Executive compensation : CFO compensation as a result of mergers and acquisition deals : a comparison between smaller firms acquiring bigger firms and bigger firms acquiring smaller firms : quantitative research among

Hele tekst

(1)Executive compensation: CFO compensation as a result of mergers and acquisition deals. A comparison between smaller firms acquiring bigger firms and bigger firms acquiring smaller firms. Quantitative research among large public companies. Name:. Sivan Taha. Student number:. 10596275. Date:. 12-08-2016. Thesis supervisor:. dr. A. (Alexandros) Sikalidis. MSc Accountancy & Control Faculty of Economics and Business.

(2) Statement of Originality This document is written by student Sivan Taha 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.. 2.

(3) Abstract. This thesis has examined the relationship between M&A deals and executive compensation. We narrowed down our research question and investigated whether CFO pay is higher when small firms acquire large firms than when large firms acquire small firms. The logic behind it is that CFOs may be rewarded more for more difficult M&A deals, namely when a small firm manages to take over a large firm. We tested this and other hypotheses using data from the Executive Compensation database and the Thompson Mergers and Acquisitions database from 2006 to 2015. Our results from ordinary linear regression models show that deal size has a positive and significant effect on CFO compensation. Furthermore, we found strong support for our main hypothesis that total reported compensation for CFOs was higher for small firms that take over large firms than for large firms that take over small firms. These findings were robust across models, regardless of the measure of compensation used. The occurrence of small firms taking over large firms is a rare and interesting economic event and this thesis contributes to the limited research done on this specific topic.. 3.

(4) INTRODUCTION. For many years the extent to which executives are rewarded has been a topic of discussion in the public media and the industry. These discussions usually focus on the exorbitant amount managers are rewarded. The economic crisis has shed light on this subject and public outrage has pressured stakeholders to take more measures to curb these excesses. It is hard to justify huge amounts of bonuses for executives when employees are being laid off at the same time or when these public companies are saved by taxpayers’ money. Despite the increasing taxes on bonuses, executive pay has been increasing over time. One case where executives are royally rewarded is with merger and acquisition (M&A) deals. This holds for top managers of the target company, but also the top managers of the acquiring company can be significantly rewarded for their effort. These rewards are often in cash or in shares. This raises the question to what extent managers act in their own interest and to what extent they have the best interest of their shareholders at heart. There are numerous reasons for mergers and acquisitions to take place. These motivations include achieving synergy effects, accelerating growth, expansion of existing production and, of course, increasing shareholder value. By taking over the competitor, a company can gain a larger and broader customer base. If a company decides to acquire its supplier, the supply chain can be shortened and the margin of profit can be increased. There are three forms of synergy effects: technical scale advantage, diversification and financial advantages of scale, which can all be motivations for M&A (Lubatkin, 1983). In his paper, Lubatkin points out the relationship between mergers and business performance after the merger. Firms need to think about whether the scale advantages are going to be met when the M&A is actually realized and whether this will translate into improved business performance. Several studies suggest that this is not always the case. Weston & Mansinghka (1971) conclude that conglomeration strategies can lead to average business performance but certainly not better. They argue that random portfolio outperforms equities conglomerate portfolios. But still, the number of M&A deals has only gone up (Lubatkin, 1983). This makes us wonder why do firms pursue this strategy, if it does not necessarily lead to improved business performance. If we look at M&A deals from the point of views of executives, then we need to see if managers are indeed getting enormously richer from a successful M&A transaction. A useful theory, related to this topic, is principal-agency theory. Does the executive (agent) really work in the interest of the shareholder (principal), or in the interests of themselves? The research in this paper focuses on the extent to which the executive pay is related to M&A 4.

(5) deal. We focus on public companies that have been active in mergers and acquisitions. There is a wealth of good research on the topic of M&A and executive compensation. Grinstein & Hribar (2004) showed that executive compensation is indeed positively related to the effort and time invested by a CEO in the overall acquisition and their final award. The size of the CEO compensation strongly depends on the size of the deal (Bliss & Rosen, 2001). Here they measured the effects of M&A on CEO compensation and came to the conclusion that there is a positive correlation between them. They find that even if the acquisition is ultimately bad for the stock price, the CEO bonus is still a high amount. The topic of the research in this paper is not unique. The relationship between M&A and executive compensation has been previously studied (Bliss & Rosen, 2001). These studies, however, focus mainly on CEO compensation as the dependent variable and choose, in the case of Bliss & Rosen, a specific sector, banking. This study is particularly unique in that it addresses the compensation of the Chief Financial Officer (CFO). The motivation behind this is that not only the CEO has invested a lot of time and effort to prepare a deal. The CFO, who has the responsibility for the due diligence, is working overtime during the M&A process. In this particular research, we focus on CFOs from small firms that acquire M&A deals bigger than their own firm valuation. We expect that because of the difficult task of financing a deal bigger than their own firm, their compensation might be higher. The remainder of this thesis is structured as follows. Section 2 presents the theoretical framework. It gives an overview of the debate in the literature concerning executive compensation and mergers and acquisitions. Section 3 formulates the main research question and the hypotheses. Section 4 describes methods of data gathering and section 5 presents the data and the results of the quantitative analysis. Finally, we will conclude our findings in section 6 and propose avenues for future research.. 5.

(6) TABLE OF CONTENT. INTRODUCTION ........................................................................................................................ 4 TABLE OF CONTENT ............................................................................................................... 6 2.. 3. THEORETICAL FRAMEWORK AND LITERATURE REVIEW ................................ 7 2.1. Mergers and Acquisitions .............................................................................................. 7. 2.2. Executive compensation ................................................................................................ 8. 2.3. Impact of M&A on executive compensation ............................................................... 10. 2.4. Impact of M&A deals on executive compensation of acquiring firm ......................... 13. HYPOTHESIS .................................................................................................................... 15 3.1 Research question ............................................................................................................. 15 3.2 M&A and CFO compensation .......................................................................................... 15. 4.. METHOD AND DATA GATHERING .......................................................................... 17 4.1. Conceptual model ........................................................................................................ 17. 4.2. Dependent variables .................................................................................................... 18. 4.3. Independent variable: .................................................................................................. 18. 4.4. Control variables.......................................................................................................... 18. 4.5. Regression analysis...................................................................................................... 21. 4.6. Data sources ................................................................................................................. 21. 5.. 6.. DATA .............................................................................................................................. 22 5.1. Dependent variable ...................................................................................................... 22. 5.2. Independent variables .................................................................................................. 24. 5.3. Control variables.......................................................................................................... 26. 5.4. Bivariate analysis ......................................................................................................... 29. 5.5. Multivariate regression analysis .................................................................................. 30. 5.6. Robustness tests ........................................................................................................... 36 CONCLUSION ............................................................................................................... 38. References .................................................................................................................................. 41. 6.

