Busy directors and executive compensation

Busy Directors and Executive Compensation

MSc Thesis

MSc Accountancy & Control, variant Control Amsterdam Business School

Statement of Originality

This document is written by student Seyfullah Kaya who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

Abstract

The impact of corporate governance on executive compensation has received increasing attention from the academic literature due to accounting scandals and the recent financial crisis. Prior research indicates that not every aspect of corporate governance in relation to executive compensation is conclusive. The purpose of this empirical research is to examine the relationship between busy directors and executive compensation. Furthermore, to contribute to research that seeks to provide a better understanding of the impact of corporate governance on executive compensation. The sample of this study consists of 1,545 observations and 309 unique publicly traded companies in the United States, which are listed at the Standard & Poor’s 500. Overall, the main findings of this examination show that there is a positive association between the busyness of directors and the level of compensation that chief executive officers obtain. Furthermore, the research did not find enough evidence to support a relationship between pay-for-performance and director busyness.

Acknowledgement

This thesis is written as part of the Master’s program Accountancy and Control of the University of Amsterdam. Before this thesis starts with the content of the subject, I would like to express my sincere gratitude to my supervisor Mr. Bo Qin. Without his advice and assistance, the thesis might not have been accomplished. Besides my supervisor, I wish to express my greatest thanks to my family, friends and colleagues, who have supported me unconditionally, especially my parents. I am grateful of the continuous encouragement, support, and attention you have given me.

Amsterdam, 2015

Table of Contents 1. Introduction ... 5 Background ... 5 Research question ... 6 2. Theoretical framework ... 7 Agency theory ... 7

Optimal contracting theory ... 7

Managerial power theory ... 8

Corporate Governance ... 8

The Sarbanes-Oxley Act ... 9

The Dodd-Frank Act ... 9

Board of directors ... 10 Board capital... 11 Executive compensation ... 13 Base salary ... 13 Annual bonus ... 13 Long-term incentives ... 13

Restricted stock and option grants ... 14

Other components ... 14

3. Hypotheses development ... 15

Busy directors – CEO compensation ... 15

Busy directors – Pay-for-performance sensitivity ... 16

4. Research methodology ... 17 Sample selection ... 17 Research method ... 18 Measurements ... 18 Busy directors ... 18 CEO Compensation ... 19 Interaction term ... 19 Control variables ... 20 Empirical models ... 22 5. Results ... 24 Descriptive statistics ... 24 Main analyzes ... 27 6. Conclusion ... 32 Suggestions ... 33 References... 34

1. Introduction

Background

In recent years, executive compensation has become a subject of scrutiny due to its alleged role in the accounting scandals of 2000-2002 surrounding large publicly traded companies such as Enron, WorldCom, and Tyco (Bebchuk and Fried 2003, 2006; Heron and Lie 2007). Furthermore, concerns with executive compensation have intensified as the financial crisis of 2007-2008 unfolded with compensations being blamed for encouraging excessive risk-taking and contributing to the collapse of the financial sector (Bhagat and Romano 2009). The crisis has been considered by many economists to have been the worst financial crisis since the 1930s Great Depression as it threatened the continuity of many large institutions, which was prevented by bailouts provided by national governments. Nevertheless worldwide market values still dropped and it played a significant role in the failure of key businesses, declines in consumer wealth, and a downward spiral in economic activity leading to the 2008-2012 global recession and that of the European sovereign debt crisis. In January 2011, during the aftermath of the crisis the United States Financial Crisis Inquiry Commission concluded in its report that: “the crisis was avoidable and was caused by widespread failures in financial regulation, dramatic breakdowns in corporate governance including too many financial institutions acting recklessly, and taking on too much risk.”1

Although the scandals and the crises increased the attention of legislators, the media, and the general public on executive compensation, particularly the relation between corporate governance and executive compensation received considerable attention from the academic literature (Adams et al. 2008; Armstrong et al. 2012; Conyon 2014). While Ozkan (2009) argues that corporate governance could reduce the conflicts between shareholders and executives, which in turn could impact the compensation policy. Armstrong et al. (2012) point out that studies document that executive compensation is greater when executives are powerful and corporate governance is weak. The findings of these studies indicate that some characteristics of corporate governance in relation to compensation are more conclusive than others. For example Core et al. (1999) and Brick et al. (2006) conclude that a chief executive officer (hereafter referred to as “CEO”) will receive a higher compensation when the same person is also the chairman of the board, who has to supervise procedures about

1

how the CEO will be evaluated and compensated. Jensen (1993), Core et al. (1999), and Bebchuk et al. (2006) conclude that the compensation of a CEO will be higher when the size of a board is larger because individual directors may focus less on the affairs of the company. In contrast, the relation between the CEO ownership and compensation is not that conclusive. While Cyert et al. (2002) concludes a positive relation between CEO ownership and compensation levels, other studies such as Lambert et al. (1993), Kahn (2005), and Sapp (2008) find a negative association. Another case in point is the relation between the number of directorships held by corporate directors on the board of a company and executive compensation. Tarkovska (2012) points out that monitoring of a board could be diminished by having busy boards and Fich and Shivdasani (2006) argue that a busy board will lead to weak corporate governance. Whether holding multiple directorships impairs the ability of a director to monitor management has become a controversial topic on his own that even spawned proposals for governance reform. Despite the controversy, research concerning the effect of multiple directorships is limited (Ferris et al. 2003).

The purpose of this study is to fill this gap by investigating the relationship between busy directors and executive compensation using a sample of publicly traded companies in the United States. Furthermore, contribute to research which seeks to provide a better understanding of the impact of corporate governance on executive compensation. Therefore, the following research question is formulated:

Research question

“What is the influence of busy directors on the compensation of executives in publicly traded companies in the United States”?

The remainder of the thesis is organized as follows. Chapter two provides an overview of the existing theories and prior literature on the relation between busy directors and executive compensation. Chapter three discusses the developed hypotheses. Chapter four addresses the research methodology used in this study. Chapter five shows the results of the conducted empirical analysis. The last chapter includes the conclusions of the study and suggestions for future research.

2. Theoretical framework

The relation between busy directors and executive compensation is explained in the literature through the use of existing theories like agency theory, optimal contracting theory, and managerial power theory. These theories should also provide an understanding of how executive compensation forms and the role of busy directors.

