Amsterdam Business School
Master Thesis Accountancy & Control
The relation between CFO & director option compensation and real earnings management
Name: Richard Houwen Student number: 10221697
Thesis supervisor: prof. d.r. V.R. (Vincent) O’Connell Date: 12-‐06-‐2016
Word count: 12.015
MSc Accountancy & Control, specialization Accountancy Faculty of Economics and Business, University of Amsterdam
Statement of Originality
This document is written by student Richard Houwen 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
In this study I explore the relationship between CFO and director compensation and real earnings management. More specifically, I focus on stock option and restricted stock compensation. I predict that options will have an effect on the behavior of CFOs and directors regarding earnings management. The results show that CFO stock option and restricted stock compensation are both negatively related to real earnings management. This means that CFOs with more options and restricted stock will engage less in real earnings management. Also, the results indicate that director option compensation is negatively related to real earnings management as well. This is in line with the agency theorist view that options can be used to align interests between executives and shareholders. This study contributes to research on CFO and director compensation.
Key words: option compensation, restricted stock, CFO, director, real earnings management
Acknowledgement
I would like to express my gratitude to my supervisor prof. d.r. V.R. O’Connell for all the useful comments and feedback I received during the process of writing the thesis. Especially, I want to acknowledge d.r. V.R. O’Connell for his positive and flexible attitude, which helped me to finish the largest part of the thesis in two months. Also, I would like to thank my supervisor Hasan Gürkan from KPMG for being able to work at my thesis at KPMG in February and March 2016. Writing this thesis gave me insight in how to scientifically conduct research, which will be useful to me in my future professional career.
Contents
1. Introduction………..6
2.Literature review & hypotheses………. 9
2.1 Agency theory & Managerial power theory………. 9
2.2 The rise of option compensation………. 11
2.3 Option compensation and earnings manipulation……….. 13
2.4 CFO option compensation and earnings manipulation………... 16
2.5 Director option compensation and earnings manipulation……….. 17
2.6 Hypotheses……….. 20
3. Research Methodology………. 23
3.1 Time period & sample...23
3.2 Measuring accrual-‐based earnings management... 24
3.3 Measuring real earnings management... 25
3.4 Empirical models... 27
4. Results... 29
4.1 Descriptive statistics & correlation ... 29
4.2 Regression results... 32
5. Conclusion and discussion... 35
6. References... 37
1. Introduction
“An infectious greed seemed to grip much of our business community. Too many corporate executives sought ways to "harvest" some of those stock market gains. As a result, the highly desirable spread of shareholding and options among business managers perversely created incentives to artificially inflate reported earnings in order to keep stock prices high and rising.” -‐ Alan Greenspan (2002).
According to the abovementioned statement, the modern option incentive system may lead to earnings management. Even though this statement is made almost 15 years ago, the use of options in executive compensation continues to be widespread. In the last decades the use of stock options in executive pay has risen tremendously (Hall & Liebman, 1997). One of the most common justifications is that equity ownership can align interests between CEOs and shareholders (Jensen & Meckling, 1976). However, some studies doubt this agency theorist view. For example, some studies relate stock options to increased accrual-‐based earnings management (Cohen et al., 2008; Cheng & Warfield, 2005; Bergstresser & Philippon, 2006) or even to financial fraud (Burns & Kedia, 2006).
After the introduction of SOX engaging in accrual-‐based earnings management became more costly. This led to a decrease of accrual-‐based earnings management and an increase in real earnings management. Managers seem to trade off both types of earnings management to their relative costliness (Cohen et al., 2008; Graham et al., 2005; Cohen & Zarowin, 2010). The effect of options on accrual-‐based earnings management described in the abovementioned literature, may hold for real earnings management as well. More specifically, real earnings management may be used to artificially inflate reported earnings as well, in order to make executive options more valuable.