(7) 2.. THEORETICAL FRAMEWORK AND LITERATURE REVIEW. This thesis aims to explore what impact mergers and acquisitions (hereafter M&A) have on CFO compensation. The literature relating to this topic can be divided into two main themes: on the one hand, the M&A literature and on the other hand, the literature on executive compensation. A number of papers also take both subjects into account. The following section describes M&A, executive compensation, and the relationship between the two. Globally, M&A has reached an extraordinary level in recent years (Barkema & Schijven, 2008). Most of the research done is extensively related to acquisitions, these researches lack integration of multiple facets to this topic.. 2.1. Mergers and Acquisitions. According to Devers & Mcnamara (2008), M&A deals can be driven by four goals: managerial self-interest (value destruction), firm characteristics, value creation (market power, efficiency, resource redeployment, market discipline) and environmental factors. Andrade et. al. (2001) give economic examples for mergers as efficiency-related, economies of scale or synergies; pursuit in creating market power, maybe by establishing monopolies or oligopolies. They indicate that mergers creates shareholder value, but with most of the gains going to the target firm it is not that evident that value creation matters to the acquiring firm. Their study relies on short stock market reactions which measures the value creation of the merger. Stock markets immediately react to M&A announcements, absorbing any future value change. The full value effect should be fused into the stock price when ambiguity surrounding the M&A is concluded, especially, when the M&A deal is completed. Their research supports results from past research of Jensen & Ruback (1983) and Jerrel at. al. (1988) that M&A creates shareholder value, but regarding the period of the announcement, the gains are entirely for the shareholders of the target firm. There are also different effects regarding M&A financed with or without stock, from the perspective of the acquiring firm there are two separate transactions, an equity issue and a merger. Myers & Majluf (1984) focus on the information contrast between investors and managers. They find that managers are more likely to issue equity when they find the price to be overvalued. The adverse market reaction in the announcement period is smaller for the acquiring firm when the merger is financed with issued stocks. A prevailing debate proclaims 7.

(8) that managers finance M&A deals with cash when they perceive their companies to be undervalued and with stocks when they find their respected firms to be overvalued (King et al. 2004). Loughran & Vijh (1997) argue that the market should perceive stock financed deals as a signal of bidder overvaluation. Multiple studies have shown that cash financed deals are more favorable to the acquiring firm (Loughran & Vijh 1997; Carow et al., 2004). Yet, this is not that evident as the researchers suggest. Fuller et. al. (2002), for example, find very different results. They show that regardless of payment type, M&A deals resulted in returns for the acquirer in case of cash and combination deals and significantly negative returns for stock deals. Interestingly, subsidiary and private acquisitions increased acquirers’ performance, regardless of the method used, and those returns were more significant when the deal was financed with stock. Examining the post-merger performance, Healy et. al. (1992) conclude that merged companies have higher asset productivity. After the merger, these companies increased their operating performance parallel to their industry peers. This study does have data limitations, which make it difficult to generalize the findings.. 2.2. Executive compensation. A well-known theory that describes conflicts of interests between the executives and the shareholder is the agency theory. Jensen and Meckling (1976) briefly summarize the core of the theory: when the owner (shareholder) and manager (executives) are different entities, they will have different interests. The principal (shareholder) wants to maximize its wealth, while the agent (manager) mainly looks at its own interest. Effective corporate governance mechanisms can relieve these agency problems by encouraging good conduct by the management. This can be achieved through incentivized compensation and monitoring by the board of directors (Agrawal & Knoeber, 1996). By binding managerial pay to company performance, opportunism by managers could be reduced and managers could be incentivized to act in shareholders’ best interest. Yet, this is not always the case. Some studies show that managerial stock ownership and long term incentive plans do have positive returns at acquisition announcement (Travlos & Waegelein, 1992). Bebchuk & Fried (2004) try to explain the political perspective by observing the compensation scheme through the power lens. They point out that managers gain extensive control over compensation setting process, by applying too much power over their boards. Board members might be in debt for being nominated to the board which might create incentives for them to make decisions that are favorable for executives. Managerial power over the board drives large cash bonuses with respect to acquisitiveness (Grinstein & Hribar, 2004). As a result, it is unclear whether M&A 8.

(9) deals are determined with shareholder wealth maximization in mind or by executives’ selfinterest. Following agency theory on the impact of executive compensation on interest alignment, other scholars argue that moderate levels of ownership can bring the interests of managers more in line with that of shareholders (Hubbard & Palia, 1995). Loderer & Martin (1997) examine whether managers stock ownership leads to acquisition performance. Their results demonstrated that acquisitions announcement returns were positively associated with higher ownership of stocks. But higher ownership of stock was not positively related with acquisition performance. CEOs that have higher power over their board do receive bigger acquisition bonus, these bonuses did not result in higher stock returns (Grinstein & Hribar, 2004). They showed that powerful CEOs do make larger acquisitions compared to CEOs with less power, the market does react unfavorable to these kind of acquisitions. Research regarding the effects of stock holdings and incentive pay on acquisitions do show different outcome. Anderson et. al. (2000) conclude that confronting agency problems lies in alternative governance mechanisms, where there are large number of outside board members. Agency theory suggests on one hand the higher ownership by mangers should allow them to conduct in the best interest of shareholders (Denis et. al., 1999), on the other hand, when ownership increases, the personal wealth of the manager becomes even more tangled with the firm, agreeably making them reduce the risk of their portfolios through diversifying with mergers. Stock ownership for managers can act as effective instrument in solving agency problems of target company (Weisbach, 1993). Executive compensation is a vital part of the agency problem; the optimal level of remuneration is tough to conclude for any executive. Bebchuk & Fried (2004) find that powerful executives are able to determine their own compensate, because they are able to influence the board of directors with respect to their own reward package. The study examines diverse varieties of incentive packages (for example: retirement benefits, options, executive loans, retirement benefits and the utilization of consultants for the compensation process, etc.) and the relationship between executive power and pay determination. Managerial power theory, which also asserts that the executive compensation does not correlate well with the business performance, discloses that executive compensation is often excessive because managers have the power to overly influence decisions of the board of directors on their own compensation. There are three factors that influence the degree of. 9.

(10) managerial power: the nature of the board election, the character of the board, and that board members often do not have the means to challenge decisions (Murphy, 2002). There is a linkage between corporate governance and agency theory. Executives with more managerial power can also influence their remuneration package. As a result, the involved firm might face high agency costs. Choe et. al. (2008) examined the relationship between executive power, executive pay and business results. They conclude that the expected relationship between managerial power and pay is widely supported. However, they did not find the relationship between performance and power to be significant. They wrap up by emphasizing the relevance of the managerial power theory in explaining the relationship between power and pay. They also point out that further research is needed for better understanding of the relationship between performance and power. Core et. al. (1999) find that the governance and ownership structure explain a significant portion of the CEO compensation. When governance structures are less effective, this will lead to higher compensation. Companies with ineffective governance structures have large agency problems and the CEO receives in this case more compensation. Companies with greater agency problems tend to perform less. This study was based on the relationship between corporate governance, CEO compensation and firm performance. In a later study by Holstrom & Kaplan (2001) it is shown that corporate governance system in US have transformed significantly since 1980s. That era was dominated by intense M&A developments, characterized by prevailing hostile buyouts. These hostile buyouts decreased from then on, but the amount of M&A deals have gone up. Corporate governance mechanisms were instrumental in focusing the attention on shareholder value.. 2.3. Impact of M&A on executive compensation. Grinstein & Hribar (2004) revealed in their study on executive compensation as a result of M&A deals, that almost half of the executives of the acquiring party are rewarded for a positive merger or acquisition. They find that the more time and effort the executive invests in the deal, and the greater the deal is, the greater the final reward. These rewards were mainly in the form of cash and the authors conclude that M&A rewards are also well predicted by managerial power. Managerial power is the power that a top manager has to determine their own reward (Bebchuk et. al., 2002). The executives that are part of the board of directors can exert great pressure in positively influencing their compensation. This implies that the 10.

