieve promotion, thereby increasing company performance. Therefore, analogous to hypothesis 4:
Hypothesis 6: Companies where the current CEO has previously been employed as a non-executive
company employee exhibit a higher performance.
B.3. Inside executives: outcome of internal promotion to executive manager level
Inside executives are defined in this paper as members of the executive board, excluding the CEO, which have been promoted to the management board from within the company. Similarly to the preceding hypothesis concerning the CEO, the proportion of inside executives is expected to have a positive effect on performance, assuming a high percentage of inside executives reflects a company-specific policy favoring internal promotion over outside hires.
Hypothesis 7: Companies where a higher proportion of executive board members, excluding the CEO,
were appointed from within the company exhibit an increased performance.
While many studies focus on the effect of insiders on supervisory boards, literature on the effect of insiders on executive boards is scarce. It is expected for the proportion of inside managers to have a similar effect to that of the positive influence on company performance found by Ang and Nagel (2011) related to insider CEOs.
II. Data collection and definitions
A. Sample selection
The sample data used in this thesis consists of companies currently (July 1st, 2013) listed on the AEX and AMX stock exchange indexes, with annual data for the period 2008 to 2012. These are
13
large and mid-sized firms in terms of market capitalization. Accounting and financial data for the sample of companies is extracted from Datastream, in Euros. All information concerning executive and employee remuneration variables, executive characteristics, large shareholders and supervisory board data is manually collected from company financial statements. Where this information cannot be found in annual statements, the data collection is supplemented by internet searches on websites such as Management Scope or Bloomberg Businessweek.
To arrive at the final sample of companies, the methodology proceeds as follows. Financial institutions are excluded from the sample. Furthermore, companies with insufficient annual financial data for the period 2007 to 2012 are excluded. This condition is necessary because control variables in the analysis that follows include a one-year lagged value of the dependent financial variable. Lastly, companies where executive characteristics and remuneration data could not be reliably retrieved are also left out of the analysis. The final sample comprises 29 corporations.
B. Variable definitions and computations
B.1. Measures of company performance
Company profitability is proxied as in Landier, Sraer and Thesmar (2006) by the net margin for the current year, calculated as the ratio of net income to total revenues. Additionally, the return on assets is used as an alternative measure of company performance, computed as the ratio of company operating income to total book value of assets, as in Sraer and Thesmar (2006). The analysis further includes a measure of performance on the stock market, calculated as the 12-month holding period return on company common stock. This is computed as in Brookman and Thistle (2009), as capital gains plus any dividends awarded during the year, divided by the initial investment in the stock. A summary of the computation of these variables can be found in the appendix, in section A.1.
B.2. Measures of explanatory variables
The collection and computation of executive and employee remuneration data proceed as follows. Annual executive salary, bonus, other cash compensation, options and share grants are collected for every individual executive on the board. All remuneration data collected is in Euros. The
14
final sample used to compute remuneration differences includes only information associated with executives who have been on the management board for at least a year and those executives who are part of the board at year-end. This ensures any temporary positions or members who resigned or retired from the board during the year are excluded. Moreover, this selection ensures that there is no bias in the computation of remuneration gaps since only full-year compensation data is included. Option awards are measured as the fair value of options granted in the year under study. Where the total fair value is not reported, the fair value per option is searched for and multiplied by the total number of options granted. This approach is analogous to the one used by Ittner, Lambert and Larcker (2003). The fair value of share grants is likewise collected. Where the total fair value or fair value per share are missing, the market stock price at year end is used as a proxy for per-share value, as in Ryan and Wiggins (2004). Executive total compensation is defined as the sum of all remuneration components to a member of the board of management: Executive salary, Executive bonus, Executive
other compensation, Executive option grants and Executive share grants. Executive total cash compensation is defined as the sum of the Executive salary, Executive bonus and Executive other compensation components.