The majority of compensation research focuses on CEO compensation. However, some studies believe that the influence of a CFO on earnings management is even larger than the CEOs influence (Jiang et al. 2010). Also, Jiang et al. (2010, p. 515) states: “Future research should consider compensation of CFOs when investigating incentives for earnings management”. Another example of a study regarding CFO and earnings management is the Graham et al. (2005) survey study. They found that earnings are the most important metric considered by investors, even more so than cash flows. Also, because of the immediate negative market reaction when missing a
benchmark, CFOs indicate that they are willing to sacrifice long-‐term gains to increase short-‐term profits. Up to 55% of the CFOs would not start a very positive NPV project if it will lead to not meeting the earnings benchmarks of this quarter.
Besides CEO and CFO compensation, literature regarding board of director compensation is interesting as well, since the whole board of a firm is responsible for the decision making process (Deutsch et al., 2011). Some literature actually studied the link between accrual-‐based earnings management and director option compensation. Boumosleh (2009) for example found a positive link between accrual-‐based earnings management and director option compensation. Furthermore, different studies found a relationship between director option compensation and financial fraud or misstatements (Kim et al., 2013; Persons, 2012; Cullinan et al. 2008). On the other hand, some research supports the usefulness of compensating outside directors with stock options. For example providing options to directors increases the likelihood of CEO turnover after poor performance (Perry, 2000). Another example is that firms who compensate directors with options generate positive cumulative abnormal returns and have significantly higher market-‐to-‐ book ratios (Fich & Shivdasani, 2005). Thus, there is no consensus about whether directors should be compensated with options.
Based on the aforementioned studies, the objectives of this thesis are twofold. First, our work will add to literature regarding the link between CFO stock option compensation and real earnings management. To the best of our knowledge, only Cohen et al. (2008) examines this relation. While Cohen et al. (2008) focuses on executives in general, this study is focused on the CFO, since their influence on earnings management is possibly even larger than the CEOs influence. Furthermore, some adjustments are made to the Cohen et al. (2008) model, e.g. creating a separate variable for restricted stock. Thus, this research aims to answer the call of Jiang et al. (2010) for more research in the CFO and earnings management field. Second, our work aims to add to the literature regarding director compensation, by trying to find a relation between director option compensation and real earnings management. As far as we know, we are the first to document this relationship.
The research question is therefore as follows:
“What is the link between CFO/director stock option compensation and real earnings management in the United States?”
A two-‐tailed hypothesis is used since there is no consensus about the incentive effects of options in the current literature. We predict that the extent of CFO option and restricted stock compensation could be associated with either less or more real earnings management. The same hypothesis holds for director option compensation. However, regarding director compensation, restricted stock will be omitted, because of data availability constraints. Thus, only the effect of director options will be examined. The results indicate that CFOs with more options and restricted stock used less real earnings management, consistent with agency theory. Further, the results also indicate that directors with more options used less real earnings management.
The relationship between CFO option and restricted stock compensation and accrual-‐based earnings management is also examined. Also, the relationship between director option compensation and accrual-‐based earnings management is studied. While this relationship regarding director compensation was still negatively significant, the relationship regarding CFO compensation was insignificant.
This paper contributes to the research about CFO and director compensation. More research into the design of incentives can help to assist remuneration committees in creating compensation packages. By aligning incentives via compensation contracts agency problems can be mitigated. There is a need to gain more knowledge about how to use stock option incentives so that we gain more insight in both the benefits and the drawbacks of compensating executives using options. When stock options provide wrong incentives, a large amount of money is spent inefficiently. Since the stock option compensation market worldwide is huge, gaining more knowledge in this field is of the utmost importance. Besides the usefulness of these results for remuneration committees, the findings from this paper can be interesting as well for accountants, shareholders and labor unions. Suppose directors with no options have a tendency to engage in more earnings management, then accountants need to design the controls of such companies differently. Furthermore, as most shareholders and labor unions strive for long-‐term value creation, firms can use this information as well to demand a different way of compensating CEOs and directors.
The structure of the paper will be as follows. First previous literature and the hypotheses will be discussed. Second the methodology will be explained. Third, the results will be presented. Finally, we present the key conclusions.