(11) compensation scheme set out by the board of directors is not fully functioning in the best interest of the firm. In a later study, Bebchuk & Fried (2003) find that the power of a CEO is significant in influencing executive compensation. Bliss & Rosen (2001) observe the outcomes of executive compensation on M&A deals. This research was based on the banking sector. Their findings indicate a significant relationship between acquisitions and executive compensation, regardless of the shares price plummeting afterwards. They conclude that majority of M&A deals ensure that the executive is individually better off at the expense of shareholders. In the paper of Wright et. al. (2002) they observe the relationship between executive compensation and M&A though from a different angle. They try to find out if monitoring plays a moderating variable in the relationship. They discover that external monitoring has an negative effect on executives’ selfish behavior and it helps in achieving shareholder objectives. They conclude that effective external monitoring can affect the executive compensation from M&A deals. Though, if there is a passive monitoring in place, executive compensation is mainly affected by the increase in company size due to the deal. Another study by Khorana & Zenner (1998) that focused on the role of executive compensation during acquisition decisions. Here they compare executive compensation for companies that undertake large acquisitions with companies that do not. They find a positive relationship at the time of the acquisition, between company and executive compensation of acquiring firms. This, however, is lacking in their control group, as the results show that larger firm size results in higher management compensation. However, large acquisitions do not lead to significantly higher total rewards but have indeed a positive effect. They conclude that good acquisitions lead to increased pay, but bad acquisitions have no effect on rewards. Kroll et al. (1990) research the effects of firm size, achieved primarily through acquisition, on CEO compensation. This study looks at both managers and owners of firms and they test whether there is an industry effect. They include the executive compensation one year after the merger has taken place in order to make the business impact of this merger more visible. They find that manager controlled firms give their CEOs higher rewards following several years of service, regardless of the company’s added performance after the merger. In owner controlled firms rewards are more based on performance. The conclusion is that acquisitions ultimately lead to a significant impact on CEO compensation for both manager as owner controlled companies.. 11.

(12) Amihud & Lev (1981) proved that executive controlled firms that are represented by reduced levels of ownership, are shown to be more likely to engage in risky mergers than firms that are owner composed. The rationale behind this is that self-serving managers of executives can more easily seek riskier strategies, that are not in the benefit of firm owners, who can gain diversification on their own in external asset markets. These findings were later confirmed by May (1995) where executive controlled firms were seen to show less risk taking nature amid periods of deregulation. The larger the executive compensation and equity holdings the more likely that firms were initiating diversifying acquisitions. However, other studies do not come to the same conclusion. Lane et. al. (1998) establish no empirical correlation among mergers and ownership. They concluded that there is no supporting data to confirm that executives are indeed diversifying through mergers to reduce risk and that these actions are expensive to stockholders. There is also another argument that can be made. Shareholders can benefit from risk reduction by mergers, because it permits making contracts with principals and agents optimal, this can enhance monitoring of executive actions. The competing views can be explained by the financial economics approach that Amihud & Lev (1981) have applied to their study compared to Lane et. al. (1998) research where they relied on strategic management view of ownership and diversification structures. In conclusion, these conflicting findings ultimately support the statement that shareholder value maximization is not the main interest when there is M&A diversifying deal. Executive ownership does also not serve as a much useful governance mechanism. Walking & Long (1984) researched the effects M&A transactions have on executives own wealth, they point out the decision executives of the targeted firm have to make: whether to resist or conspire in a particular offer. Targeted firm managers have two particular motivations contradicting each other. The first is that they might lose their job and with that their cash compensations when the merger is complete. The second one is that managers’ incentives related to the possible wealth increase they might extract from shedding their shares of the company at a higher price to the acquiring firm. However, if executives of the targeted firm pursue their own agenda and do not take the shareholders best interest in mind, they are more inclined to cooperate when there is a tender offer that their personal wealth will increase as a result the merger. Executives that cave in to the offer are found to be experiencing a significant gain in stock ownership as related to executives that did not accept the tender offer. Managers who are overpaid are more inclined to have a negative attitude toward a merger as they would like to see continuation of their job at the targeted company, while managers that 12.

(13) can gain personal wealth because of the bid are more incentivized in accepting. Studies also show (Weisbach, 1993) that executive resistance can be softened by giving managers more equity stake in the company. The merger can give them wealth gains, the perk from selling their stocks outweighs losing their jobs, this implies that the magnitude of stock ownership needs to be high enough for the targeted firm executives if there to be an alignment of interest to that of shareholders. Manager compensation is more motivated by the size of the corporation than its performance (Schmidt & Fowler, 1990). There are other papers that demonstrate the same aspect, where there is shown that manager compensation is marginally influenced by company performance (Khorana & Zenner, 1998). In this case, researchers showed that the probability of CEO being layoff after a poor acquisition, they also showed that poor transactions did not relate to higher executive pay.. 2.4. Impact of M&A deals on executive compensation of acquiring firm. How do M&A deals impact acquire managers’ compensation after the deal is complete? Are the post-acquisition era and the growth of firm size associated with monetary gains for managers? Conventional understanding in general agrees that managers have monetary incentives to initiate acquisitions. This is primary due to the positive effects of firm size on executive compensation. Avery et. al. (1998) presented empirical evidence against this belief. They followed multiple CEOs in their careers and investigated whether managers secured internal compensation (cash and stocks) or external compensation (i.e., outside director position) for their acquisitiveness. The authors showed that manager cash reward after the acquisition does not increase, however they did find an increase of external rewards. CEOs of acquiring firms are more likely to secure outside directorship. The acquisition may indicate to the outside world that the CEO has the expertise to manage a more complex company. They point out that executives are more interested in standing out in the business community when seeking M&A instead of their own compensation. Guest (2009) demonstrated that there is a significant increase in cash compensation in the postacquisition year for poor and good deals. When taking stock ownership into account, executives lose significant wealth when they forge poor acquisitions. Bliss & Rosen (2001) found that M&A transactions increase executive pay and total compensations of lower and higher acquiring managers, they showed that CEOs wealth increases at the expense of shareholders. In a similar paper (Anderson et al., 2004) found also that confirm this finding, 13.

(14) where poor deals appear to lead to significant increase in total compensation of management pay. Guest (2009) demonstrated that there is a significant increase in cash compensation in the post-acquisition year for poor and good deals. Harford & Li (2007) reported that acquiring firm that has a strong board, which was measured in relation to the executive tenure, will result in less strong increase of compensation around M&A deals. However, this is contradictory to the findings of Guest (2009), here there was no support for the link between weak corporate governance structures of acquiring firm and M&A deals. Grinstein & Hribar (2004) observed executives with more power over the board, they find that their executive compensation is much higher. They conclude that managerial power is the main driver of M&A deal bonuses. The results suggest that executives receive higher compensation after the post M&A period, because the board in their own company is now larger and less independent.. 14.

(15) 3. HYPOTHESIS. 3.1 Research question The research question of this thesis is summed up as the following: Does M&A deal size impact CFO compensation? And is there difference between CFO compensation when the deal size is bigger than the acquiring firms’ valuation? This research question is divided in three hypotheses, which are developed by the theory and by combining different studies mentioned in the theoretical part, where they touch on the two subjects separately.. 3.2 M&A and CFO compensation In the first hypothesis we want to look at, if there is a relationship between M&A and CFO compensation. The theory suggests there is significant effect. Weisbach (1993) indicated that if the executives’ personal wealth gains from the M&A they might me more inclined not to resist the deal. Anderson et. al. (2004) found evidence that even with poor M&A deals there is a significant increase in the total executive pay. Hypothesis 1: There is a positive relationship between M&A deals and CFO compensation.. 3.3 Firm size and Executive compensation The positive relationship between firm size and executive compensation can be validated by researchers as Kroll et. al. (1990) that proves that acquisitions lead to a significant increase on executive compensation, here they looked at the effect of firm size on these increases. They established this finding by looking at the bonuses of executives a year after the M&A has taken place, it is also indicated that higher executive pay is remunerated after multiple years of service to the firm. In Later studies Kroll et al. (1997) show that CEOs rewards gain from both acquiring successful company and also from simply enlarging the overall size of their firm. Harford & Li (2007) provide the same conclusion in their research, they show that equity wealth and compensation of executives are always higher from making M&A deals, same applies when these deals are value destroying. Blissen & Rosen (2001) concluded that the amount of management compensation significantly depends on the size of the deal. The theory implies the size of the deal is related to capacity of 15.