The analysis at the level of the executive board focuses on the remuneration gap between the CEO and other executive managers. It is therefore necessary to separate executive compensation granted to the CEO from that of other members of the management board. CEO total compensation therefore represents the sum of all remuneration components awarded to the CEO during the year:
CEO salary, CEO bonus, CEO other compensation, CEO options and CEO shares. On the other hand, Other executive total compensation represents the average total remuneration to a member of the
board of management other than the CEO, and is composed of Other executive salary, Other executive
bonus, Other executive other compensation, Other executive options and Other executive shares. With
this in mind, we define the compensation gap (relative difference) at the level of the executive board as CEO compensation divided by the average compensation of other executives on the board:
15
The remuneration gaps between the individual components of total compensation awarded to executives are defined in a similar manner:
Salary gap = CEO salary / Other executive salary (2) Bonus gap = CEO bonus / Other executive bonus (3) Other compensation gap = CEO other compensation / Other executive other compensation (4) Options gap = CEO options / Other executive options (5) Shares gap = CEO shares / Other executive shares (6)
To compute annual remuneration variables for non-executive employees, the total values for the items reported as “Salaries and Wages” and “Other personnel costs” are first collected from company annual statements. The “Salaries and Wages” item includes any bonuses awarded to employees during the year and is therefore comparable to compensation paid out to executives as salaries and bonuses.
Therefore, the next step is to subtract the total value of Executive salary and bonus (Executive
salary plus Executive bonus) awarded to all executives on the board from the item “Salaries and
Wages” and likewise the total value of Executive other compensation awarded to all executives from the item “Other personnel costs”. The total amounts of Executive salary and bonus as well as
Executive other compensation subtracted from the items “Salaries and Wages” and “Other personnel
costs” include compensation to all executives on the board, including that of any members who were on the board for less than one year or members who retired or resigned during the year. This ensures there is no upward bias in the computation of the per-employee compensation that follows.
Finally, the items “Salaries and Wages” and “Other personnel costs”, now with amounts awarded to executives excluded, are divided by the number of full-time employees reported for the year by the company, excluding the number of executive board members. Where the number of full-time employees is not disclosed in annual statements, this variable is replaced by the average number of employees reported for the year. The resulting per-employee remuneration variables are denoted
Employee salary and bonus and Employee other compensation. Employee total cash compensation is
16
The remuneration gap analysis between employees and executives focuses on these two remuneration categories only, Employee salary and bonus and Employee other compensation, since options and shares are frequently not awarded to all employees, but rather to a limited category comprising only high-ranking employees. Dividing the total reported fair value of options or shares by the number of full-time employees would therefore be misleading. For this reason, only binary variables Employee options and Employee shares are created as in Jones and Kato (1995), if any options have been awarded to employees outside the executive board during the year.
The following remuneration gap variables are therefore defined at the level between executive and non-executive employees:
Total cash compensation gap = Executive total cash compensation / Employee total cash compensation (7)
The individual components of the total cash compensation gap are similarly defined:
Salary and bonus gap = Executive salary and bonus / Employee salary and bonus (8) Other compensation gap = Executive other compensation / Employee other compensation (9)
A summary of remuneration variables and their computations is included in appendix A.2.
In addition, data on the age of every individual executive on the board at year-end is collected from annual statements. This data is only collected for executives who were on the board throughout the entire year, as in the case of data collected to compute the remuneration gap. For non-executive employees, only two companies from the data sample reported usable statistics on average non-executive employee age. Age statistics are therefore only collected for non-executives. With this data the following variables are constructed: CEO age is the age of the company CEO at year-end while Other
executive age is the average age of executives on the board at year-end, excluding the CEO. Age difference then represents the difference between the age of the CEO and the average of that of other
executive board members:
17
Furthermore, for every executive, data on whether they were previously a non-executive employee of the company before joining the board is collected. In the case of the CEO, it is also taken note of whether he or she was a member of the board before being promoted to CEO position. Tenure as a board member, non-executive employee or CEO (if applicable), measured in years, is collected for every executive on the board.