2. Literature review & hypotheses
In order to come to develop our hypotheses, the findings and conclusions of previous literature will be described. Also, agency theory and managerial power theory will be discussed. Agency theory will be discussed to show how compensation helps to align interests of managers and shareholders. Managerial power theory will be discussed in order to add to our discussion of agency theory.
2.1 Agency theory & Managerial power theory
Even though research on executive compensation continues to thrive, there is still a lack of consensus regarding the forces shaping executive compensation. On the one hand academics advocate market-‐based explanations, for example agency theory. On the other hand this explanation is challenged by academics highlighting the importance of power. The most prominent recent challenger of market-‐based explanations is managerial power theory (Van Essen et al., 2015). Next, both theories will be explained in more detail.
Agency theory describes that executives of companies are the managers of other people’s money. It cannot be expected that they watch over it with the same vigilance as people would watch over their own money. Further, agency theory assumes that both executives and shareholders try to maximize their utility. The main interest of shareholders is a higher stock price and executives strive to increase their compensation. This can lead to a conflict of interest problem. For example, a CEO makes an acquisition and therefore his pay increases because the revenue of the firm increases. Shareholders may wonder if the CEO really acted in the best interest of the firm, or that he mainly made the acquisition to increase his paycheck. Thus a CEO may take an action in his best interest even though this hurts the firm (Jensen & Meckling, 1976). This is the so-‐called agency problem. Stock option or share ownership can potentially overcome this problem, by making CEO income dependent on the share price. This will align the interests between managers and shareholders (Haugen & Senbet, 1981; Agrawal & Mandelker, 1987). Also, stock options counterbalance a CEOs natural tendency toward risk aversion, which originates from striving to avoid personal losses, e.g. reputation loss or job loss (Milgrom & Roberts, 1992).
Managerial power theory (MPT) contrasts with agency theory. However, the goal of MPT is not to prove that agency theory is wrong, but to extend it by arguing that managerial power can cast some doubt on the assumption of optimal contracting. MPT states that executives use their power
to increase their income (Bugeja et al., 2012). This will be especially visible when a CEO is present on the supervisory board. This may lead to an environment of mutual ‘back scratching’ between the chairman and the CEO. Bebchuk & Fried (2003) and Crystal (1991) provide several reasons that can explain the mutual ‘back scratching’. First, directors are essentially hired by the CEO and can also be fired by the CEO, so board members may be unwilling to express a conservative opinion regarding CEO compensation. Second, social mechanisms e.g. friendship, collegiality are frequently found in boards, which can lower the objectivity of directors. Finally, CEOs are in the position to reward directors via larger compensation.
Evidence regarding mutual ‘back scratching’ is also found in the paper of Oxelheim & Clarkson (2015). They studied the determinants of the compensation of the supervisory board chairman, and specifically the relationship between chairman and CEO compensation. The compensation of a CEO is set before the compensation of the supervisory board chairman, thus they hypothesize that the chairman may conspire with a CEO in order to increase his income. Their hypothesis is confirmed, because they found evidence that the gap of compensation between the chairman of the supervisory board and CEO is less in firms where the chairman has first served in the executive team. A chairman may engage in mutual back scratching since it increases his income as well.
Besides the just discussed studies regarding the MPT, a lot of other research extends our knowledge of theory of managerial power. However, this literature is characterized by conflicting findings that lead to evidence supporting different claims. For example, some studies find evidence that CEO duality is related to larger compensation (e.g., Yermack, 1996), while other studies find a non-‐relation between CEO duality and compensation (e.g., Boyd, 1994). Van Essen et al. (2015) assesses this literature, namely 219 studies, with the help of meta-‐analytic methods. First, the study focuses on the relationship between managerial power and CEO compensation. Second, the study examines the relationship between managerial power and CEO pay-‐ performance sensitivity. They find overall support for MPT regarding the relationship between managerial power and total compensation levels. Two of three indicators of CEO power (board size and CEO duality) have the predicted positive relation with total compensation, meaning that most of the times when a CEO has more power, there will be a higher compensation package. Another indicator of CEO power (CEO tenure) was insignificant. Further two of three indicators of board power (ownership concentration and institutional ownership) are negatively associated with total pay, meaning that more board power leads to lower compensation, consistent with
their hypothesis. However, another indicator of board power (board independence), led to higher compensation, inconsistent with their hypothesis.