Agency theory

Bertrand et al. (2001) and Conyon (2014) indicate that the compensation of executives is commonly viewed through the lens of the principal-agent models. Under this contracting view, compensation is used to diminish the moral hazard problem that arises because CEOs regularly own little of the companies they control. The principal-agent model is a part of the agency theory, which rooted in economics and finance thinking (Jensen and Meckling 1976) and has become a cornerstone of the corporate governance field. In general terms, an agency is a relationship between two parties, where an agent (management) represents the principal (shareholders) in transactions with a third party. As governance mechanics serve to address agency conflicts, they can arise when the principal is unable to verify what the agent is doing and when the principal and agent have diverse attitudes towards risk. This may lead to the principal and agent taking different actions. One way to partly alleviate principal-agent problems is by aligning the interest of executives with those of shareholders through the use of compensation packages that make the agent’s compensation vary according to the company’s performance (Jensen and Murphy 1990).

Optimal contracting theory

Introduced by Jensen and Meckling (1976), the optimal contracting model assumes that compensation is set by a remuneration commission consisting of independent non-executive board directors on behalf of the shareholders. This at arm’s length bargaining process should result in a cost-effective compensation package that provides risk reluctant executives with incentives to enhance shareholder value. Since this model has been the leading theory on compensation, Bebchuk et al. (2006) brand it as the official view. However, Knop and Mertens (2010) point out that various empirical studies provide evidence inconsistent with the optimal contracting theory, such as Finkelstein et al. (1995), Yermack (1997), Core et al.

(1999), and Bertrand and Mullanaithan (2000).The results of these studies have led to a new theory, which is called the managerial power model.

Managerial power theory

According to Bebchuk et al. (2002) managerial power theory explains that executives have the power to influence their own compensation and that they use their power to extract rents at the expense of shareholders. Furthermore, the more power an executive has, the more an executive will be able to inflate its own compensation. This power is determined by how the board of directors is composed and the ownership structure of the company. Bebchuk et al. (2002) point out for instance that the fraction of independent and inside directors is significant for the influence a CEO has in a company. Changes in compensation can be expected when the constraints that executives and directors face change.

Overall, where the optimal contracting theory assumes that executive compensation is used by shareholders to reduce agency costs, the managerial power theory sees inefficient executive compensation arrangements as an agency problem in itself.

Corporate Governance

According to the evidence presented by Core et al. (1999) companies with weaker corporate governance structures have greater agency problems and that CEOs at companies with such agency problems receive higher compensation. Merchant et al. (2007) define the term corporate governance as a set of mechanisms and processes that help ensure that companies are directed and managed to create value for their owners while concurrently fulfilling responsibilities to other stakeholders such as suppliers, employees, and society at large. The focus of corporate governance is on controlling the behaviors of executives and also while less directly, those of the other employees in the company. There are for the most part two corporate governance orientations, the Rhineland model approach, which has a broader concern for the rights of other stakeholders, and the Anglo-American approach, which prevails in the United States with focus on the primacy of shareholders as the beneficiaries of fiduciary duties2 (Merchant et al. 2007). Furthermore, in nations that are influenced by the latter approach of corporate governance usually have a one-tier board structure in which boards consists of all the directors and decisions are made together. In

2

contrast, two-tier boards have a management board consisting solely of executives and a supervisory board consisting only of non-executive outside directors. Regardless of the style of corporate governance, all publicly traded companies are bound by the regulations and rules of the stock exchange on which their shares are listed. Important examples of legislations specifically aimed at strengthening corporate governance in the United States are the Sarbanes-Oxley Act of 2002 and the Dodd-Frank Act of 2010.

The Sarbanes-Oxley Act

Following the introduction of the Sarbanes-Oxley (SOX) Act in 2002, corporate governance became a pressing issue for many companies. All companies registered with the Securities and Exchange Commission (SEC) must comply with the Sarbanes-Oxley Act regardless of whether their headquarters are based in the United States or abroad. The purpose of the Sarbanes-Oxley Act was to improve the transparency and quality of financial reporting in order to restore public confidence in markets after accounting fraud bankrupted large companies such as WorldCom and Enron. Adams et al. (2008) indicate that as a result of the Act boards have become larger, more independent, meet more often, and have more responsibility. One of the provisions worth mentioning is Section 404, which mandated an evaluation of the effectiveness of a company’s internal controls by both management and external auditor and formal written opinions about the effectiveness of those controls.

The Dodd-Frank Act

In the wake of the financial crisis, the Dodd-Frank Wall Street Reform and Consumer Protection Act of 2010 was an attempt to repair ‘too big to fail’ and other corporate governance problems. The aim of the Act was to avoid another financial crisis by developing new financial regulatory processes that enforced accountability and transparency. Conyon (2014) mentions that when drafting the Act, legislators believed that corporate governance arrangements prior to 2010 were ineffective and more was needed to be done to reduce the excess compensation of executives. An important provision of the Act was providing investors with the opportunity to vote on executive ‘say-on-pay’3 compensation. Overall, the Act provides owners additional control over executive compensation and requires the board of directors and their compensation committees to be more accountable and independent.

3

Board of directors

In publicly traded companies, shareholders usually diversify their risks and own a portfolio of shares in several companies. Individually, they rarely have an incentive large enough to dedicate resources to ensure that the management is acting in the best interest of the shareholders. The usual solution for shareholders is to delegate their authority to monitor the actions of the management to a board of directors, a group of people who have the ultimate decision-making authority in a company. Therefore, the board of directors also set the compensation of executives by acting on the behalf of the shareholders (Hermalin and Weisbach 1998). The board of directors typically consists of inside directors, who are full-time employees of the company and outside directors, whose primary employment is not with the company. The latter group is often taken to be independent directors. Fama and Jensen (1983) argue that outside directors value their reputation as directors because good reputation signals the value of their services to the director labor market. Reputation will be more important to them than to inside directors whose careers are tied to the CEOs. Adams et al. (2008) mention that the Sarbanes-Oxley Act contained a number of requirements that increased the workload of and the demand for outside directors. Consequently, in recent years this has led companies to have majority-outsider boards.

In the United States, the basic fiduciary duties of the board of directors is to make decisions in an informed way, advance company interest over personal ones, be devoted to shareholders interests, and avoid deliberate destruction of shareholder value. Merchant et al. (2007) indicate that in order to carry out their responsibilities, boards must ensure that they are independent and accountable to shareholders. They are solely responsible for the selection and evaluation of the company’s CEO, and ensure the quality of the top management. According to Global Investor Opinion Survey (McKinsey 2002), institutional investors perceive board practices to be at least as important as financial issues such as performance or growth potential when evaluating companies for investment. However, it must be recognized that there are also limits to what a board can do. Directors are expected to advise on corporate strategy but do not develop the strategy. They are expected to ensure the integrity of the financial statements but do not prepare the statements themselves. The board of directors is not an extension of management, but a governing body elected to represent the shareholders’ interests (Larcker and Tayan 2011). Therefore, they cannot be held responsible for the day-to-day management of the company.