The number of executives on the board is denoted as Executive board size and includes all executives who have served on the board for at least a year (therefore at least since the beginning of the current year) and are part of the board at year-end. Inside board CEO is a binary variable which takes the value 1 if the current CEO has been promoted to this position from among the members of the executive board at the time. Employee CEO is also a binary variable taking the value 1 if the current CEO was an employee of the company in the past, aside from any positions on the executive board. Finally, Inside executives represents the fraction of members on the board of management, excluding the CEO, which were promoted to the board from among company non-executive employees.
B.3. Measures of governance control variables
The explanatory variable Independent executives is defined as the fraction of executive board members which were part of the board before the current CEO was appointed to this position. Independent executives are identified by comparing an executive’s tenure on the board to the tenure of the current CEO. Those executives whose tenure on the board is longer than that of the current CEO are defined as independent. To compute the fraction of independent executives, the number of independent executives is divided by the total number of executives on the board, excluding the CEO.
In addition, the analysis includes the variable Large shareholders, which measures the number of shareholders owning at least 5% of company equity, and Supervisory board size, which represents the number of directors on the company supervisory board.
18
B.4. Measures of financial control variables
Control variables included in the analysis that follows consist of firm size, market-to-book ratio, financial leverage, risk, and a one-year lagged value of the independent variable measuring company performance. Lagged performance is therefore, depending on the analysis, the one-year lagged value of net margin, return on assets and company stock performance, respectively. These variables are defined in section B.1. Firm size is measured as in Brookman and Thistle (2009) by the natural logarithm of the market value of assets. Financial leverage is computed as the value of total company liabilities divided by the book value of assets, following the approach by Mehran (1995). Risk is defined as the standard deviation of daily stock returns over the prior year. This definition is adapted from Core et al (1999), who use 5 prior years. Due to the limited availability of company data for years prior to the sample period studied, this thesis uses stock returns over one prior year. Following the definitions by Landier, Sraer and Thesmar (2006), the market to book value represents the ratio between the market value and book value of assets, as a proxy for company growth opportunities. Finally, all models include year- and company-fixed effects. Further details on the definitions of these control variables can be found in the appendix, in section A.1.
B.5. Descriptive statistics
Tables II, III and IV in appendix B present descriptive statistics of all company-year variables used in the analysis. With 5 years of data over the period 2008 to 2012, and 29 companies surveyed, the maximum sample size of company-year observations is 145. A number of observations are lost due to the fact that the executive boards of some companies were composed, in one or more years, of only the CEO, resulting in less data for several variables.
From the collected data, Table II shows that CEOs are, on average, 2.79 years older than other members of the board. This supports the assumption that other executive board members have more career-related promotion opportunities within the company, being younger than the current CEO. Moreover, in 62% of the cases across company-year observations CEOs were appointed from among
19
executive board members. This provides support for the hypothesized role of promotion to CEO rank from among the members of the board.
The average fraction of executives (excluding the CEO) which were appointed to the board from inside the company is 61%, with a median of 75%. Moreover, in 62% of the cases across company-year observations CEOs were previously non-executive employees of the company (regardless of any potential positions on the executive board before being appointed CEO). This shows that between a half and three quarters of executive board members were appointed from the non-executive employee level across company-year observations and supports the assumption that the promotion channel from non-executive to executive rank is at work within the data sample.
Furthermore, the table shows that the average number of members on the executive board in the sample is 3, with a minimum number of 1 (in which case this is the CEO) and a maximum of 6. Additionally, about a third of the members of the executive board are defined as independent from the CEO, on average. However, the median percentage of independent executives is 0, showing that most boards have no executives which are defined as independent from the CEO. Finally, the average supervisory board size is 6 across the sample, while the average company in the sample has 3 significant shareholders.