(16) the manager, better managers are capable to manage bigger firms. In Khorana & Zenner (1998) research there is an positive relationship demonstrated an between firm size and executive compensation, the bigger the firm the higher the management compensation. There is also the notion that executive compensation is more motivated by the size of the corporation than its performance (Schmidt & Fowler, 1990). There is universal understating that executives have incentives to instigate M&A deals, this mainly do the monetary incentives that come do to these deals (Avery et. al., 1998). This mainly do the effects of company size on management compensation. To answer our research question we have setup up the following two hypotheses to test the relationship between deal size on management compensation: Hypothesis 2: Does the size of the M&A have a positive effect on CFO compensation? Hypothesis 3: Does small firm taking over big firm matter in CFO compensation?. The third hypothesis will answer the main theme of this research. Small firms that acquire big firms are unique events. There is not a lot research done in this regard, thus we expect the data to be limited for these events. Scholars have observed the connection amid company size and growth. Historical research has revealed a undesirable connection between firm size and growth, demonstrating that small firms normally expand quicker than larger firms (Dunne & Hughes, 1994). Acquisitions can feed this hunger for fast expansion. Additional problem in connection to M&A and firm size is in the conducts of small and large acquirers. Moeller et al. (2004) describes that small acquirers incline to have better profits while large acquirers tend to display undesirable returns. The variance could be because of the fact that small firms usually acquire rather large firms, but still relative to their own size, while large acquirers have a habit to acquire rather small firms. Other studies have also researched the size (the proportion of acquiring firm size to acquired firms size) and they discovered that size might be of a dire influence to the performance of M&A deals (Seth, 1990). They also found that acquisitions of big firms produce more combined effects relative to small firms. Therefore, it is sound to take into account the relative size of an M&A deal in relation to executive compensation, because a large M&A might lead to additional growth and with added growth executives are usually payed more. Furthermore, M&A would influence small acquirers growth than that of large acquirers.. 16.

(17) 4.. METHOD AND DATA GATHERING. In this chapter we will construct our hypothesis based on different models and variables. These will help in defining the regression model to test the different hypothesis. We have used the model of Blissen & Rosen (2001) for formulating our variables.. 4.1. Conceptual model. Bodolica et. al. (2009) defines three M&A phases: Pre-M&A transaction (prior to the announcement), during the conduct of the M&A transaction (from the announcement until its complete) and post-M&A transaction (period after the deal is complete).. Figure 1. Framework on M&A transactions and executive compensation. Source: (Bodolica et. al., 2009, page 118). The figure above consists of the following three roles: the bidding company, the targeted firm that is going to be taken over and the acquiring company, which succeeds in completing the M&A transaction. This framework gives the focal point to the degree to which executive compensation is correlated with M&A deals. Executive compensation accounts to internal rewards (structure, magnitude and changes) and external rewards (membership in outside board of directors) which is appointed to executives of the acquiring company.. 17.

(18) 4.2. Dependent variables . Total CFO compensation as listed in SEC (Security and Exchange Commission) filings. This is a compensation of base salary, equity and bonuses..  . 4.3. Current compensation which represents salary and bonus in US Dollars. Equity Compensation, options and stocks awarded to the CFO.. Independent variable: . M&A deals with a valuation of more than 100 million, this will make our analysis more relevant, because such a value is considered significant economic events (Grinstein & Hribar, 2004). o The following data is considered within the independent variable:. . . Size of M&A deal (deals value > 100 million). . Owned percentage of shares after the transaction (51% or more). . Status of deal completed. Small takeovers, here we look for small firms that acquire bigger firms. We consider this event to happen when the deal value is more than the valuation of the acquiring firm at the date of the M&A.. 4.4. Control variables . CFO age. . CFO gender. . Firm Value (Acquirers market value, 4 weeks prior to deal announcement). . Deal value. . Time To Complete. . Economic Crisis (M&A deals before 2008 and after). . ROA (Return on Assets). . Region (4 dummy variables). . Industry (4 SIC codes, 4 dummy variables). . Years (dummy variables for each year of the completed M&A deal). 18.

(19) Table 1: Definition of used variables Variable:. Description:. LnReportedComp. Natural logarithm of the total compensation in US Dollars. Data gathered in COMPUSTAT/ Execucomp: Total compensation As reported in SEC filings. In thousands of dollars per year.. LnCurrentComp. Natural logarithm of present compensation (salary and bonus) in US Dollars. Data from COMPUSTAT / Execucomp: “Total current compensation”. In thousands of dollars per year.. LnEquityComp. Natural logarithm of equity component (options and stock) in US Dollars. Data gathered from COMPUSTAT / Execucomp: Value of Option and Stock Award in thousands of dollars per year.. LnDealSize. Natural logarithm of the value of M&A deal size in millions of US Dollars. THOMSOM M&A database.. LnFirmSize. Natural logarithm of the acquiring firms’ Total assets in millions of US Dollars. THOMSON M&A database. LnTimeToComplete. EconomicCrisis. Natural logarithm of the total number of days between announcement and completion of the deal. THOMSON M&A Database Dummy 0 if a deal is completed between 2006 and 2008, 1 if the deal is completed after 2008.. ROA. Return on Assets of the acquiring firm. THOMSON M&A database, acquirers net income of the last twelve months divided by acquirers net assets.. AssetsToDealSize. Control variable = Deal size as a percentage of the acquiring firms’ total assets.. SmallTakeovers. Small firms acquiring bigger firms. THOMSON M&A database Dummy 1 if firm value is smaller than deal value.. Region(1). (Dummy 1) Acquirers nation: United states (reference category), THOMSON M&A database. Region(2). (Dummy 1) Acquirers nation: Canada and Australia, THOMSON M&A database. Region(3). (Dummy 1) Acquirers nation: Europe, THOMSON M&A database. Region(4). (Dummy 1) Acquirers nation: The rest of the world, THOMSON M&A database 19.

(20) CFOage. Age of the CFO during the completion of the deal. COMPUSTAT / Execucomp database. CFOgender. Woman (dummy 1), man (dummy 0). Gender of the CFO during the completion of the deal. COMPUSTAT / Execucomp database. DebtToAsset. Long-term debts as a percentage of total assets. Debt that is held for longer than 12 months. THOMSON M&A database. FeesToDealsize. Fees paid to M&A advisers as a percentage of the deal size. THOMSON M&A database. AfterTheDeal. CFO compensation a year after M&A deal has taken place compared to before the deal. COMPUSTAT / Execucomp database (Dummy 1) Industry of the acquiring firm; Gas, electric (reference category). Sic(1) Sic(2). (Dummy 1) Industry of the acquiring firm; trade organizations. Sic(3). (Dummy 1) Industry of the acquiring firm; service organizations. YearDummy(1). (Dummy 1) M&A deals completed in year 2006 (reference category). YearDummy(2). (Dummy 1) M&A deals completed in year 2007. YearDummy(3). (Dummy 1) M&A deals completed in year 2008. YearDummy(4). (Dummy 1) M&A deals completed in year 2009. YearDummy(5). (Dummy 1) M&A deals completed in year 2010. YearDummy(6). (Dummy 1) M&A deals completed in year 2011. YearDummy(7). (Dummy 1) M&A deals completed in year 2012. YearDummy(8). (Dummy 1) M&A deals completed in year 2013. YearDummy(9). (Dummy 1) M&A deals completed in year 2014. YearDummy(10). (Dummy 1) M&A deals completed in year 2015. 20.