Compensation committee

Not all corporate matters are deliberated by the full board of directors. Some of them are delegated to committees. In most companies, there is at least an audit committee4, a compensation committee, and nominating and governance committees5. In accordance with the needs of the company’s industry, there might be other committees such as finance, investment, public policy, technology, and risk management committees. Various stock exchanges have regulations that necessitate companies to have executive compensation approved by a majority of independent directors. Conyon (2014) mentions that such issues are delegated to the compensation committee, which is a specialized committee of the board that only includes outside directors. They are in principle free from the influence of executives they advise. The responsibilities associated with serving on the compensation committee of a company’s board have increased significantly in recent years with the enactment of the Dodd-Frank Act. These responsibilities of the compensation committee are reviewing and approving the compensation of the company’s CEO and other executives, overseeing the company’s benefit plans, and reviewing and making recommendations to the board of directors regarding the compensation of other directors. In order to help the compensation committee with their responsibilities often outside consultants are employed (Merchant et al. 2007).

Board capital

The presence of outside directors in a company is not a random occurrence. Companies want to have outside directors who are distinguished individuals who also have an ability to add value as directors. Hence, these directors usually represent an important constituency or have backgrounds that make them valuable to a board. For instance, many companies have bankers on their boards since they can provide financial expertise, which can result for example in overall debt ratio being lower (Byrd and Mark 2005). Companies that have important government contracts or deal often with the government, place a high value on directors with political connections as they might influence government decisions (Adams et al. 2008). Another common occupation of outside directors is that of being a CEO of another company given that they have an understanding of the issues facing the top management.

4

Required since 1977 by the Securities and Exchange Commission (SEC) for all publicly traded companies. 5

When there is an announcement that CEOs of well-performing companies will be added to the boards this generates positive abnormal returns (Fich 2005). Conyon and Read (2006) argue that serving on the boards of other companies also helps to build the CEO’s human capital. Moreover, Tian et al. (2011) argue that boards have access to better-quality information and develop more effective information processing capabilities if they have high levels of human and social capital. These two elements of board capital deviate from each other as human capital refers to the knowledge, experience, and proficiency embedded within an individual, whereas social capital refers to the access of useful information and resources through social relationships (Hillman and Dalziel 2003), which affect how both individual directors and the board as a whole function (Johnson et al. 2013).

While social capital lends itself to various definitions, interpretations, and practices due to the broad range of outcomes it can explain (Halpern 2005). Social capital researchers in the context of corporate governance argue that directors with high levels of social capital can bring information about other companies’ strategies, the external environment, and prospective managerial talent (Haunschild 1993; Certo 2003). Empirical research in this area confirms the consistent positive impact of high social capital (Johnson et al. 2011). A major source of social capital for boards is the network of external directorship ties of outside directors. For instance, directors who held board seats in strategically related companies were more effective in strategic decision making for similar companies than directors without such network ties (Carpenter and Westphal 2001). Moreover, it can help companies avoid bankruptcy (D’Aveni 1990), contribute positively to company value (Hillman et al. 1999; Perry and Peyer 2005), and serve as a window to the world (Andrews 1980). However, ties can be harmful to shareholders, entrench the management of the focal company, and coordinated action among directors is possible when directors are interlocked6 (Johnson et al. 2013). This is in line with Hallock’s (1997) evidence that the compensation of the CEO is significantly higher in companies with interlocked outside directors. Overall, the findings indicate that outside directorship ties are associated with improving the flow of information to the focal company. This can be in the form of new ideas for the focal company, often to the benefit of company shareholders.

6

Executive compensation

While an executive may be any corporate officer in any company, the source of most attention is the compensation of the CEO of a large publicly traded company (Bebchuk and Fried 2006). The CEO is charged with running a company by instituting the policies and rules set by the board of directors. For this endeavor, the CEO receives compensation from the company. Prior studies point out that compensation levels can be expected to increase with company size (Baker et al. 1988), company performance (Core et al. 1999), and to vary across industries (Ely 1991). Although there is substantial heterogeneity in compensation practices across companies and industries, most compensation packages contain five basic components. These are the base salary, annual bonus, payments from long‐term incentive plans, restricted stock grants, and restricted option grants (Frydman and Jenter 2010).

Base salary

The base salary is the standard wage paid to an executive that typically is the largest share of an annual compensation package. Since salary establishes the executive’s basic standard of living it does not fluctuate in relation to company performance and thus set at market rates. In most cases, the larger the company, the smaller the compensation package is made up of salary (Bebchuk and Fried 2006).

Annual bonus

The annual bonus is a short-term incentive that provides additional cash compensation if the company’s yearly performance exceeds specified targets. Consequently, it is tied to some variant of accounting earnings, which can lead to earnings management7 in order to extract rents from shareholders. Such gains could take the form of increased compensation (Healy 1985; Holthausen et al. 1995).

Long-term incentives

Payments from long-term incentive plans usually refer to grants where the payment is based on performance for a period beyond a year. These long-term rewards are often paid out in cash or stock over several years.

7

Restricted stock and option grants

Lastly, the two final components are granting of shares and options that are restricted in terms of transferability and subject to a time-based vesting schedule. When vested, they are economically equivalent to a direct investment in a company stock. The purpose of option awards is to tie remuneration directly to share prices and thus give executives an incentive to increase shareholder value.

Other components

The remaining types of CEO compensation include perquisites, pensions, and severance pay. Due to insufficient disclosure these types of compensation have sometimes been labeled as stealth compensation that may allow executives to covertly extract rents (Jensen and Meckling 1976; Bebchuk and Fried 2006). Perks comprise a variety of services and goods provided to executives from personal use of company aircraft to club memberships. They appear to be a more general signal of weak corporate governance as reductions in company value upon the disclosure of perks substantially exceed their actual cost (Yermack 2006). The pension plans, which typically consist of non-qualified retirement plans, are offered at the discretion of the company and are reserved for the executives. CEOs often defer receiving their bonuses till their retirement years, storing the cash away in a retirement plant that usually allows them to pay lower taxes once they draw on the cash. Severance pay includes provisions such as the golden parachute, which compensates CEOs if they lose their job due to their company being acquired. The stock market tends to react positively to the inclusion of a golden parachute (Lambert and Larcker 1985).

Overall, both the level and composition of executive compensation packages have changed significantly over time (Frydman and Saks 2010). Prior to the 1970s, one can observe low levels of compensation and largely composed of salaries and annual bonuses. Subsequently to the end of the 1990s, all components grow significantly and differences in compensation across companies widen. The largest increase comes from options, which becomes the single largest component of executive compensation during the 1990s. After the stock market decline of 2002, stock options lost their luster and restricted stock grants have become one of the most popular form of compensation by 2006 (Frydman and Jenter 2010).

