The effect of CEO compensation structure on M&A performance : evidence from American acquiring companies

(1)

The effect of CEO compensation structure on M&A

performance: Evidence from American acquiring

companies

Master Thesis

Thesis Author: Jun Shi

Thesis Supervisor: Dr. Tomislav Ladika

MSc Business Economics, Finance track

Faculty of Economics and Business, University of Amsterdam

(2)

Acknowledgements

I take this opportunity to express my gratitude to my entire family for their endless support. Without my family I would not be enable to pursue my academic life.

I also take this opportunity to express my gratitude to my thesis supervisor, Dr. Tomislav Ladika, who inspired and guided me in completing this master thesis.

Lastly, I thank my friends, Faculty of Economics and Business and University of Amsterdam for their encouragement and help without which this thesis would not be possible.

(3)

1

Abstract

This paper aims to provide insights into CEO compensation structure design. By using a sample of 2,552 completed acquisition from 2003 to 2013 within American non-financial firms, this paper studies how equity-based compensation fraction is affecting merge and acquisition performance around and following the investment decision announcement. The result documents that equity-based compensation negatively related to the M&A performance. Further equity-based compensation leads to worst performances in firms with high growth opportunity when splitting samples into high and low growth opportunity levels. Firms are examined to grant wrong incentives in the form of stock option to decrease shareholders’ wealth.

(4)

2

I Introduction

In corporate finance field, an “agency problem” has been raised where the confliction between the manager self-interest and the shareholder value may makes shareholders lose from value-destroying decision. While most investment decisions are relatively small, merge and acquisition are major and observable long-term investments. Executive compensation, designed to prevent agency problem, grew enormously during the last two decades. The growth is mainly caused by the parts of restricted stock and stock option. Thus, it seems that companies are increasing equity-based compensation to prevent loss from management and maximum shareholder value in M&A deals.

Companies may grant equity-based compensation differently due to the companies’ characteristics, which result to different effects on M&A performance. Consider these two types of firm. A young software firm will pay its CEO mainly in stock and option because the firm generates rare cash in the begging period of business. However, an established consumer goods firm will pay its CEO mainly in case-based compensation because the CEO may have many job options and wishes to receive a low-risk compensation. The fraction of equity-based compensation between these two firms are obviously different. On the other hand, the CEO in high growth opportunity firm makes more volatile investments that could contribute to worse performance. Meanwhile, the low efficiency of EBC paid to CEO in low growth opportunity firm could incur agency cost. In this case, company age and growth opportunity are the two main indicators to distinguish company type and affect M&A performance. Based on this intuition, this paper will pay attention to classifying companies by different levels of firm age and growth opportunity, then examining the effect of equity-based compensation fraction on M&A performance of each type.

(6)

4 This paper examines the effect of equity-based compensation on M&A performance and expect that equity-based compensation has a positive impact to young firms with high growth opportunity around announcement period and post-acquisition period. Increasing equity-based compensation fraction provides CEO with right incentives in M&A deals, this will be the first hypothesis which is then consistent with the “optimal contracting view” hypothesis. Furthermore, this paper splits companies into high and low firm age, high and low growth opportunity, expecting that the effect of equity-based compensation will be different. So the second hypothesis will be that the firm with high growth opportunity or low firm age will provide better M&A performance.

This paper analyses the data of American public, non-financial acquiring companies with completed deal status, because stock price information can only be obtained from public companies and only completed deals provide accurate and integrated market reaction to M&A. Data range is selected from 2003 to 2013 to examine the latest trend of effect, compared to prior studies. In regression analysis, M&A performance are measured by 62 days (-1,+60) Cumulative Abnormal Return in short run and three-year Buy-and-Hold Return in long run. Equity-based compensation fraction is calculated by the sum of restricted stock and stock option divided by the total compensation. Year effect and industry effect are controlled as well.

For acquiring firms, this paper finds that equity-based compensation paid to CEO provides incorrect incentives which leads to stock return decrease around and follow the acquisition decision announcement: 62 days (-1,+60) cumulative abnormal return decreases by 0.04 percent if increasing equity-based compensation fraction by 1 percent; Three-year buy-and-hold return decreases by 0.12 percent if increasing equity-based compensation fraction by 1 percent. Moreover, this paper also finds that equity-based compensation contributes to the worst short run performances in firms with high growth opportunities and worst long run

(7)

5 performances in firms with low growth opportunities. To test the robustness of the result and the reason of skewed incentive, the stock option fraction is examined to have negative effect on M&A short run and long run performance by splitting EBC fraction into restricted stock fraction and stock option fraction. By using samples before and after 2005, the significantly negative coefficient of EBC fraction before 2005 also indicates that firms grant wrong incentives in the form of stock option.