(21) 4.5. Regression analysis. The dependent variable (executive compensation) consists of base salary, bonuses, stock and options. Therefore, to better analyze the data we will make 3 regression analyses. This will add to the overall explanatory power of our analysis. In the first model we will look at the total CFO compensation. In model 2, the salary and bonuses are evaluated and in model 3 we will examine the options and stock awarded to the CFO.. 4.6. Data sources. The following databases are used to conduct the research: . CFO compensation - COMPUSTAT Executive Comp (WRDS). . Mergers & Acquisition deals - Thomson ONE M&A database. 21.

(22) 5.. DATA. The analysis in this dissertation draws on data from the COMPUSTAT Executive Comp (WRDS) database and the M&A - Thomson ONE M&A database. While the former contains data on the compensation of executives and their characteristics, the latter includes information about the merger and acquisition and characteristics of the companies. We focus on mergers and acquisitions that have taken place in the last decade, between 2006 and 2015. The sample includes those CFOs who were at the company during the year of the M&A and a year after. In total, there are 657 observations. (We deleted one observation because it had negative values for compensation. This is probably an error in the data.) The majority of the observations are from executives in the United States (89.8%), followed by Europe (5.0%), Canada and Australia (4.7%), and the rest of the world (0.5%).. 5.1. Dependent variable. The dependent variable is CFO compensation. We use three different measures of compensation. Firstly, total compensation as reported in SEC filings. This measure captures the whole compensation amount. Secondly, total current compensation which consists of the CFO’s salary and bonuses. Thirdly, total equity compensation. This is the difference between the first and the second measurement. It is likely that executives are rewarded mostly through equity compensation. All measures are in thousands US dollars. By investigating three different measures of compensation, we can explore whether successful mergers and acquisitions have a stronger effect on some types of compensation than on others. Figure 2 displays the trend in compensation for CFOs from 2006 to 2015.. 22.

(23) Figure 2. Average CFO compensation (2006-2015). We see that the current compensation (the red line) does not fluctuate much during this period. The fluctuations in the total SEC compensation seem to be driven mainly by changes in the equity compensation. There are two low points: in the period of the economic crisis (20072009), rightly after the compensation increases again. Then a sharp decline in 2013. This can be explained by taking for example the US (United States) GDP growth from 2012 to 2013, the US did slowdown in GDP growth from 2.2 to 1.5 (World Bank and OECD data files). We see in the economic data that in the year 2012 there was an increase in GDP in comparison with the previous year. This decline is due to the declining growth. Based on the economic data and the acquisition deals in these years we see a correlation between these two, the acquisition deals size is higher in economic growth. After 2012 the average deal size is going down and if we look at the economic data in figure 3, the economy is worsening after 2012.. 23.

(24) Figure 3. US GDP growth rate (2006-2015). The correlation between total SEC compensation and total current compensation is positive, r=0.47, and significant (p<0.01). The correlation between total SEC compensation and equity compensation is almost perfect: r=0.98 and p<0.01. The correlation between total current compensation and equity compensation is moderately strong and positive: r=0.31, p<0.01.. 5.2. Independent variables. The first independent variables is the size of the M&A deal. We expect that CFOs will be rewarded more when they manage to close a deal of higher value. Figure 4 displays the average deal size for the period 2006 to 2015. The line follows the trends in compensation. We see sharp declines in the average deal size in 2008 and in 2013. These developments are related to the broader economic changes.. 24.

(25) Figure 4. Average deal size (2006-2015). The second independent variable is the size of the acquiring firm’s valuation relative to the deal size. In other words, it measures whether small firms are taking over large firms, and vice versa. The following graph shows the distribution of this variable. Of all 657 observations, 59 (or 9%) are CFOs of small firms who took over large firms and 598 (or 91%) are CFOs of large firms who took over small firms. It is thus more common for large firms to take over small firms than the other way around. In the rare instances that small firms take over large firms, we would expect a higher compensation because it may reflect greater effort in achieving this success. Figure 5 displays the proportion of small firms taking over large firms between 2006 and 2015. We see two things. Firstly, these types of takeovers have always been rare events. Secondly, there is a declining trend after 2009. These types of takeovers seem to have occurred less in the most recent years.. 25.

(26) Figure 5. The proportion of small firms taking over large firms (2006-2015). 5.3. Control variables. We also include a number of control variables in the regression models. These variables may influence compensation and have been found to be important by previous studies. The control variables are: a dummy for the gender of the CFO (1 if the CFO is a woman, 0 if a man), age of the CFO, the number of days to complete the deal, a dummy for the economic crisis, four dummies for the region in which the company is settled (the US as the reference category, three dummies for Canada and Australia, Europe, and the rest of the world), three dummies for the industry in which the company operates, the value of the target’s assets as a percentage of the deal value, value of the acquiring firm, the value of the target firm, and returns on equity, long-term debts as a percentage of total assets, and fees as a percentage of the deal size. We also include year dummies to control for yearly fluctuations. Table 2 describes how all variables are measured.. 26.

(27) Table 2: Summary statistics of the variables. VARIABLES LnReportedComp LnCurrentComp LnEquityComp SmallTakeovers AfterTheDeal LnTimeToComplete LnDealSize LnFirmSize Age CFOgender DebtToAsset Return on Assets EconomicCrisis Sic group Region AssetsToDealSize FeesToDealsize YearDummy1 YearDummy2 YearDummy3 YearDummy4 YearDummy5 YearDummy6 YearDummy7 YearDummy8 YearDummy9 YearDummy10 ** See table 1, page 19 for the definition of used variables and their specific number format.. (1) N. (2) mean. 657 7.666 656 6.242 656 7.284 657 0.0898 657 0.502 655 4.698 657 6.704 657 8.432 657 50.73 657 0.0852 487 23.83 657 0.0997 657 0.501 657 2.484 657 1.161 657 251.6 116 0.633 657 0.149 657 0.184 657 0.116 657 0.0457 657 0.126 657 0.0715 657 0.0989 657 0.0974 657 0.0548 657 0.00609. (3) sd. (4) min. (5) max. 0.848 5.095 0.542 3.912 1.116 1.163 0.286 0 0.500 0 0.598 2.303 1.352 4.606 1.548 4.852 6.289 34 0.279 0 25.62 0.00500 0.980 -16.77 0.500 0 0.769 1 0.513 1 758.0 2.772 0.510 0.00782 0.357 0 0.388 0 0.320 0 0.209 0 0.332 0 0.258 0 0.299 0 0.297 0 0.228 0 0.0778 0. 11.14 9.577 11.12 1 1 6.405 10.79 13.21 69 1 323.2 2.972 1 3 4 7,700 2.718 1 1 1 1 1 1 1 1 1 1. Table 2 presents the descriptive statistics of all variables. With regards to the dependent variables, we see that the average log compensation is 7.7 for total compensation as reported in SEC filings, 6.2 for current compensation, and 7.3 for equity compensation. The differences between the minimum and maximum values are relatively large, which indicate that there is a lot of variation between companies in terms of their compensation. While some CFOs are rewarded with very high compensation, others receive lower compensation. Looking at personal characteristics of the CFOs, we find that CFOs are on average 50.7 years old and 91.5% of all CFOs are male. The average time to complete a deal, the number of days between announcing the deal and closing the deal, was 130.7 days. One deal was announced and completed in the same day 27.