3. Hypotheses development

This section includes an overview of the hypotheses that are developed from the reviewed literature concerning busy directors and executive compensation.

Busy directors – CEO compensation

Shareholders and institutional investors criticize companies for appointing directors who hold multiple directorships in various companies, contending that such directors are incapable of effectively monitoring the management of several companies (Ferris et al. 2003). Furthermore, some directors believe that too many board appointments place an excessive burden on directors (Korn 1998). This would mean that overcommitted directors have the potential to be negligent in their oversight or unavailable at critical moments due to being busy. Conversely, having a busy director can bring potential benefits, which can offset the effect of their lack of time. For instance, outside directorship ties may provide boards with expertise in evaluating CEO performance and hence an alternative angle regarding what should be the optimal compensation for the CEO. These ties have been variously viewed as a source of expertise, experience and information (Carpenter and Westphal 2001; Perry and Peyer 2005; Larcker and Tayan 2011). Moreover, busy directors might have sound reputations and high integrity, which are driving factors in the demand for their services. Hence, they are more likely to learn the expertise of assessing CEO performance (Tian et al. 2011) and thereby make positive contributions in determining the level of compensation. However, the findings are inconclusive, studies such as Fich and Shivdasani (2007), Jiraporn et al. (2008) and Hoitash (2011) find that busy directors are associated with less monitoring, excessive CEO compensation, and poorer company performance. On the other hand, the study conducted by Ferris et al. (2003) found no evidence on the relation between busy directors and the degree of monitoring. Nevertheless based on these observations I predict that the amount of directorships held by directors and the level of CEO compensation will be significantly related. Therefore, the following hypothesis is developed:

Busy directors – Pay-for-performance sensitivity

Another controversial issue in corporate governance is whether executive compensation contracts exhibit ‘pay-for-performance’ (Larcker and Tayan 2011). The notion that the level of compensation awarded to an executive should be related to the value of the services rendered during a particular period. Some argue that there is a disconnect between these two and that pay-for-performance does not exist in the United States (Larcker and Tayan 2011). The board of directors can influence an executive to focus on shareholder interests and avoid self-serving choices by designing a compensation package that provides the executive with an incentive to increase shareholder value. While outside directors are not responsible for the company’s performance since they cannot directly influence it, they are held responsible of the supervision of the board given that they are monitoring the company’s operations (Boumosleh 2009), attracting and dismissing board members (Conyon 2014), and set the compensation of the executives (Hermalin and Weisbach 1998). Hence, these directors should know more about establishing appropriate performance metrics for CEOs (Conger et al. 1998). According to Weisbach (1988) outside directors do appear to be more responsive to performance than inside directors in decisions such as CEO retention. Core et al. (1999) investigated a number of governance variables and concluded that companies with busy boards also exhibit lower operating performance and stock market returns in certain years. These findings suggest that having busy directors on a board can fail to be in the company’s interests. Conversely, other evidence suggests that multiple directorships are associated with company success. Kaplan and Reishus (1990) and Booth and Deli (1996) find that there is a positive relationship between a company’s performance and directors who hold multiple directorships. Moreover, Brown and Maloney (1999) point out that a company enjoys superior returns from acquisitions when they have directors who held multiple directorships. Based on these observations the following hypothesis is developed:

Hypothesis 2: There is an association between the pay-for-performance sensitivity of busy directors on the board of the directors and the level of CEO compensation.

4. Research methodology Sample selection

This paper’s sample is based on U.S. companies who are listed at the Standard & Poor’s (hereafter referred to as “S&P”) 500 from the period 2009-2013. The U.S. market is chosen because of data availability and prior literature on executive compensation mainly originating from the United States (Knop and Mertens 2010). The S&P 500 index is followed among various investors because of its comprehensive view of large companies. The list contains companies from multiple industries and is the index for the 500 biggest market capitalization companies who are listed on the U.S. stock market exchanges. The study required data from multiple databases. Consequently, companies with missing observations after merging the data were excluded from the sample. Furthermore, financial (SIC 6000 - 6999) and public administration (SIC 9000 - 9999) companies, such as mutual funds and government-owned business establishments have been excluded as well due to their components being different from other industries (Richardson et al. 2005) and difficulty of comparability of their financial data (Lin et al. 2010). Lastly, observations have been winsorized in order to exclude the extreme outliers (1 percent) since there is a possibility that the outliers can impact the results that are presented in a normal distribution. The final sample is composed of 1,545 observations derived from 309 unique companies. The following table lists a detailed overview of the sample derivation.

Table 1: Sample and industry composition

Description N

Initial sample from Compustat ISS

Removal because of not being listed on the S&P 500 index Removal due to not having five consecutive years of data Removal of directors not marked as CEO

Removal of missing data and financial companies after merging with data from ExecuComp Removal of outliers

Sample of observations for models of impact on CEO compensation for 2009-2013 The number of unique companies

69,588 -43,039 -3,876 -20,583 -520 -25 1,545 309

Research method

The research methodology is based on a quantitative approach for the purpose of having findings for statistical generalizations. In order to answer the research question, the needed information is obtained from databases and a regression analysis is applied. The data from the financial statement components and the control variables are collected from Compustat North America. This is a database of American and Canadian fundamental and market information on active and inactive publicly traded companies. It provides more than 300 annual and 100 quarterly income statement, balance sheet, cash flow statement, and supplemental data items. For the data on the compensation of executives the S&P Executive Compensation (also known as ExecuComp) is used. This database is a part of Compustat and contains information about the compensation of executives in large U.S. companies. The dataset includes all the S&P 500, S&P Mid-Cap 400 and the S&P Small-Cap 600 companies. Together, these companies constitute approximately 91 percent of the available market capitalization in the United States.

Measurements

In order to examine the hypotheses, the busyness of the directors forms the primary independent variable while CEO compensation forms the dependent variable. Furthermore, several variables are used as control variables in order to not distort the results. Also, natural logarithm transformation is used for some of the variables, such as CEO compensation, current compensation, S&P core earnings, and CEO tenure to obtain a more homogeneous variance of series and to mitigate the influences of outliers (Lütkepohl and Xu 2010).

Busy directors

Researchers such as Core et al (1999) and Fich and Shivdasani (2006) refer to directors who hold multiple board seats as busy directors. While the numeric threshold that constitutes a busy director is subject to discretion, researchers usually consider it to be directors that hold three or more directorships. Likewise, they refer to a busy board as one in which a large number of directors are busy. Given this subjective cut-off point in determining busy directors, in this study the percentage of busy directors is determined by the average number of directorships for all directors of a company. The result of this calculation is indicated by the busyness of the directors.