Next, Van Essen et al. (2015) studied the relationship between managerial power and CEO pay-‐ performance sensitivity. In this case less evidence is found to support MPT regarding the performance-‐pay sensitivities. Only one indicator of CEO power (CEO tenure) is significant in the predicted negative direction, meaning that CEOs with a longer tenure will have a lower performance-‐pay sensitivity. Regarding board power two of the three indicators (board independence and institutional ownership) are positively significant, consistent with their hypothesis. This means that the higher the board independence and institutional ownership, the higher the performance-‐pay sensitivity.
In conclusion, MPT is useful to predict core compensation variables such as total compensation, but less equipped to predict the sensitivity of pay to performance. In most situations when CEOs have more power, they receive more total compensation. Also, when a board has more power, CEOs receive less total compensation. Furthermore, more powerful directors seem to increase the CEO performance-‐pay sensitivity. The meta-‐analysis shows that MPT is a useful expansion of agency theory. Now that some theoretical background on compensation is provided, we will provide additional information on the history of option compensation.
2.2 The rise of option compensation
Before describing the history of option compensation, some details about the different kinds of options can be useful. Options can be divided into e.g. unexercised exercisable options, unexercised unexercisable options, newly awarded options, in-‐the-‐money options and out-‐of-‐the-‐ money options. Restricted stock can be considered as an option as well. To be more specific, restricted stock can be considered as an unexercised unexercisable option with a zero exercise price (Kadan & Yang, 2005). Each option may have different incentive effects. For example, managers will have an incentive to lower the stock price around the stock option award date (Cohen et al., 2008). On the other hand, unexercised unexercisable and unexercised exercisable options may give managers an incentive to opportunistically manage earnings upward. At least, this may be the case for unexercised unexercisable options that will become exercisable on the short term. However, another incentive effect is also possible regarding unexercised unexercisable options that will become exercisable on the long term. Unexercised unexercisable options may
give managers less incentives to use real earnings management, since this burns real cash flows, and can lead to less long term value creation. This study will focus on unexercised unexercisable options, unexercised exercisable options and restricted stock. Also, this study will focus on in-‐the-‐ money options, due to data availability constraints. Also, no distinction will be made between options that will be become exercisable on the short term or the long term, again due to a lack of data availability. Next, the history of option compensation is described.
Jensen & Murphy (1990) engaged in a study to link pay and performance in the 1969-‐1983 time frame. Their main conclusion was that CEO wealth raises $3.25 per every $1.000 increase in firm value and they believed this amount to be immaterial. Thus, they found a non-‐relation between pay and performance. In addition, Rosen (1990) also found almost no correlation between performance and pay. They concluded that the elasticity of top executive pay with respect to stock market returns lay around 0.1. This means that for example a CEO whose company gets a 20% return on stock would get paid 1% more compared to a CEO who earned a 10% return on stock. This amount can be considered immaterial as well.
However, between 1980 and 1994, an increasing amount of firms started using options to incentivize their CEOs. The number rose from 30% to 70% (Hall & Liebman, 1997). The Hall & Liebman (1997) study recreated the Jensen and Murphy paper in this time frame. They found that the median elasticity of CEO compensation to market value of a firm rose from 1.2 in 1980 to 3.9 in 1994. Compared to Rosen (1990) the 1994 elasticity is 39 times higher. This means that the CEO whose company gets a 20% return on stock would get paid 39% more compared to a CEO who earned a 10% return on stock. The boom of stock option compensation led to a really significant pay for performance sensitivity.
The Hall & Liebman (1997) paper states that the most direct solution to agency problems is to align incentives of CEOs and shareholders by providing derivatives such as stock options. The years after the paper, this indeed happened on a large scale, and in 2001, stock options accounted for 50% of the pay of CEOs of large firms in the United States (Sanders & Hambrick, 2007).