Table III focuses on the relative executive compensation gap. Absolute levels of executive compensation are reported in Panels A and B, while Panel C shows relative remuneration gap statistics. Executive compensation is separately reported as CEO compensation in Panel A, and in Panel B as the average compensation of other executive board members, excluding the CEO. Panel C shows the relative total compensation gap as well as the distinct components of the relative remuneration gap between the CEO and the average of other executive board members. Panel C shows that, indeed, CEOs earn more on average than other executive board members. Specifically, total compensation is on average 1.65 times higher for CEOs. The separate components adding up to the total executive compensation gap are all likewise higher than 1 on average, the highest ratio being that of the Other compensation gap.
Table IV presents descriptive statistics related to the relative remuneration gap between executive and non-executive employees. As mentioned previously, the analysis focuses on the
20
difference in cash compensation, namely salaries and bonuses, and other benefits remuneration. As would be expected, the differences in this case are significantly larger than in the case between CEOs and other executives. Specifically, executive board members (including the CEO) earn 24.35 times the average employee cash compensation: 25.36 times the amount in salaries and bonuses and 13.09 times the amount in compensation classified as other benefits.
III. Methodology
A. Measuring the effect of compensation levels on company performance
In order to test the hypotheses outlined in section I, ordinary least squares models with control variables and company- and year-fixed effects are constructed. Panel data is used to avoid any endogeneity biases that may arise due to omitted variables or the possibility of reverse causality.
As a first step in the analysis, company performance measures are regressed on executive and subsequently on non-executive employee compensation. The equation used in the analysis focusing on executive compensation is the following:
Performancei,t = α + β1Executive compensationi,t + β2Executive agei,t + β3Zi,t + γi + λt + ɛi,t (11)
Where Performancei,t stands in turn for company net margin, return on assets and stock
performance for company i in year t. Compensation variables for the CEO and other executives are denoted by the term Executive compensationi,t. Natural logarithms of executive compensation
variables are used in order to work with a normalized distribution. Moreover, a value of 1 is added to the value of stock options and stock grants to avoid taking the natural logarithm of zero. The variable
Executive compensationi,t therefore takes the values of the natural logarithms of Executive total
compensation and individual components of total executive compensation: Executive salary, Executive bonus, Executive other compensation, Executive option grants and Executive share grants, separated
into CEO compensation and the average compensation of other members of the executive board. Furthermore,the term Executive agei,t stands for the age of the CEO and the average of that of other
21
Market-to-book value, Financial leverage and Risk. Contingent on the measure of performance
represented by the dependent variable, Lagged performance is the one-year lag of net margin, return on assets and stock performance, respectively. The terms γi and λt represent in turn firm-fixed and
year-fixed effects. The results of the regressions are shown in Table V, VI and VII with the dependent variables net margin, return on assets and stock performance, respectively.
A similar equation is used for the regression of company performance measures on non-executive employee compensation:
Performancei,t = α + β1Employee compensationi,t + β3Zi,t + γi + λt + ɛi,t (12)
The variable Employee compensationi,t includes in turn the natural logarithm of Employee
salary and bonus and Employee other compensation, as well that of as the sum of these two
components: Employee total cash compensation. Additionally, binary variables Employee options and
Employee shares are created. These are equal to 1 in the cases where options or shares were granted to
any employees outside the executive board during the year. The variables Zi,t, γi and λt are equivalent to
those included in equation (11). The output of these regression models can be found in Tables VIII, IX and X.
B. Measuring the effect of the relative compensation difference and long-term career horizons on company performance
The next model serves as the basis of the first half of the main analysis, and measures the effects on company performance of the relative remuneration difference and of the proxies for long-term career incentives at the level of the executive board:
Performancei,t = α + β1Compensation Gap1i,t + β2Age differencei,t + β3Executive board sizei,t +
β4Inside board CEOi,t + β5Wi,t + β6Zi,t + γi + λt + ɛi,t (13)
where Compensation Gap1i,t represents in turn the total relative compensation gap between the CEO
22
the total relative compensation gap: Salary gap, Bonus gap, Other compensation gap, Options gap and
Shares gap. The variable Performancei,t comprises the net margin, return on assets and company stock
performance, respectively. Financial control variables are included in the term Zi,t as before, and
executive board monitoring controls outlined in section II.B.3 are included in Wi,t. Output for the
regressions can be found in Tables XI, XII and XIII.