CEO Compensation

Following the standard definition of executive compensation reported in the ExecuComp database, the value used for the compensation of the CEO is the number that is accounted in the SEC filings.8 Hence, CEO compensation is defined as the sum of the base salary and annual bonuses earned by the executive during the fiscal year, the value of stock and option-related awards that vested during the year as detailed in FAS 123R, the value of amounts earned during the year pursuant to non-equity incentive plans, the change in pension value and non-qualified deferred compensation earnings, and other compensation received by the executive including perquisites, life insurance premiums, discount share purchases etc. The following variables and measurements related to CEO compensation are used:

Table 2: Measurements of compensation components

Variable Measurement in Compustat

CEO compensation Current compensation Option awards

TOTAL_SEC: Total compensation as reported in SEC filings. TOTAL_CURR: Total current compensation (Salary + Bonus). OPTION_AWARDS: The total of the option awards.

Interaction term

Since the second hypothesis predicts a relationship between pay-for-performance sensitivity and director busyness, an interaction term is used concerning performance and busyness. Hence, the variable β11(Interaction term)at page 22 consists of two measures, which are β5(Performance)ln and β1(Busyness). The interaction term is the outcome of the multiplication between the company’s performances and the directors’ busyness. If the coefficient on the interaction term is significant positive it would mean that busyness improves the pay-for-performance sensitivity. However, in case of a negative coefficient, it would imply that the busyness of a director attenuates the pay-for-performance sensitivity.

8

Control variables

A number of control variables such as company size, board size, CEO duality, company performance, year effect, CEO tenure, financial leverage, growth opportunity, and industry classification are used in the investigation. The operationalization of each of these measures is consistent with prior literature and adopted conventional variable definitions in order to ensure comparability with similar studies.

First of all, the level of compensation increases with company size given that a study by Baker et al. (1988) show that a 10 percent larger company will compensate its executives an average of 3 percent more. The researchers indicate that larger companies may employ better qualified and better paid CEOs. In this study the size of a company is measured by the value of its assets. Hence, a natural log of total assets to control for the inherent skewness.

Second, when there are more board members on the board, the compensation of a CEO will be higher as individual directors may focus less on the affairs of the company (Core et al. 1999 and Bebchuk et al. 2006). Hence, in order to control the impacts there is a control variable for the board size of a company.

And third, when a company's CEO is also the chairman of its board, executives have opposing objectives. According to Core et al. (1999) and Brick et al. (2006) this means the CEO will receive a higher compensation. Therefore, a dummy variable is used in order to control the influences of this board characteristic on the dependent variable.

Fourth, theory suggests that the level of compensation is an increasing function of the company’s performance (Core et al. 1999) and hence the effect of it needs to be controlled. The performance is measured by a natural log of S&P Core Earnings9.

And fifth, in order to control the effects of different years, a control variable is used so that the results are not distorted due to the number of years.

Sixth, Hill and Phan (1991) conclude that the tenure of the CEO gives executives time to build influence within a company and thus to tie their compensation more closely to their own preferences. So, as CEO tenure grows the higher the level of CEO compensation becomes. In order to control this effect, a natural log of CEO tenure is used. CEO tenure is measured by the number of years an individual has been the CEO of a given company.

9

The Standard and Poor’s (S&P) measurement of core earnings in order to provide for transparency and consistency, in addition a more stringent definition a company’s core earnings, clearly setting out exactly what can and cannot be considered earnings and expenses.

And seventh, increasing financial leverage raises the chance that companies will fall into financial distress. A company with notably more debt than equity is considered highly leveraged. Madura et al. (1996) argue that leverage should have a positive impact on CEO compensation given that financial leverage raises the risk that CEOs bear personally. Therefore, the effect of financial leverage is controlled and measured by the total debt ratio, the ratio of total liabilities to total assets.

Eighth, Smith and Watts (1992) provide evidence that there are systematic differences in compensation across a company’s growth opportunities. They argue that since it is more difficult to monitor the management’s actions in companies with more growth options, companies are more likely to make use of incentive plans. For this reason, growth opportunity is controlled and measured by the variable market to book ratio. The market to book ratio is the market value of a company divided by the book value of a company.

Lastly, a study by Ely (1991) reports significant differences in compensation practices across industries. Hence, in examining the impact of busy directors on executive compensation in a cross-industry sample, it was considered prudent to explicitly control for such inter-industry variations. The following distinction in industry classification is used and subsequently made operational through the use of six dummy variables:

Table 3: Explanation SIC codes

SIC Codes10 Industry classification

0100 – 1499 1500 – 1799 1799 – 3999 4000 – 4999 5000 – 5999 7000 – 8999

Agriculture, Forestry, and Mining sector. Construction sector.

Manufacturing sector.

Transportation and Public utilities. Trade organizations.

Service organizations.

10

Standard Industrial Classification (SIC) codes are four-digit numerical codes assigned by the U.S. government to business establishments to identify the primary business of the establishment in order to facilitate the collection, presentation and analysis of data.

Empirical models

Empiric models are used for each of the mentioned hypotheses in section three. The first hypothesis assumes that there is an association between busy directors and the compensation of the CEO. The second hypothesis predicts a relationship between pay-for-performance and director busyness. Hence, empirical models one, two, and three are applied for the first hypothesis while the remaining model is used for the latter.

1. Total CEO compensation:

(Compensation)ln = β0 + β1(Busyness) + β2(Company Size)ln + β3(Board Size) + β4(CEO Duality) + β5(Performance)ln + β6(Year) + β7(CEO Tenure)ln + β8(Leverage)+ β9(Growth)+ β10(Industry)+ ε

2. Total current compensation:

(Current)ln = β0 + β1(Busyness) + β2(Company Size)ln + β3(Board Size) + β4(CEO Duality) + β5(Performance)ln + β6(Year) + β7(CEO Tenure)ln + β8(Leverage)+ β9(Growth)+ β10(Industry)+ ε

3. Total long-term compensation:

(Options)ln = β0 + β1(Busyness) + β2(Company Size)ln + β3(Board Size) + β4(CEO Duality) + β5(Performance)ln + β6(Year) + β7(CEO Tenure)ln + β8(Leverage)+ β9(Growth)+ β10(Industry)+ ε

4. Pay-performance sensitivity:

(Compensation)ln = β0 + β1(Busyness) + β2(Company Size)ln + β3(Board Size) + β4(CEO Duality) + β5(Performance)ln + β6(Year) + β7(CEO Tenure)ln + β8(Leverage) + β9(Growth) + β10(Industry)+ β11(Interaction Term)+ ε

Table 4: Variable Definitions Variable Description (Compensation)ln (Current)ln (Options)ln (Busyness) (Company Size)ln (Board Size) (CEO Duality) (Performance)ln (Year) (CEO Tenure)ln (Leverage) (Growth) (Industry) (Interaction Term) Symbol ln Symbol ε Symbol β

Natural logarithm of total CEO compensation as reported in SEC filings in U.S. Dollars ($).