According to agency theory, stock options can overcome three problems that arise when you just use a base salary (Haugen & Senbet, 1981; Sanders & Hambrick, 2007): shirking, shortsightedness, and risk aversion. First, shirking will be less because of aligning CEO payoffs with shareholders
payoffs. Second, stock options have a vesting period before they can be exercised. Thus CEOs may have a long-‐term view when making their investments. Third, firms can solve risk aversion, because with stock options, risky projects that turn out well can make a CEO very wealthy. However, the last two decades, it became clear that the usage stock options might also have some downsides, such as manipulation of earnings. Manipulation of earnings will be split into accrual-‐ based earnings management and real earnings management.
2.3 Option compensation and earnings manipulation
Before discussing the literature about option compensation and earnings manipulation, the introduction of SOX will be discussed, because this led to shift in the way of managing earnings. Since the introduction of Sarbanes-‐Oxley Act (SOX) in 2002, the financial reporting environment changed. For example, Section 201 of SOX forbids outside auditors to provide several non-‐audit services to their audit clients (Bartov & Cohen, 2009). Also, other non-‐audit services, such as tax advisory, should be approved in advance by the audit committee. This can lead to more auditor independence and higher audit quality. Another part of SOX, Section 404, states that companies must provide assessments about the effectiveness of their internal controls. Furthermore, due to Section 302 and Section 906, CEOs and CFOs must state under oath that their annual and quarterly financial reports are correct. They can receive significant penalties for false certification, such as financial penalties or imprisonment. The aim of the measures is to decrease fraudulent financial reporting. Those measures made engaging in accrual-‐based earnings management more costly. Therefore after the introduction of SOX accrual-‐based earnings management has become less prevalent (Cohen et al., 2008; Graham et al., 2005; Cohen & Zarowin, 2010).
Regarding real earnings management, a different effect post-‐SOX can be expected. Real earnings management is harder to detect than accrual-‐based earnings management, since it is hard to distinguish from optimal business decisions (Graham et al. 2005; Cohen et al. 2008). Therefore, post-‐SOX, when accrual-‐based earnings management becomes more risky, managers may prefer real earnings management.
Also, Zang (2011) found significant positive results between the usage of real earnings manipulation activities and the costs associated with accrual-‐based earnings management. This supports the hypothesis that managers trade off both types of earnings management to their relative costliness. The costs of real earnings management however can be significant, since
managers may be willing to burn ‘real’ cash flows to reach desired numbers, instead of just manipulating accounting numbers.
There are multiple ways to engage in real earnings management. Roychowdhury (2006) distinguishes three ways to manage earnings in a real way. Firstly, through abnormal levels of cash flows from operations, for example by providing discounts to temporarily increase sales. Secondly, through manipulating expenses, for example cutting back on R&D, advertising, SG&A expenses. Lastly, through production costs, for example engaging in overproduction to lower costs of goods sold.
Even though real earnings management has become more prevalent in the last decade, previous literature studied stock options mostly in relation to accrual-‐based earnings management. Cheng & Warfield (2005), for example, find that CEOs with high equity incentives, are more likely to report earnings that meet or just beat analysts’ forecasts. They find that CEOs with a lot of stock options, report on average more income-‐increasing abnormal accruals, which indicates accrual-‐ based earnings management. Lastly, they also find that CEOs sell more shares after income-‐ increasing abnormal accruals, possibly indicating a short-‐term focus.
Bergstresser & Philippon (2006) found results similar to the Cheng & Warfield (2005) paper. Namely, they found evidence that when CEO income is more closely tied to the value of stock and option holdings, more accrual-‐based earnings management is used. These CEOs seem to use the discretionary components of earnings more in order to inflate earnings. Furthermore, in years with high accruals, CEOs exercise large amounts of options. This is again a possible indicator of a relation between earnings management and options. Bergstresser & Philippon (2006), also provide examples of companies where inflated earnings coincided with significant option exercises and the selling of shares. For example Xerox, which is a company where executives manipulated earnings in the 1990s. In 2002 Xerox was sued by the SEC and had to restate their earnings between the period 1997 and 2001. Due to this restatement, net income was reduced by $1.4 billion. During this period, the total value of options exercised was over $20 million, which was almost three times as much as the value of options exercised over the prior five years.