Similarly, at the level between executive and non-executive employees, the relative compensation and company performance model is the following:
Performancei,t = α + β1Compensation Gap2i,t + β2Employee CEOi,t + β3Inside executivesi,t + β4Zi,t + γi +
λt + ɛi,t (14)
Where Compensation Gap2i,t includes measures of the relative compensation gap between
executive and non-executive employees: Total cash compensation gap, Salary and bonus gap and
Other compensation gap. Employee CEO is a binary variable taking the value 1 if the current CEO
was previously a non-executive employee of the company and Inside executives measures the proportion of executives on the company board, excluding the CEO, who were previously non-executive employees of the company. Financial control variables Lagged performance, Firm size,
Market-to-book value, Financial leverage and Risk are included in the term Zi,t. Regression results are
presented in Tables XIV, XV and XVI.
IV. Empirical results and analysis
In this section, the results of this study are discussed with respect to the hypotheses outlined in section I. A preliminary analysis tests for the effect of absolute compensation levels of executive and non-executive employees on company performance. Subsequently, the main analysis focuses on the relative difference in remuneration at the level of the board of management and at the level between members of the executive board and non-executive employees. The models also include tests for the impact on company performance of internal promotion to key executive positions. Moreover, the executive-level analysis includes as explanatory variables the age difference between the CEO and other executives, the size of the executive board and executive monitoring mechanisms proxied by the
23
size of the supervisory board, number of significant shareholders and percentage of independent executives on the management board.
A. Analysis of the effect of compensation levels on company performance
The results of the preliminary analysis testing the effects of the levels of compensation and ages of executive board members on company profitability are shown in Table V of appendix B. From this analysis, CEO total compensation emerges as a significant factor with a positive relation to company profitability in every model in which it is included. It can therefore be concluded that the level of compensation alone is an effective incentive for CEO performance. From among the distinct components of CEO total compensation, the value of shares awarded to the CEO emerges as significant in model (2); however, the coefficient becomes insignificant when sufficient control variables are added in model (6). The component of CEO compensation which surfaces as significant in model (6) is the value of options awarded. The coefficient is negative, likely indicating a shorter time horizon associated with the company for CEOs. As options frequently have vesting periods of at least a few years, and are not awarded if the executive leaves the company, this negative performance response to option awards can indicate that CEOs anticipate a higher possibility of resigning or retiring before options have vested. In contrast, the coefficient on option awards to other executives is positive and significant, indicating that, as opposed to the CEO, other executives on the board have longer-term horizons and stronger career incentives tied to the company. Another significant factor in model (6) is CEO age. The positive coefficient indicates that company profitability benefits from the experience of an older CEO, over and above any effect of a short-term horizon associated with the company for a more senior CEO.
The model shown in Table VI uses an alternative measure of company performance, namely the return on assets. This analysis shows that the results found related to CEO total compensation, CEO age and option compensation to CEOs and other executives are robust with respect to the measure of company performance used.
24
In Table VII, company performance is measured as the return on common stock over the year. Remuneration classified as “Other” awarded to the CEO is found to be positively related to stock performance when sufficient control variables are added to the regression model. This component of remuneration includes, among others, all costs that CEOs incur while carrying out company-related activities, including, for instance, travel costs, team-building activities and training costs. It is likely that high expenses allocated to this category indicate that CEOs spend a significant amount of time and resources promoting company activities, interacting with stakeholders (including employees) and generally working to build value, and this has a positive effect on stock performance. The coefficient on other compensation awarded to executives excluding the CEO is also positive, but insignificant.