Natural logarithm of current compensation of the CEO, which is the sum of base salary and annual bonus in U.S. Dollars ($).

Natural logarithm of the options awards of the CEO, which is the total of the options awards in U.S. Dollars ($).

A variable indicating the busyness of the directors, which is calculated by the average number of directorships for all directors of a company.

Natural logarithm of the total assets of companies in U.S. Dollars ($) in order to control the effects of company size.

The number of board members indicating the board size of a company.

A dummy variable indicating the results whether the CEO is also the chairman of the board of the company in question (1 = yes; 0 = no).

Natural logarithm of the performance as calculated with the S&P Core Earnings in U.S. Dollars ($).

A dummy variable indicating the results of the mentioned year, which is used as a control variable in order to make a distinction between years.

Natural logarithm of the length of time that a CEO has worked for the company. A variable indicating the ratio of total liabilities to total assets.

A variable indicating the market to book ratio for a company’s growth opportunities. A dummy variable indicating the results of the mentioned industry, which is used as a control variable in order to not distort results.

A variable that represents the outcome of the multiplication between the variables (Performance)ln and (Busyness).

The abbreviation for natural logarithm.

The lowercase Epsilon represents the regression error term. The lowercase Beta represents the regression slope.

5. Results

In this section, the findings of the hypotheses developed in chapter three of the study are presented and explained. The first part describes the statistics of the collected variables. The subsequent part provides evidence of the conducted tests based on the hypotheses.

Descriptive statistics

Table 5 presents the descriptive statistics for the full sample consisting of 1,545 observations. The table states some centrality measures like the mean, some dispersion measures, such as the standard deviation, and the value for the certain percentiles (p25, p50 (median), and p75).

Table 5, Panel A shows that the sample CEOs mean (value of ln) total compensation is $12.2 million (value of 9.2), the current compensation is $1.4 million (value of 7.1), and the option-based compensation is $3.2 million (value of 7.7). The mean busyness of a director amounts to around 1.1 while the average board size is about 11, which is comparable to other researches (Stuart 2014). The average asset-based company size is $26.2 billion (value of 9.5) while the mean liabilities accounts for $15.7 billion. Consequently, this results in a mean leverage of about 59 percent, which indicates that most of the sample companies are relatively strongly financed through debt. Furthermore, the average market value is $27.4 billion with a market-to-book ratio (growth) of around 3.4, which suggests a high availability of growth opportunities to the sample companies. Moreover, the average company earnings based on the S&P Core Earnings is $1.7 billion (value of 6.7). The mean value of the interaction term is near 7.3. The average CEO tenure is 7 years (value of 1.6), which is similar to levels between the years 1998 and 2008 (Larcker and Tayan 2011). Last of all, panel B shows that around 36 percent of the CEOs of the sample are also chairman of the board of directors and almost 50 percent of the sample companies belong to the manufacturing sector. The remaining half is mostly represented by the transportation sector, service companies, and trade organizations with 18 percent, 13 percent, and 11 percent, respectively.

Table 5: Descriptive statistics of the variables used in the analyzes

Panel A. N Mean St. Deviation Minimum Maximum p25 p50 p75

(Compensation)ln 1,545 9.217 .629 6.631 11.474 8.853 9.224 9.602 (Current)ln 1,545 7.071 .661 .148 10.374 6.856 7.026 7.244 (Options)ln 1,545 7.676 .966 1.032 11.415 7.250 7.767 8.230 (Busyness) 1,545 1.074 .417 .000 2.500 .800 1.083 1.364 (Company Size)ln 1,545 9.528 1.091 7.150 12.757 8.696 9.441 10.319 (Board Size) 1,545 10.639 1.937 5.000 20.000 9.000 11.000 12.000 (Performance)ln 1,545 6.660 1.215 1.315 10.655 5.958 6.579 7.357 (CEO Tenure)ln 1,545 1.629 .873 -1.702 3.600 1.099 1.770 2.260 (Leverage) 1,545 .585 .187 .074 1.584 .453 .586 .701 (Growth) 1,545 3.370 10.142 -102.781 156.739 1.662 2.558 3.971 (Interaction Term) 1,545 7.274 3.376 .000 20.240 4.930 6.930 9.310 CEO Compensation 1,545 12,252.223 9,007.074 758.459 96,160.696 6,994.980 10,134.391 14,790.725 Current Compensation 1,545 1,490.282 2,106.822 1.160 32,013.461 950.000 1,125.000 1,400.000 Option Awards 1,545 3,227.225 4,838.757 2.806 90,693.400 1,407.761 2,362.241 3,750.000 Assets 1,545 26,215.476 39,870.706 1,273.984 346,808.000 5,980.100 12,592.000 30,312.500 Liabilities 1,545 15,727.825 24,449.852 175.071 201,365.000 3,187.748 7,197.100 17,714.000 SP Core Earnings 1,545 1,687.190 3,397.627 -5,723.000 42,419.750 371.138 702.733 1,545.825 Market Value 1,545 27,363.659 42,974.645 1,303.404 438,702.000 7,270.339 12,576.305 26,820.733 CEO Tenure 1,545 7.009 5.378 .181 36.583 3.000 5.872 9.586

Table 5: Descriptive statistics of the variables used in the analyzes

Panel B. Frequency Percent

CE

O

Du

alit

y CEO is not Chairman 986 63.8 CEO is also Chairman 559 36.2 Total 1,545 100.0 In d u stry Clas si ficatio

n Agriculture, Forestry, and Mining sector 110 7.1 Construction sector 30 1.9 Manufacturing sector 770 49.8 Transportation and utilities 270 17.5 Trade organizations 170 11.0 Service organizations 195 12.6 Total 1,545 100.0

Table 6: Pearson Correlation Matrix

Variables 1 2 3 4 5 6 7 8 9 10 11 1. (Compensation)ln 1 2. (Current)ln .360 1 3. (Options)ln .674 .188 1 4. (Busyness) .231 .136 .151 1 5. (Company Size)ln .489 .283 .243 .256 1 6. (Board Size) .261 .175 .136 .154 .449 1 7. (Performance)ln .466 .225 .265 .219 .727 .362 1 8. (CEO Tenure)ln .074 -.071 .026 -.063 -.071 -.106 -.022 1 9. (Leverage) .033 .083 -.053 .185 .132 .138 -.024 -.101 1 10. (Growth) -.002 -.013 .000 -.050 -.081 -.041 .002 .038 .017 1 11. (Interaction Term) .380 .204 .221 .900 .519 .275 .597 -.063 .141 -.040 1