(28) (minimum=0). This concerned the take-over of Heracles General Cement Co SA based in Greece by Lafarge SA in France. The longest time to complete a deal was 605 days. This was the MA of TOnline International AG by Deutsche Telekom in 2006. The size of deals is on average 2,397 and ranges from 100.1 to 48,766 million. The variable EconomicCrisis has a mean of 0.5 which means that 50% of all deals in the sample took place after the 2008. The acquiring firms are generally much larger than the target firms. While the former is worth on average 17,045 the latter is on average 1993 million. The return on equity is positive, on average 0.1. This means that the assets as a percentage of the deal size is on average 251.3.. Table 3: Correlation matrix of the independent variables.. AfterTheDeal SmallTakeovers DealSize CFOgender CFOage TimeToComplete EconomicCrisis Sic rv Region AssetsToDealSize ~e FeesToDealsize Acq. FirmSize Target FirmSize. After The Small Deal size Deal Takeovers. CFO gender. 1.0000 -0.0391 -0. 0325 -0.0127 -0.0157 -0.0304 0.1064* -0.0019 -0.0493 -0.0041 0.0645 -0.0052 -0.0433. 1.0000 0.2318* 0.0949* -0.0281 0.2544* -0.0266 0.0166 0.0051 -0.0417 -0.0357 -0.0838* 0.1247*. 1.0000 -0.0065 0.0907* 0.2501* 0.0145 -0.1310* 0.0493 -0.0066 -0.3996* 0.3416* 0.8394*. 1.0000 -0.0945* 0.0306 -0.0327 -0.0152 0.0210 -0.0518 0.0072 0.0105 -0.0226. Sic. Region. Assets To Fees To Acq. Target DealSize Dealsize Firm Size FirmSize. Sic 1.0000 Region -0.0401 1.0000 AssetsToDealSize 0.0886* 0.1719* FeesToDealsize 0.0152 -0.1928* ~e Acq. FirmSize -0.0012 0.0165 Target FirmSize -0.1000* 0. 1905* Note: ‘*’ denotes p-value <0.05. 1.0000 -0.0334 0.0220 0.1590*. CFO age. 1.0000 0.0199 0.0864* 0.0306 0.0491 0.0269 0.1179 0.0718 0.0576. Time To Economic Complete Crisis. 1.0000 -0.0220 0.0018 0.0123 0.1138* -0.1965* 0.0204 0.2478*. 1.0000 -0.1048* 0.0000 -0.0194 0.0525 0.0214 0.0010. 1.0000 -0.2742* 1.0000 -0.3143* 0.2952* 1.0000. None of the variables are highly correlated with each other. This is a good sign, because high correlations could lead to problems in the estimation. This does not seem to affect us here. For example, small takeover has a positive and significant relationship with deal size. The Pearson correlation coefficient is 0.23. This could be interpreted as a weak to moderate relationship. 28.

(29) 5.4. Bivariate analysis. In order to test whether a M&A deal has an effect on CFO compensation, We compare compensation of CFOs before and after the M&A deal has taken place. We used an independent samples t-test first and created a dummy which is 1 for the period after the M&A deal (including the year of the M&A itself). The log of CFO reported compensation is nearly as high in the period after the M&A deal as in the period before the M&A deal. The p-value is 0.33, which means that this difference is not significant, although it is in the expected direction. The t-test does not provide support for the first hypothesis. We then repeated the test for the two other measures of compensations. Again, the differences are small and insignificant. This tells us that CFO compensation does not seem to be much higher in the period after the M&A deal. Table 4: Two-sample t test with equal variances Group. Obs. Mean. 0. 327. 1 combined. Std. Err.. Std. Dev.. [95% Conf.. Interval]. 7.65236. .0440269. .7961439. 7.565747. 7.738972. 330. 7.680343. .0494229. .8978113. 7.583119. 7.777568. 657. 7.666415. .0330915. .8482025. 7.601437. 7.731393. -.1580231. .1020557. diff. -.0279837. .0662252. diff = mean(0) - mean(1) Ho: diff = 0 Ha: diff < 0 Pr(T < t) = 0.3364. t = -0.4226 degrees of freedom = 655 Pr(|T| > |t|) = 0.6728 Ha: diff != 0. Pr(T > t) = 0.6636 Ha: diff > 0. The second hypothesis tests whether the size of the M&A deal has a positive effect on the level of CFO compensation. In bivariate correlations between deal size and the various measures of compensation, we find that deal size is positively and significantly (p<0.01) correlated with compensation. This means that the larger the size of the deal, the higher the compensation of the CFO. The correlations are moderately strong in general.. 29.

(30) Table 5. Bivariate correlations. Deal size. Reported compensation. Current compensation. Equity compensation. 0.46**. 0.34**. 0.41**. Note: ‘**’ means that p-value is below 0.01.. We will also explore the relationship between compensation and the acquiring firm’s valuation relative to the deal size. An independent-samples t-test was conducted to compare compensation for small firms taking over large firms and large firms taking over small firms. For total reported compensation, the compensation for small takeovers was higher than for large takeovers. This is in the expected direction of my hypothesis. This difference is significant: p-value <0.01. For current compensation, small takeovers also received more than large takeovers. Again, in the right direction and the difference (-0.25) is significant (p<0.01). For equity compensation, small takeovers received more than large takeovers. Again, in the right direction. The p-value for this difference is also significant (p<0.01). In summary, for three different measures of compensation we find that CFOs of small firms that take over large firms receive significantly higher compensation than CFOs of large firms that take over small firms. The difference is not very large though. The next analyses will explore whether this effect holds once we control for other variables.. 5.5. Multivariate regression analysis. The bivariate analyses provide some support for the hypothesis, in particular for the hypothesis concerning small firms taking over large firms. In this section, we will use multivariate regression analysis to test the hypothesis. We include several control variables in order to make sure that the found effects of the main independent variables of interest are robust to the inclusion of alternative explanations. Then we will run three models. The main difference between these models is the dependent variable. Model 1 has total reported compensation as the dependent variable, model 2 total current compensation, and model 3 total equity compensation. The results are reported in Table 3 below and we will discuss and interpret each model are separately.. 30.

(31) The models can be summarized with the following regression formula: CFO compensation = b0 + b1 * Deal size + b2 * Small takeover + b3 * CFO age + b4 * CFO Gender + b5 * Completion Time + b6 * Acquiring firm size + b7 * Return on Assets + b8 * Economic Crisis + b9 * Region + b10 * Industry + b11 * Year dummies + error. Table 6. Regression analysis. VARIABLES AfterTheDeal LnDealSize SmallTakeovers CFOage CFOgender LnTimeToComplete LnFirmSize AssetsToDealSize ROA EconomicCrisis 2.Region 3.Region 4.Region 2.Sic 3.Sic Constant. Observations R-squared. (1) m1 LnReportedComp. (2) m2 LnCurrentComp. (3) m3 LnEquityComp. 0.016 (0.051) 0.054* (0.025) 0.680*** (0.098) 0.008* (0.004) 0.072 (0.088) -0.139** (0.044) 0.337*** (0.020) -0.000 (0.000) -0.003 (0.025) 0.199 (0.195) 0.382** (0.117) -0.333** (0.117) -0.589 (0.358) -0.113 (0.085) -0.110 (0.068) 4.550*** (0.316). -0.030 (0.038) 0.014 (0.019) 0.398*** (0.073) 0.006* (0.003) 0.170** (0.065) -0.044 (0.033) 0.171*** (0.015) 0.000** (0.000) 0.001 (0.018) 0.167 (0.146) 0.218* (0.087) -0.164 (0.088) -0.332 (0.267) 0.003 (0.063) -0.038 (0.051) 4.514*** (0.236). 0.010 (0.072) 0.051 (0.036) 0.834*** (0.138) 0.010 (0.005) -0.019 (0.123) -0.155* (0.062) 0.409*** (0.029) -0.000* (0.000) -0.008 (0.035) 0.147 (0.274) 0.470** (0.164) -0.338* (0.165) -0.659 (0.504) -0.170 (0.119) -0.181 (0.096) 3.505*** (0.445). 655 0.501. 654 0.318. 654 0.429 31.