Another well-‐known example of excessive stock option compensation and earnings manipulation is Enron. Hall & Murphy (2003) state that the problem of stock options is that they are granted to
too many people. Stock options can be linked to excessive risk taking and excessive focus on stock prices and are thought to be one of the main causes of the Enron debacle.
Donoher et al. (2007) also raises concerns about the efficacy of incentive alignment via options and states that the problem of misleading disclosures did not end after the collapse of for example Enron or after the introduction of SOX. They hypothesize that there is a positive relationship between CEO equity ownership or options and the incidence of misleading disclosures, which is confirmed by their results. These results are not in line with agency theory, since agency theory implies that options will lead to incentive alignment. Donoher et al. (2007) interpret their results with the help of behavioral agency principles and state that ownership and options both alter executive behavior, albeit in a different way. Ownership will expose the executive to downside risk, whereas option compensation has no downside risk and gives executives a chance to realize costless gains. A situation sometimes referred to as ‘head I win, tails you lose’ (Sanders, 2001). Lastly, they also examined whether powerful boards were better able to deter misleading disclosures. This was confirmed since misleading disclosures occurred less in firms where boards had a high level of business experience and long tenure.
Another study examining the use of stock options and misreporting is Burns & Kedia (2006). They examine how management incentives influence the likelihood of restatements of financial statements. Accrual-‐based earnings management can affect stock prices, thus managers with compensation linked to the share price may have an incentive to artificially inflate earnings. CEO wealth increases when the stock prices increases due to aggressive accounting. However, the loss of CEO wealth when the share prices declines is limited. Burns & Kedia (2006) find that the higher the sensitivity of CEO income to the stock price, the higher the tendency to misreport. Stock option incentives stimulate aggressive accounting practices.
The abovementioned studies all doubt the agency theorist view that stock options can align interests between executives and shareholders. They found that stock options are positively related to accrual-‐based earnings management or misreporting. However, the Cohen et al. (2008) paper, also found some other effects. The Cohen et al. (2008) paper studies the link between SOX and earnings management. Also, they examine the link between earnings management and stock options. Both accrual-‐based earnings management and real earnings management are examined. One of their variables is stock options. Options can both have a downward and upward incentive
effect with respect to stock prices. Around the stock option award date, managers prefer a lower stock price, since that will decrease the exercise price of the stock option. However, for unexercised options that become exercisable in the short-‐term managers have an incentive to increase the stock price in the short-‐term. So managers with more unexercised options are more sensitive to short-‐term stock prices, and can use their discretion to affect reported earnings. Their results suggest that unexercised options are indeed positively associated with income increasing accrual-‐based earnings management. However, regarding real earnings management, they found that more option compensation actually led to less real earnings management. This may be explained by agency theory. According to agency theory, options align incentives between executives and shareholders (Jensen & Meckling, 1976). Real earnings management can be considered as costly, thus when incentives between executives and shareholders are better aligned, options actually decrease the amount of real earnings management.
Most of the abovementioned studies focused on CEO compensation, therefore one can conclude that there seems to be an either positive or negative link between CEO stock option compensation and earnings management. However, not only the CEO is responsible for decision-‐making and earnings management. The CFO and board of directors also have significant influence, which will be discussed next.
2.4 CFO option compensation and earnings manipulation
Prior research focused mainly on how CEO equity incentives are related to earnings management. This is likely because CEO compensation packages are much higher compared to the CFOs and therefore believed to be the most influential (Jiang et al., 2010). However, the CFO does have significant influence over the financial reporting area. Geiger & North (2006) for example examine the changes in discretionary accruals after the appointment of a new CFO. They found a significant reduction in the use of discretionary accruals after a new CFO joined the firm, compared to a control group of firms that did not hire a new CFO. Also, the changes were not driven by the appointment of a new CEO in the same period. Thus, Geiger & North (2006) provide empirical evidence that a CFO has significant influence over the firm’s reported financial results.