Table 6 gives an overview of the Pearson correlations in order to examine the dependence between multiple variables at the same time. Those correlation coefficients that are significant at the 5 percent level are italicized, while bold correlation coefficients are significant at the 10 percent level. The coefficient indicates the strength and direction (±) of the correlation. Cohen (1998) argues that when A > 0.50 Pearson correlation coefficient, it is an indication of a large correlation between variables. Except for company size and company performance, most variables are not very high correlated with total CEO compensation. The interaction term is highly correlated with busyness and company performance since the term is a multiplication of those two variables. Lastly, the table shows a high correlation between company size and company performance.

Main analyzes

Tables 6, 7 and 8 report the regression estimates for the empirical models described by equations (1), (2), and (3). An Ordinary Least Squares (OLS) regression model is used to examine the first hypothesis that predicts that directors’ busyness influences the compensation of CEOs. R-Square is included to indicate the proportion of variance in the dependent variable, which can be explained by the independent variables. Industry and year dummy variables are involved in this test but are not shown in the tables.

Table 6: OLS analysis CEO compensation

Variables Unstandardized B coefficient t-statistic p-value

(Constant) -53.892 -2.699 .007 (Busyness) .174 5.082 .000 (Company Size)ln .163 8.335 .000 (Board Size) .018 2.295 .022 (CEO Duality) .058 1.993 .046 (Performance)ln .106 6.366 .000 (CEO Tenure)ln .078 4.995 .000 (Leverage) -.040 -.517 .605 (Growth) .001 1.062 .288 R-Square .296 Adjusted R-Square .293 F-statistic 64.862 N 1,545

As shown in table 6, the independent variable busyness has a strong positive association with the dependent variable CEO compensation at the 1 percent significance level. This result is consistent with similar research, such as Core et al. (1999) and Fich et al. (2010), which have shown an increase in executive compensation caused by busy directors on the board of companies. Similarly, consistent with prior literature, the control variables company size (Baker et al. 1988), company performance (Core et al. 1999), and CEO tenure (Hill and Phan 1991) have a strong positive relationship with CEO compensation at the 1 percent significance level. Furthermore, the control variables CEO duality and board size have a positive relationship with CEO compensation at the 5 percent significance level. This outcome is consistent with the conclusion of Core et al. (1999) that having more members on the board of directors leads to a higher compensation of the CEO. However, the results also show that there is not enough evidence to support a significant relationship between the remaining control variables leverage and growth and the dependent variable CEO compensation. Lastly, the R-Square shows that the regression line approximates CEO compensation with 29.6 percent.

Table 7: OLS analysis Current compensation

Variables Unstandardized B coefficient t-statistic p-value

(Constant) 9.771 .415 .678 (Busyness) .090 2.229 .025 (Company Size)ln .103 4.463 .000 (Board Size) .018 1.958 .050 (CEO Duality) .006 .169 .865 (Performance)ln .032 1.654 .098 (CEO Tenure)ln -.034 -1.847 .065 (Leverage) .222 2.460 .014 (Growth) .001 .503 .615 R-Square .117 Adjusted R-Square .111 F-statistic 20.287 N 1,545

Table 7 shows that the regression coefficient of busyness is 0.090 (p<0.05), which indicates a significant and positive association with the dependent variable current compensation. However, the control variable company size has a strong positive relationship with current compensation at the 1 percent significance level. This is consistent with prior literature that larger companies compensate their executives more concerning salary and bonus (Baker et al. 1988). In addition, the variable leverage has a positive relationship with current compensation at the 5 percent significance level. Moreover, the control variables board size, performance, and CEO tenure have a weak relationship with current compensation at the 10 percent significance level. The negative coefficient on CEO tenure could be explained by boards being more likely to discipline long-tenured CEOs through cuts in salary and bonuses than through firing. On the other hand, the outcomes show that there is not enough evidence to conclude a significant relationship between the other variables growth, CEO duality, and current compensation. Last of all, the R-Square indicates that the regression line estimates current compensation with 11.7 percent.

Table 8: OLS analysis Option awards

Variables Unstandardized B coefficient t-statistic p-value

(Constant) -58.961 -1.695 .090 (Busyness) .248 4.160 .000 (Company Size)ln .081 2.372 .018 (Board Size) .019 1.400 .162 (CEO Duality) .036 .706 .480 (Performance)ln .119 4.113 .000 (CEO Tenure)ln .041 1.492 .136 (Leverage) -.397 -2.975 .003 (Growth) .001 .566 .572 R-Square .096 Adjusted R-Square .091 F-statistic 16.370 N 1,545

According to table 8 the primary independent variable busyness has a strong positive relationship with the dependent variable options at the 1 percent significance level. Similarly, the control variables performance and leverage also have a strong positive relationship with options at the 1 percent significance level. Furthermore, the control variable company size has a positive association with option-based compensation at the 5 percent significance level. However, the results point out that there is not enough significant relationship between the remaining control variables board size, CEO duality, CEO tenure, growth and the dependent variable options. Lastly, the R-Square shows that the regression line approximates options with 9.6 percent.

The overall conclusion for hypothesis one is that there is a significant positive association between CEO compensation and busyness. Furthermore, it is more significant positive when the compensation is in the form of options than when it consists of current compensation. Additionally, the size and the performance of a company have shown to be significantly and positively related to all three of the dependent variables. The results also indicate that CEO tenure, CEO duality, and board size have a significant relationship with CEO compensation. For current compensation, this is the case with leverage, board size, and CEO tenure respectively. Last of all, the results indicate that leverage has also a significant relationship with options.

Tables 9 reports the regression estimates for the empirical model described by equation (4) in the previous section. An Ordinary Least Squares (OLS) regression is performed to test the second hypothesis that predicts an association between directors’ busyness and pay-for-performance. Industry and year dummy variables are involved but are not shown.