However, the coefficient of Other executive salary is negative and significant. The effect of
CEO salary is also negative, but not statistically significant. The annual salary component of executive
compensation therefore has a negative effect on stock performance, indicating that investors take notice of this figure and react negatively to it. This is in line with the discussion on the perceived fairness (or unfairness) of high executive compensation.
The effect of non-executive employee compensation on company profitability is investigated in Table VIII, using the net margin as a first measure of company performance. The positive effect of remuneration classified as “Other” indicates that companies who spend more funds on non-executive employee training, travel, team-activities, housing and other such costs are more profitable. It is therefore likely that investing in this way in non-executive employees’ development, team-building activities and other benefits raises their productivity and expertise, and contributes to a higher company operating performance. The effect does not appear, however, when return on assets is used as a company performance indicator in Table IX.
Companies awarding options to employees outside the executive board perform better on the stock market, as indicated in Table X. The coefficient on the binary variable for share awards is also positive, but insignificant. The use of long-term incentives for non-executive employee compensation in the form of awarded options is therefore found to have a positive relation to common stock performance, likely reflecting non-executive employees’ long-term horizons associated with their
25
company. Indeed, in their 2005 paper, Oyer and Schaefer conclude that employee retention is one of the main reasons for which companies award options to employees below the executive rank.
B. Analysis of the effect of the relative difference in compensation between corporate ranks on company performance
This section begins with a discussion of the results found regarding company performance and the relative difference in remuneration at the level of the executive board. Subsequently, it looks at the results of the analysis carried out at the level between executive and non-executive employees.
The executive-level analysis finds evidence to show that the relative compensation gap between CEOs and other executives is positively related to company profitability. From the results shown in Table XI, the Total compensation gap has a positive and significant coefficient in every model it is included in. This provides evidence in support of hypothesis 1. From among the distinct components of the Total compensation gap, the Other compensation gap appears to be driving this positive effect. This would imply that executives are particularly motivated to supply effort and be promoted to CEO position due to the fringe benefits associated with this rank. The coefficient on the
Bonus gap is also positive, but becomes insignificant when sufficient explanatory variables are
included in the model. The negative effect of the relative difference in the value of stock options awarded to CEOs versus other executives is an interesting finding. It appears that awarding less option value to executive board members relative to the CEO in fact has a negative impact on profitability, potentially through a demotivating effect and consequently a decrease in their performance. This implies that executives other than the CEO would be interested in this long-term component of remuneration, supporting the assumption that they have relatively longer time horizons and longer-term interests in company performance.
There is also evidence in support of hypothesis 2, in that the age difference between CEOs and other executives has a positive effect on company performance. Companies where executives on the board are younger than the CEO perform better in terms of profitability measured by the net margin, suggesting that these subordinate executives are motivated by longer-term horizons and more possibilities for future promotion to CEO position. Is it possible that the positive effect is simply due
26
to fresher perspectives and innovative ideas that boost profitability, brought by younger executives. However, if this were the case, this effect would have appeared in the analysis of the previous section, in tests for the effect of CEO age and average age of other executives on performance shown in Table V. The relativity to CEO age is therefore an important consideration.
Executive board monitoring mechanisms found significant are supervisory board size and the number of significant shareholders. The negative relation between the size of the supervisory board and profitability is in line with findings by Yermack (1996) and Eisenberg, Sundgren and Wells (1998). Supervisory board effectiveness therefore indeed decreases with board size. There is also a positive relation between the number of significant shareholders and company profitability. Blockholders are therefore an important supervision mechanism, generating positive changes in company policies and activities, as found by Gillan and Starks (2000). There is, however, no evidence in this analysis to indicate any relation between the fraction of independent executives on the board and company profitability.
The findings with respect to the Other compensation gap, Options gap and Supervisory board
size are replicated in the analysis shown in Table XII, demonstrating that these variables similarly
affect company return on assets.