(32) Year Dummies Adjusted R-squared F test. Yes Yes 0.482 0.292 26.36*** 12.22*** Standard errors in parentheses *** p<0.001, ** p<0.01, * p<0.05. Yes 0.408 19.72***. Before we move on to the interpretation of the coefficients, we note that we have tested the models for multicollinearity. Multicollinearity indicates whether the independent variables are highly correlated with each other. If this is the case, then linear regression models cannot accurately estimate the effect of each of these independent variables. Multicollinearity is high when one independent variable can be written as a function of another independent variable. It is generally recommended to test for multicollinearity and we have inspected the tolerance and Variance Inflation Factors (VIF) after each model. As a guideline, VIF scores above 5 indicate problems of multicollinearity. Most VIF scores are well below 2 and the average VIF of all independent variables is 2.8. The only variables with high VIF scores are the dummy variable for crisis and the dummy variables for 2010 to 2014. As a robustness test, we re-ran the analysis without the dummy for crisis. This solved the problem of high VIF. The results without the dummy for crisis are very similar results. We should also note that we used a log transformation of the continuous variables expressed in US dollars to deal with the problem of heteroscedasticity. This means that the variances are not constant. This would violate an important assumption of linearity, namely that the errors are normally distributed. While heteroscedasticity does not affect the coefficients, it will lead to biased standard errors. If left unaddressed, this can lead to misleading results. By rescaling these variables by using the log transformation, this problem can be addressed. The log transformation changes the shape of the data distribution from very skewed to a more normal distribution. We therefore proceed with the analysis and interpret the results. A few words on the models in general. The adjusted R-square measures the amount of variation in the dependent variable that is explained by the independent variables included in the model. Unlike the normal R-squared, the adjusted version takes into account the number of variables included in the model. In other words, it penalizes large models and rewards more parsimonious models. We see that the adjusted R-squared is more conservative than the normal R-squared: its values are always lower than of the latter. Of the variation in total reported CFO compensation, we find that 48.2% is explained by the model. This is substantially lower, namely 29.2%, for total current compensation. It is 40.8% for total equity compensation. The explanatory variables are thus better at explaining total reported and equity compensation for 32.

(33) CFOs than at explaining total current compensation. The F-test tests the overall significance of the model. In other words, it tells us whether the model with the explanatory variables is better than a model without any independent variables. A significant F-statistic rejects the null hypothesis that the intercept-only model explains as much as the full model. In all three cases, we find that the F-test is significant which means that our models are better than a model without independent variables. Note that year dummies are included in all three models to account for year-specific events, macro-economic effects, and other factors that may vary yearly. We have not reported them in the table to save space. Compared to 2006, CFO compensation was significantly higher in 2007, 2012, and 2014. The coefficients for the year dummies of other years are not significantly different from 2006. Model 1: This model tries to explain the variation in total reported compensation of CFOs. The first variable, the period after the M&A deal, has a small and positive effect on compensation. This means that CFO compensation is higher after the M&A deal than in the year before the M&A deal. The difference, however, is not significant. This confirms our previous finding that the results do not provide support for the first hypothesis. Deal size has a small positive and significant effect on compensation. This is in line with our expectation that larger deals are rewarded because more effort was required from the CFO. The results thus seem to support the second hypothesis. The dummy variable for small takeovers is positive and highly significant (p<0.001). This means that CFO compensation is higher when small firms take over large firms, than when it happens the other way around. The findings in the bivariate analysis are thus confirmed here in the multivariate analysis as well. This can be interpreted as support for the third hypothesis. We will now interpret the control variables. Firstly, age has a positive and significant effect on reported compensation. Older CFOs generally receive higher compensation than younger CFOs. The gender of the CFO does not have a significant effect on reported compensation. Women have slightly higher compensation than men, but the standard errors are very large. This could be because there are very few women in the sample. Next, completion time has a negative and significant effect. The longer it takes to complete the deal, the lower the compensation for the CFO. Again, this suggests that reward is to some extent based on the effort of the CFO. Long processes are punished by lower compensation. The size of the acquiring firm has a positive and highly significant effect (p<0.001) on reported compensation. The larger the acquiring firm is, the higher the compensation for the CFO. This makes a lot of 33.

(34) sense because larger firms have more resources and can therefore more easily pay out higher compensation to their executives. The remaining control variables are not significant, but it is still worth to interpret them. Firstly, assets as a percentage of deal size is negative, but insignificant. Although we expected that the higher the deal value with respect to the acquiring firms assets would lead to higher compensation, because this would mean bigger and more complex firm to manage, this is not found in the data. Return on assets has a positive effect: the higher the return on assets, the higher the compensation. The coefficient, however, never reaches significance. The dummy for the economic crisis has a positive, but insignificant effect. This is counterintuitive. It means that CFO compensation was higher in the period after the crisis than in the period before the crisis. The dummy, however, is not significant. For the geographical region, the US is the reference category. The dummies can be interpreted as follows: CFO compensation in Australia, Canada, and New Zealand is higher than in the US. The difference is significant. Compensation in Europe is significantly lower than in the US. This makes sense because executive compensation is generally lower in European countries. Compensation is also lower in the rest of the world than in the US, but these differences are not significant. Finally, we see that the Sic of the company also does not matter for CFO compensation. CFO compensation in trade and service organizations are lower than in gas and electric organization, but the differences are not significant. Since most independent variables are not measured on the same scale, we cannot easily compare their effect sizes. we have therefore also calculated the standardized effect sizes. These are reported in the table below. we will limit the discussion here to the main independent variables of interest and other significant control variables. One standard deviation increase in AfterTheDeal (the period after the MA deal) leads to a 0.008 unit increase in reported compensation. This is very small. One standard deviation increase in deal size, however, leads to a 0.073 unit increase in compensation. The effect of size is much larger for small takeover. One standard deviation increase in small takeover leads to a 0.195 unit increase in compensation. One standard deviation increase in completion time leads to a 0.083 unit decrease in compensation. Completion time is thus half as important as small takeover. One standard deviation increase in the firm size of the acquiring firm leads to a 0.522 unit increase in CFO compensation. This is by far the most important determinant of CFO compensation.. 34.

(35) Table 7. Standardized beta coefficients of model 1. VARIABLES. (1) Beta LnReportedComp. AfterTheDeal 0.008 LnDealSize 0.073* SmallTakeovers 0.195*** CFOage 0.049* CFOgender 0.020 LnTimeToComplete -0.083** LnFirmSize 0.522*** AssetsToDealSize -0.028 ROA -0.003 EconomicCrisis 0.100 2.Region 0.081** 3.Region -0.071** 4.Region -0.040 2.Sic -0.043 3.Sic -0.052 Note: The beta coefficients show the effect of one standard deviation change in the independent variable on the dependent variable. *** p<0.001, ** p<0.01, * p<0.05. For the interpretation of model 2 and 3, we will be more brief. Model 2 attempts to clarify the difference in salary and bonuses, while model 3 strives to explain stocks and options, related to executive compensation following an M&A deal. We find that the SmallTakeover is significant for both model 2 and 3, this means that when CFO has the opportunity of acquiring another firm with higher deal value than its own assets, it will more likely lead to higher compensation. Firm size is also significant, but assets to deal size is not significant, however it does have some impact on CFO pay.. 35.