Table 9: OLS analysis CEO compensation – Interaction term

Variables Unstandardized B coefficient t-statistic p-value

(Constant) -52.788 -2.641 .008 (Busyness) .355 2.024 .043 (Performance)ln .138 4.008 .000 (Interaction Term) -.028 -1.055 .292 (CEO Duality) .060 2.049 .041 (CEO Tenure)ln .078 4.979 .000 (Board Size) .018 2.262 .024 (Company Size)ln .165 8.391 .000 (Leverage) -.039 -.513 .608 (Growth) .001 1.072 .284 R-Square .298 Adjusted R-Square .293 F-statistic 59.071 N 1,545

Dependent variable: (Compensation)ln

As presented in table 9, the column p-value indicates that the interaction term has no relationship with the dependent variable CEO compensation at the 10 percent significance level. The variables performance, CEO tenure, and company size have a strong positive relationship with CEO compensation at the 1 percent significance level. Likewise, the variables busyness, CEO duality, and board size have a positive relationship with CEO compensation at the 5 percent significance level. However, the results indicate that there is not enough evidence for a significant relationship between the remaining variables leverage, growth and the dependent variable CEO compensation. Additionally, the R-Square shows that the regression line approximates CEO compensation with 29.8 percent while the adjusted R-Square indicates that all the variables explain 29.3 percent of the variability of CEO compensation.

Overall, the interaction term does not show a relationship between performance and busyness with CEO compensation at the 10 percent significance level. Thus, no moderating effect exists of busyness on pay-for-performance sensitivity. Therefore, the second hypothesis is rejected given that there is not enough evidence for the hypothesis.

6. Conclusion

This section presents the conclusions of the study, provides an overview of the contribution, and gives suggestions for further research.

While executive compensation lingers to command the center stage in the public discourse about corporate governance, this study has investigated that having busy directors does significantly and positively influence the level of executive compensation. Moreover, the relationship is more significant positive when the component of the compensation consists of options than when it contains a combination of salary and annual bonus. Furthermore, several control variables, such as company size, CEO tenure, CEO duality, company performance, board size, and leverage have shown a significant relationship with the compensation variables. This study also examined whether busyness of a director improves or weakens the pay-for-performance sensitivity. The results have shown that the interaction term does not indicate a significant relationship between performance and busyness with CEO compensation. Furthermore, the results of this model have shown that some control variables, for instance, CEO tenure, company size, CEO duality, and board size have a significant relationship with CEO compensation. Summarized, there is enough evidence to validate an association between busy directors and CEO compensation. Therefore, the first hypothesis is accepted. However, there was not enough evidence found for a relationship between busy directors and pay-for-performance. Therefore, the second hypothesis is rejected.

Understanding the role of the board of directors is vital both for our understanding of corporate behavior and with respect to establishing a policy to regulate corporate activities. This study contributes by investigating a characteristic of the board in relation to determining the compensation of executives. Furthermore, this study contributes to limited research concerning the impact of busy directors. The study differs from similar research by primarily focusing on the topic of multiple directorships and by including all directorships in the investigation instead of arbitrary cutoffs. Lastly, the study sought to provide a better understanding of corporate governance and executive compensation by covering an overview of related issues.

Suggestions

The research is subject to limitations. For instance, the results of this study are only focused on the U.S. market and companies that are listed at the S&P 500, essentially only large publicly traded U.S. companies. Furthermore, certain companies such as financials, publicly administrated, and newly listed were excluded for various reasons explained in the methodology section. Hence, future research could be in different countries and/or company sizes. When this relationship is investigated in diverse settings, the results can be compared with each other in order to reach a comprehensive view about the subject.

References

Adams, R., Hermalin, B.E. and Weisbach, M.S. (2008). ‘The role of boards of directors in corporate governance: A conceptual framework and survey’. Working Paper, National Bureau of Economic Research, no. 14486. Andrews, K.R. (1980). ‘Directors’ responsibility for corporate strategy’. Harvard Business Review, vol. 58, no. 6: pp. 30-42.

Armstrong, C.S., Ittner, C.D. and Larcker, D.F. (2012). ‘Corporate governance, compensation consultants, and CEO pay levels’. Review of Accounting Studies, vol. 17, no. 2, pp. 322-351.

Baker, G., Jensen, M. and Murphy, K. (1988). ’Compensation and Incentives: Practice vs. Theory’. The

Journal of Finance, vol. 43, no. 3, pp. 593-616.

Bebchuk, L.A., Fried, J.M. and Walker, D.J. (2002). ‘Managerial power and rent extraction in the design of executive compensation’. University of Chicago Law Review, vol. 69, pp. 751-846.

Bebchuk, L.A. and Fried, J.M. (2003). ‘Executive compensation as an agency problem’. Journal of Economic

Perspectives, vol. 17, no. 1, pp. 71-92.

Bebchuk, L.A. and Fried, J. (2006). ‘Pay Without Performance: The Unfulfilled Promise of Executive Compensation’. Cambridge, MA: Harvard University Press.

Bertrand, M. and Mullainathan, S. (2000). ‘Agents with and without Principal’. The American Economic Review, vol. 90, no. 2, pp. 203-208.

Bertrand, M. and Mullainathan, S. (2001). ‘Are CEOs rewarded for luck? The ones without principals are’. Quarterly Journal of Economics, vol. 116, no. 3, pp. 901-932.

Bhagat, S. and Romano, R. (2009). ‘Reforming Executive Compensation: Focusing and Committing to the Long Term’. Working Paper, Yale Law School.

Booth, J.R. and Deli, D.N. (1996). ‘Factors affecting the number of outside directorships held by CEOs’. Journal

of Financial Economics, vol. 40, no. 1, pp. 81-104.

Boumosleh, A. (2009). ‘Director compensation and the reliability of accounting information’. The Financial

Review, vol. 44, no. 4, pp. 525-539.

Brick, I.E., Palmon, O. and Wald, J.K. (2006). ‘CEO compensation, director compensation, and firm performance: Evidence of cronyism?’. Journal of Corporate Finance, vol. 12, no. 3, pp. 403-423.

Brown, W.O. and Maloney, M.T. (1999). ‘Exit, voice, and the role of corporate directors: Evidence from acquisition performance’. Available at SSRN 160308.

Byrd, D.T. and Mark, S.M. (2005). ‘Bankers on the Board and the Debt Ratio of Firms’. Journal of Corporate

Finance, vol. 11, no. 1-2, pp. 129-173.

Carpenter, M.A. and Westphal, J.D. (2001). ‘The strategic context of external network ties: Examining the impact of director appointments on board involvement in strategic decision making’. Academy of Management

Journal, vol. 44, no. 4, pp. 639-660.

Certo, S.T. (2003). ‘Influencing initial public offering investors with prestige: Signaling with board structures’.

Academy of management review, vol. 28, no. 3, pp. 432-446.

Cohen, J. (1998). ‘Statistical power analysis for the behavioral science (2nd ed.)’. Hillsdale, New Jersey: Lawrence Erlbaum Associates.