The survey study of Graham et al. (2005) is examining CFOs and earnings management. They interviewed 401 CFOs and determined the key factors that drive decisions related to financial reporting. Earnings are the key metric considered by investors, even more so than cash flows,
according to the CFOs. Due the immediate negative market reaction when missing a target, CFOs are willing to sacrifice long-‐term value in order to meet benchmarks. Meeting or exceeding benchmarks is very important and CFOs describe a trade-‐off between delivering the expected earnings on the short-‐term and long-‐term value creation. CFOs may use their significant influence over the firm’s financial report to manage earnings. Also, Graham et al. (2005) finds that CFOs prefer to take actions that have negative consequences in the long term than make within-‐GAAP accounting choices to manage earnings. Up to 78% of the CFOs admits to sacrifice long-‐term gains in order to make earnings less volatile. Also, 55% of the CFOs would not start a very positive NPV project if it meant falling short on the earnings expectations of this quarter.
Based on the abovementioned studies one can conclude that a CFO uses his significant influence over the financial reporting of a firm frequently to manage earnings. In addition, studies examined the effects of CFO compensation and earnings management. Jiang et al. (2010) states that increasingly compensating CFOs with stock options will stimulate CFOs to boost short-‐term stock prices at the expense of long-‐term gains. They found that the influence of a CFO on earnings management is even larger than the CEOs influence. Namely, accrual-‐based earnings management is more increasing in CFO equity incentives than CEO equity incentives.
Katz (2006) expresses the concerns of Internal Revenue Service Commissioner Mark Everson at a senate hearing in 2006, who states that CFOs should not be paid in options. Everson was worried about the temptations that go along with compensating CFOs with stock options. CFOs may need to possess ‘heroic’ virtue to keep them from wrongdoing. CFOs who are in charge of ‘minding the cookie jar’ should not be paid in options. Concluding, compensating CFOs with options may have some adverse effects. On the other hand, according to agency theory, compensating CFOs with options can align interests between CFOs and shareholders. This is the same argument that is used regarding CEO compensation. Next, earnings management and director compensation will be discussed.
2.5. Director option compensation and earnings manipulation
Similar to CFO compensation literature, director compensation literature is relatively scarce (Deutsch et al., 2011). Also, there is no consensus in the current literature about the effects of director option compensation on earnings management. Before discussing the relationship between director option compensation and earnings management, a definition of the different
categories of directors can be useful (Cullinan et al., 2008). Directors can be split between inside, outside/independent and grey directors. Inside directors are employed full time by the firm, e.g. as CEO or CFO. Outside directors, also known as independent directors, are not full time employees of the firm and their role is to monitor the executives of the firm. Grey directors have certain relationships with the firm in addition to their director role, e.g. providing consulting services to the firm.
There are two possible effects of option compensation on earnings management. Firstly, more option compensation can lower the amount of earnings management. According to agency theory, directors are self-‐interested agents, and their incentives need to be aligned with the interests of shareholders (Jensen & Meckling, 1976). Stock options could mitigate the possible misalignment of incentives, since directors become dependent on the stock price and thus more involved in various board tasks. So, by aligning their interests with shareholders, directors may become more inclined to disclose relevant information to investors (Boumosleh et al., 2012). Also, providing options to directors may attract more qualified directors (Boumosleh, 2009). Other research supporting the usefulness of providing stock options to outside directors include e.g. Perry (2000) and Fich & Shivdasani (2005). Perry (2000) states that compensating directors with options will increase the likelihood of CEO turnover after a poor performance. Also, director options and stock may provide directors with more incentives to monitor managers.
Fich & Shivdasani (2005) examines whether the adoption of director stock option plans does affect shareholder wealth. They find that providing stock options to outside directors generate positive cumulative abnormal returns (CARs) as well as favorable revisions in earnings forecasts. Also they find that firms that provide outside directors with stock options have significantly higher market-‐ to-‐book ratios.
Even though providing options to directors could be beneficial according to agency theory and some of the earlier mentioned studies, several studies provided results that doubt this agency theorist view. Those studies found director option compensation to be related to accrual-‐based earnings management, opportunistic option grant timing, earnings misstatements, SEC violations and even financial fraud.