Prior studies concentrate on the effect of CEO compensation fraction of all firm types (Jensen Murphy 1990, Minnick et al. 2010 and Bebchuk and Fried 2003), some paper further examine the effect of CEO compensation, especially in equity-based compensation, on M&A performance (Shleifer and Vishny 1988, Cheng and Farber 2003). Recent papers tell the difference between high and low EBC firms (Datta et al. 2001). However either “optimal contracting view” or “managerial power view” is proved in these papers, the conclusion of dividing EBC into high and low level is still limited which also cannot identify a firm explicitly. The idea of “optimal level of incentives” is more like a common structure rather than an ideal compensation model. This paper differs from previous works in three primary ways. Firstly, the effect of equity-based compensation is examined by dividing acquiring companies into different firm age and growth opportunity levels. Secondly, I further explore the reason why the equity-based compensation has impact on M&A performance. Finally, firm characteristics and CEO characteristics that affect firm performance and CEO pay are controlled in the regression to make better prediction. It sheds light on how theboard of directors design the CEO compensation structure to align the manager interest and shareholder value to maximize the company wealth.

The remainder of this paper is organized as follows. Section 2 describes the literature review. Section 3 describes the data and descriptive statistics. The research methodology is explained

(8)

6 in detail in Section 4. Section 5 presents the empirical findings. Conclusions are drawn in Section 6.

II Literature Review

2.1 Agency theory

2.1.1 Agency theory definition

Jensen and Meckling (1976) define agency relationship explicitly as “a contract under which one or more principals engage another agent to perform some service on their behalf which involves delegating some decision making authority to the agent”. The agency problem from the view of the company is regarded as the divergence between manager’s decision and those decision that could maximize the shareholder’s welfare, which originated from the separation of ownership and control. The possibility that the manager pursuing private benefit incurs agency cost: the monitoring expenditure by the shareholder, the bonding expenditures by the agent and the reduction in dollar equivalent of shareholders.

2.1.2 Main reasons incur agency cost

There are several reasons that agency cost happens. Cai and Vijh (2007) document that CEOs with higher fraction of equity-based compensation are more likely to make acquisitions. Partially the board sets requirements to CEOs when CEOs want selling stock or option, for example, CEOs can sell stock only if they meet a certain amount of sales. In that situation, CEOs will stimulate short run performance and cash out by making merge and acquisition whether it is a good decision or not.

Similar to the reason stated above, Bliss and Rosen (2001) examine the relationship between mergers and CEO compensation during 1986–1995, they find CEO compensation

(9)

7 automatically increase after merge. Guest (2009) even finds that CEOs are equally rewarded even in a good or bad acquisition, though bad acquisition doesn’t increase CEO wealth because value of restricted stock and stock option holding goes down.

Another reason is associated with corporate control, Jensen and Ruback (1983) define “corporate control as the rights to determine the management of corporate resources—that is, the rights to hire, fire and set the compensation of top-level managers”. If the CEO was the chairman of the board, the lack of efficient monitoring could incur excessive compensation, empire-building and spending money on personal comfort. Dutta et al. (2011) suggest that more powerful CEOs are likely to make acquisitions, as measured by excessive compensation. When the wealth of a CEO is well-diversified, the risk aversion reduces the CEO incentives to take more risk than what would be optimal from the perspective of shareholders.

2.2 Optimal contracting view

2.2.1 Optimal contracting view argument

For years, many previous studies examine executive compensation as a solution to prevent loss of shareholder value from value-destroying decision made by the manager both empirically and theoretically (Murphy 1999). It is what optimal contracting view argues, the restricted stock and stock options provide executive with right incentives to maximum firm’s profit (Weisbach 2007).