The analysis on stock performance, presented in Table XIII, corroborates the evidence found on the positive relation between company profitability and the difference in age between CEOs and other executives. Companies where key officers are younger than the CEO therefore also perform better in terms of stock returns. Additionally, Table XIII indicates that company performance on the stock market is worse for companies where CEOs were appointed from inside the board. This implies that investors perceive outside hires as better choices for CEOs.
Turning to the analysis carried out between executive and non-executive employee levels, Tables XIV and XV show models where the net margin and return on assets are regressed in turn on the relative compensation gap between executives and non-executive employees and variables indicating whether the current members of the management board were previously non-executive employees. The relationships between these variables and company performance are not statistically significant. These findings are mirrored by the analysis in Table XVI, with stock market performance
27
used as the regression dependent variable. A comparison of these results with those found at the executive board level suggests that relative compensation and long-term career incentives determined by the prospect of a future promotion within the company have stronger effects at higher levels of management. The stronger effect found at this level is also likely due to the fact that executives are in a better position drive company performance, as they directly participate in decisions regarding company strategy and major value-enhancing activities such as mergers or acquisitions.
V. Drawbacks and suggestions for further research
This thesis focuses on annual hand-collected data from 29 companies over a time span of 5 years, from 2008 to 2012, adding up to a total of 145 observations. Future research would benefit from a larger data sample as this would increase the statistical significance of regression results.
Furthermore, using averages of non-executive employee compensation, as is done in this thesis, is likely to mask important differences in the distribution of compensation across employees and corporate ranks below the executive level. High-ranking non-executive employees are expected to be receiving considerably higher compensation than lower-ranking employees. For this reason, using the total average amount of compensation paid to non-executive employees introduces an amount of inaccuracy compared to data retrieved for members of the executive board, for which amounts paid out to each executive are disclosed in annual statements. Ideally, a research on non-executive employee compensation should take into account actual individual employee compensation as well as employee-specific corporate rank within the company and other controls such as experience, gender or qualifications. Alternatively, compensation could be averaged out over employees with the same rank within the corporation. However, companies do not report such data in annual statements and individual non-executive employee data is confidential, except for total amounts reported under company overhead expenses. A better way to carry out such data collection would therefore be by using anonymous surveys distributed to employees of a sample of companies, where respondents would disclose their compensation, corporate rank and other relevant characteristics. This would also allow research on long-term compensation components awarded to non-executive employees, such as
28
grants of options or restricted stock as well as an analysis of the age gap effect at the level of non-executive employees.
VI. Summary and conclusions
This thesis researches the effect on company performance of the relative difference in compensation between corporate ranks within public corporations. By doing so, it focuses on the compensation incentives of employees subordinate to the CEO and their long-term prospects associated with the company they are currently working for, such as those provided by the possibility of internal promotion. Specifically, this paper hypothesizes that employee productivity, and therefore company performance, is positively related to the relative difference in compensation between corporate ranks. The higher relative difference in remuneration is expected to act as motivation for employees to pursue promotion to a higher rank within the company. For this reason, the analysis also investigates the outcome of internal promotion to key executive positions within the company, rather than filling these positions by hiring externally. Additionally, this paper takes into account the effect of board size and relative age difference between corporate ranks at the executive level. The sample used in this study comprises 29 large and medium-sized companies in terms of market capitalization listed in the Netherlands, from which annual data variables are derived for the period 2008 to 2012.
The analysis of this paper is carried out at the level of the executive board and subsequently at the level of non-executive employees. The research performed at the executive level centers on the members of the executive board subordinated to the CEO. At this level, where the possibility of a future promotion to CEO position is comparatively high, the relative remuneration gap between the CEO and other executives is expected to act as motivation for these executives to supply effort in order to achieve promotion and at the same time to ensure long-term company value preservation. Other factors affecting the likelihood of promotion are taken into account within the analysis, such as whether the current CEO previously served as a board member, the age difference between the CEO and the average of that of other executives, and the size of the executive board.
At the non-executive level, the analysis tests for the effect on company performance of the ratio of average executive compensation to the average compensation of a non-executive employee, a