(36) 5.6. Robustness tests. This section describes several robustness tests, which are reported in Table 8 below. We were interested in another control variable, namely debt to deal size. Debt financing can be seen as an alternative control mechanism, when linked with executive compensation, can boost the effectiveness of corporate governance in the affected companies. In countless of cash M&A deals, bidders has to incur debt to pay for the acquisition. Jensen (1986) points out the debt can reduce agency cost, when managers are confronted with borrowing costs to finance the deal, they are pressured by debt providers to make sane judgements regarding M&A deals. Debt increases bankruptcy risk and higher leverage can produce better outcomes of performance of M&A deals. We did not include this in the main models because this variable has more missing values and the sample size drops substantially: from 657 to 487 observations. Long term debts have a positive, but insignificant effect on CFO compensation. Importantly, however, we find that the effect of smalltakeover does not change much. It is robust to the inclusion of this control variable. A second robustness test concerns the operationalization of the dummy for economic crisis. It is possible that we didn’t find an effect initially due to the coding choice. Remember that we coded all years after 2008 as crisis years. For the robustness test, we recoded the years 20062008 and 2013-2015 as non-crisis years, and 2009-2012 as crisis years. Model 5 shows that the new measure of economic crisis also does not affect CFO compensation. Finally, we wanted to test fees as a percentage of deal size. The logic behind this is that the more companies have to spend on the deal in terms of fees, the lower the compensation for executives will be, because the more adviser expertise services is attracted, the less the CFO is needed. Unfortunately, the data on fees contains many missing values and the sample drops to 116 observations. We therefore limit the number of independent variables and only include the most important control variables in additions to fees as a percentage of deal size. The results have to be interpreted with more caution. Fees have a negative effect on CFO compensation. This is in the expected direction: the higher the fees, the lower the compensation. The coefficient, however, is not significant. The coefficient of small takeover is no longer significant. This is probably the result of the smaller sample size.. 36.

(37) Table 8. Regression analysis. VARIABLES AfterTheDeal LnDealSize SmallTakeovers age CFOgender LnTimeToComplete LnFirmSize assets_to_dealsize roa EconomicCrisis 2.Region 3.Region 4.Region 2.Sic 3.Sic ltdebts EconomicCrisis2. (1) m4 LnReportedComp. (2) m5 LnReportedComp. (3) m6 LnReportedComp. -0.004 (0.056) 0.052 (0.028) 0.747*** (0.107) 0.007 (0.004) 0.071 (0.097) -0.190*** (0.051) 0.338*** (0.025) -0.000 (0.000) 0.004 (0.024) 0.024 (0.225) 0.506*** (0.144) -0.350** (0.133) -0.548 (0.345) -0.047 (0.092) -0.034 (0.075) 0.001 (0.001). 0.036 (0.049) 0.053* (0.025) 0.682*** (0.098) 0.008* (0.004) 0.074 (0.088) -0.140** (0.044) 0.337*** (0.020) -0.000 (0.000) -0.001 (0.025). 0.038 (0.124) 0.359* (0.151) 0.220 (0.214). 0.583 (0.547) 0.379** (0.116) -0.315** (0.118) -0.586 (0.358) -0.115 (0.084) -0.110 (0.068). 0.164 (0.118). FeesToDealsize Constant Observations R-squared Year Dummies Adjusted R-squared F test. -0.040 (0.134) -0.001 (0.149). 4.830*** 4.535*** (0.350) (0.316) 485 655 0.516 0.502 Yes Yes 0.489 0.483 19.54*** 26.44*** Standard errors in parentheses *** p<0.001, ** p<0.01, * p<0.05. 0.186 (0.131) 4.560*** (0.762) 116 0.503 Yes 0.422 6.253***. 37.

(38) 6.. CONCLUSION. This thesis has examined the relationship between M&A deals and executive compensation. The research question was to answer whether executive compensation differs between M&A deals done by big firms acquiring small firms and small firms acquiring big firms. In previous sections of this paper we presented a theoretical framework that built on an extensive literature on executive compensation, mergers and acquisitions, and the effects of M&A deals on executive compensation. Whereas this literature has predominantly focused on CEO compensation, we have used these insights to explain CFO compensation. In the empirical analysis, we examined the causal relationship between M&A and CFO compensation, firm size and executive compensation. We tested several hypotheses using OLS regression analysis and found support for some of our hypotheses. Our analysis allows us to draw the following conclusions. Firstly, we find no evidence for our first hypothesis, which was that compensation for CFOs is higher after the M&A than before the M&A. Although there was a small difference in compensation between these two periods, it was not statistically significant. We can conclude that CFO compensation is not significantly higher after the M&A deal than in the year before the M&A deal. This makes us reject our H1 hypothesis. Secondly, deal size has a positive and significant effect on CFO compensation. This means that the larger the size of the deal, the higher the compensation of the CFO. The effect is significant for reported compensation and almost significant for equity compensation, but it is insignificant for current compensation. We can conclude that deal size affects reported compensation, but not the salary components. The empirical analysis thus provides some support in favor of our second hypothesis. Finally, we find strong support for our third hypothesis that the size of the acquiring firm relative to the deal size for CFO executive compensation. We explored the relationship between compensation and the acquiring firm’s valuation relative to the deal size in an independent t-test and a multivariate regression analysis. We found that total reported compensation for CFOs was higher for small firms that take over large firms than for large firms that take over small firms. This is in line with our theoretical expectation, although we did not expect to find such strong support from our model. The results are strong and highly significant across models, regardless of the measure of compensation used.. 38.

(39) In conclusion, for three different measures of compensation we find that CFOs of small firms that take over large firms receive significantly higher compensation than CFOs of large firms that take over small firms. It must be noted, however, that the occurrence of small firms taking over large firms is a rare economic event. In the decade under study, this happened in less than 10% of all observations. Nevertheless, our expectations were confirmed that higher compensation is rewarded to CFOs, because of the perceived efforts in achieving this.. 6.1. Limitations. This study is not without limitations. Because our topic of interest is an economic event that does not happen often, resulting in limitations when taking into account the data and the scope of the research. Our data is only between 2006 and 2015. It is possible that the results are different when a different time period is studied. Moreover, our data is also dependent on the number of the M&A deals conducted. From 657 observations only 59 (or 9%) were CFOs of small firms who took over large firms. It would be interesting to explore further in which respects the deals of small takeovers differ from the deals by large takeovers. The moderate adjusted R-squared value also indicates that not all variance is explained by our model. Future research would benefit from including additional variables in order to explain the remaining variance of CFO compensation. The empirical analysis here was not limited to any particular industry. Instead, we grouped observations from several industries together in the analysis and found no significant differences between the various industries. This suggest that the results can be generalized to some extent to other industries. The majority of cases, however, originated from developed economies, and the US in particular. This is also true for the number of small firms taking over large firms: of all 59 instances, 53 took place in the US. Given the small number of instances in Canada, Australia, and Europe, we cannot exactly say how this will play out in different settings. It would be interesting to examine small takeovers in greater detail, for example in case studies to learn more about the conditions under which they take place. This will increase our understanding of how these special events may affect CFO compensation. Moreover, the analysis has relied on cross-sectional data. Ideally, however, we would like track the same executives over a period of time. This would strengthen our causal claims by controlling for unobserved differences between firms and individual CFOs.. 39.

No results found