The use of restricted stock and stock option increases due to the desire of decreasing natural managerial risk, because the value of a call option goes up in the underlying stock and volatility. Jensen and Murphy (1990) measure how executive stock option awards value affected by the change in firm’s value. They estimate the pay-performance relation of over 2,000 CEOs and find that CEO wealth changes $3.25 for every $1,000 change in shareholder

(10)

8 wealth. Even though the relation between compensation and firm performance is positive and statistically significant, the effect is smaller than their expectation. They also hypothesize that private and public political forces limit the pay-performance sensitivity. Based on the study of Jensen and Murphy (1990). Minnick et al. (2010) test how pay-for-performance sensitivity affects acquisition in the banking industry. They find banks whose CEO has higher pay-for-performance sensitivity contribute significantly better abnormal return around the announcement period, because higher pay-for-performance sensitivity encourages CEOs to make value-enhancing acquisition rather than value-destroying acquisition. However, the effect of pay-for-performance sensitivity doesn’t work in large banks, only in small and medium-sized banks.

2.2.2 Equity-based compensation

Previous studies also start to research the executive compensation structure. Normally, executive compensation includes salary, bonus, restricted stock, stock option, long term incentive plan and others. The part of restricted stock and stock option are commonly regarded as equity-based compensation.

Some researches examine the relation between equity-based compensation and firm performance by separating acquisitions into high and low EBC firms when the stock price response to the acquisition announcements is insignificant for the full sample. Datta et al. (2001) construct cumulative abnormal return around acquisition announcement and document a significantly positive relation between acquiring managers’ equity-based compensation fraction and firms’ stock performance. Besides, they expend existing studies by using three-year buy-and-hold return to detect long run performance following acquisition announcement. They conclude that managerial incentives effectively shape long run corporate investment policies and motivate CEO to make value-enhancing investment. Furthermore, they find that

(11)

9 high EBC firms experience significant positive stock price effect after announcement and pay lower acquisition premiums then low EBC firms. Datta et al. (2001) is also the most closely related paper to this paper. By appraising the acquisition process from the managerial view, Shleifer and Vishny (1988) also predict that equity-based compensation has the effect of reducing the non-value-maximizing behavior of acquiring managers. Smith and Stulz (1985) develop a theory of hedging behavior of value-maximizing firm and imply that shareholders can decline the likelihood of executives passing up valuable and risky investment by increasing the convexity of the relation between executives’ wealth and firm performance.

Restricted stock and stock option increase partially and consistently due to the public outrage, and there has been some criticism of large salary and bonus taken by executives, especially during financial crisis. The board of a firm decides to convert part of basic salary and bonus into equity-based compensation to avoid public outrage such as unemployment and media, while satisfying the executives’ requirements and incenting them to work as usual.

2.2.3 Optimal level of incentives

To study the optimal compensation structure, Core and Guay (1999) develop a well-acknowledged model for the optimal level of incentives and use the residuals between the estimated and reality compensation to predict deviations between CEO’s holding of equity incentive and optimal levels. The CEO equity incentives are defined as the change in the dollar value of the CEO’s stock and option for a 1% change in the stock. They find that companies use annual grants of option and restricted stock to manage CEOs’ optimal compensation structure. The weakness of this optimal level of incentives is that the regression model is based on a certain sample, it is more like a common used structure rather than optimal structure with utility testing. Even most firms grant their CEO compensation in this

(12)

10 way, no evidence can support that they pay the optimal one. Given that, this paper will assume that no optimal compensation structure exists.

2.3 Managerial power view

2.3.1 Managerial power view argument

A big debate is there, on whether executives could be incented by compensation. The managerial power view, a controversial view to the optimal contracting view, is developed present. The managerial power view suggests that restricted stock and stock option provide meaningless incentives but transfer rents to executives (Weisbach 2007). Bebchuk and Fried (2003) suggest that executive compensation is not only a tool for addressing the agency problem, but also can be a part of the agency problem itself. Bebchuk and Fried (2004) develop two views in a more accepted way by audience. They review extensive empirical findings that distinguish optimal contracting view and managerial power view and conclude that the findings support the notion that CEO has great control over their own compensation.

2.3.2 Equity-based compensation

Cheng and Farber (2008) investigate whether firms reduce the option-based compensation in their CEO new contract and if so, whether this action incurs improvement on firm performance. Surprisingly, they find an opposite effect of equity-based compensation on stock performance against previous literature. They find that the reduction in option-based compensation has come with a decrease in the riskiness of investment decision, which reflects in subsequent better operating performance and lower equity return volatility. The intuition behind this is, the reduction in option-based compensation rewards CEO lower value of option than when they engage into a value-enhancing investment. This directly reduces the willingness of CEO making risky investments.