For example Boumosleh (2009) argues that the extent to which outside directors perform their monitoring role depends on the incentives they receive. He found that option compensation is positively related to accrual-‐based earnings management. These results doubt the statement that agency problems between directors and shareholders are actually reduced when compensating directors with option contracts. Stock option rich contracts for outside directors may incentivize them to allow more favorable reporting to avoid for example the violation of a debt contract.
Bebchuk et al. (2010) and Byard & Compton (2005) suggest that outside director stock option compensation could inflate their independence. They examine opportunistic option grant timing practices. The occurrence of opportunistically timed options to executives has been extensively described in previous literature, whereas Bebchuk et al. (2010) and Byard & Compton (2005) describe this phenomena regarding outside director option compensation. Bebchuk et al. (2010) examines at-‐the-‐money, unscheduled options awarded to both CEOs and outside directors. The amount of lucky grants is examined, which is defined as a grant given at the lowest price of the month. The results showed that 15% of the grants to CEOs were lucky, whereas 11% of the grants to outside directors were lucky. Those lucky grants were caused by deliberate choices and were aimed to make the options more profitable.
Also, Cullinan et al. (2008) examines whether outside directors who receive no options are more effective in detecting earnings misstatements, compared to outside directors which receive options. They examined 105 firms that misstated revenues and 105 firms that did not misstate revenues. Cullinan et al. (2008) finds that when the directors receive no options, there is a reduced chance of financial misstatement. The problem with providing stock options to directors may be that they create a mutuality of interest between executives and directors, rather than between shareholders and directors. Concluding, outside directors who receive no options may perform their tasks better than directors who do receive option compensation.
Kim et al. (2013) studies both the composition and the compensation of the board of directors in relation to financial fraud. Regarding the composition of the board they find that SEC violations occur less when the board consists of more women, financial experts, independent members. Also less fraud occurred when a CEO does not serve as chairman as well. Shorter director tenure, which may increase their independence also led to less SEC violations. Regarding compensation, stock and especially options have a positive relation with financial fraud.
Another example is Persons (2012), who examines 111 fraudulent companies and 111 non-‐ fraudulent companies. He found a positive relationship between director stock option compensation and the likelihood of fraud. Interestingly, there was no relationship between director stock ownership or director cash and the likelihood of fraud. This indicates that companies should choose director cash and stock compensation over option compensation.
The findings of those papers can be explained by the fact that disclosing earnings restatements usually means a decline in share price (Palmrose et al. 2004), so Hamdani & Kraakman (2007) argue that outside directors with stock options have an incentive to overlook wrongdoing, because their income is dependent on the share price as well. In other words, Kim et al. (2013) states that while performance-‐based compensation can align director and shareholder profit motives, outside directors also play a key role in monitoring management and preventing wrongdoing. Outside directors may have difficulties performing this role of monitoring, when option compensation is compromising their independence.
2.6 Hypotheses
Option compensation was introduced to align incentives between executives and shareholders. However, option compensation may lead to unintentional effects, which may be more severe the higher the amount of option compensation. For example, option compensation can lead to earnings management or financial fraud (Cohen et al., 2008; Cheng & Warfield, 2005; Bergstresser & Philippon, 2006; Donoher et al., 2007; Burns & Kedia, 2006). Most studies focus on CEO compensation, however not only the CEO is responsible for decision-‐making and earnings management, but other executives and directors are also responsible. Interestingly, Jiang et al. (2010) found that the influence of a CFO on earnings management is even larger than the CEOs influence.
All the above-‐mentioned studies regarding earnings management focus on accrual-‐based earnings management, except Cohen et al. (2008). After the introduction of SOX firms partly substituted accrual-‐based earnings management to real earnings management. The Cohen et al. (2008) paper studied the relation between both accrual-‐based earnings management and real earnings management and executive stock option compensation. In this study, we extend the Cohen et al. (2008) paper to the CFO and director field. To the best of my knowledge, no previous literature studied the link between CFO or director option compensation and real earnings management.