(13)

11 Cai and Vijh (2007) examine relations between restricted stock and stock option holding and the premium paid to the target firm, they collect all firms during 1993 to 2001 and conclude that CEOs with higher holding or illiquidity discount are more likely to make acquisitions or get acquired. Acquirer CEOs with higher restricted stock and stock option pay a higher premium.

2.4 Explanation variables review

Besides the main effect of equity-based compensation on stock performance, Yermack (1995) uses Black-Scholes approach to study whether stock option’s incentives significantly relates to explanatory variables. He find those studies do not always agree to each other, in addition, their results are difficult to compare due to the differences in sample selection, methodology and time periods. For example, Jensen & Murphy (1990) imply that firm’s stock return is significantly related to the equity-based compensation, nevertheless Murphy (1985) find the relation is not significant. Lewellen et al. (1987) find that CEO’s age has positive effect on equity-based compensation, however Eaton and Rosen (1983) reject this notion. The effect of firm size on equity-based return is trickier, where four papers prove it is insignificant, two papers prove it to be significantly positive and one paper proves it to be significantly negative. Leverage has positive effect on equity-based return, examined in Lewellen et al. (1987). Loughran and Vijh (1997) research on whether long-term shareholders benefit from corporate acquisitions by testing the form of payment and mode of acquisition. They find firms using stock mergers earn about 86.7 percent which is significantly less excess return, than firms using cash tender offers.

(14)

12

III Data and Descriptive Statistics

3.1 Sample Construction

I construct M&A deal sample using the Thomson ONE Merge and Acquisition database. This paper includes samples that are announced between January 2003 and December 2013 and meet the following criteria:

(1) Both acquiring firm and target firms are U.S. firms which removes international effect. (2) Acquiring firms are public firms so that stock price change could be observed.

(3) Both acquiring firms and target firms are non-financial firms (excluding SIC code: 6000-6799) which removes the interruption of financial restriction on these firms.

(4) The acquisition has been completed.

(5) The deal value is between 0.1 million to 1,000 million to eliminate outlier influence. (6) The financial and executive compensation’s information are available in Compustat,

ExecuComp, Eventus and Center for Research in Security Prices(CRSP).

The chosen period of time assures that the latest trend of effect are investigated. The fourth criterion assures that the full effect of merge and acquisition will react with completed deals. Applying these criteria results in a total of 9,293 deals. The availability of executive compensation reduces sample to 3,049 deals. Moreover, the availability of company financial information reduces sample to 2,552 deals. In the sample construction for long run performance examination, Datta et al. (2001) only include the first merge and acquisition by a company during the study period to maintain the observations’ independence and exclude acquisitions completed in the last two years. This paper uses a new method for the overlapping sample within three-year research period that includes all deals and treat each merger as a separate observation, then cluster standard errors by the GVKEY(company ID)

(15)

13 firm identifier. By the availability of three-year BHR information in Eventus database and excluding deal announced between 2012 and 2013, the long run sample reduces to 2,070 deals .

Panel A of Table I presents descriptive statistics of 2,552 completed acquisitions between 2003 and 2013. The second column indicates that the amount of deal per year has increased from 232 in 2003 to 345 in 2007 which arrived at its peak time, then decreased from 2007 to 2009 during the most difficult nightmare period of financial crisis. It has recovered slowly since 2010. The reason of small amount in 2013 is that most of the deals are not completed, the findings in the table are finished within two years to present. Surprisingly, unlike the trend of the deal amount, the trend of average deal value looks steadier. The last column of Panel A shows that the average deal value has increased from 209.42 million in 2003 to 389.49 million in 2013, while a slightly decrease occurs during financial crisis. The aggregate’s average deal value is 333.38 million which is lower than what Datta et al. (2001) research. The difference can be explained by the different criterion.

3.2 Equity-based compensation fraction

I collect CEO compensation information using the ExecuComp database, including total compensation (TDC1), stock option (1992 format: OPTION_AWARDS_BLK_VALUE1992 format; 2006 format: OPTION_AWARDS_FV), restricted stock (1992 format: RSTKGRNT; 2006 format: STOCK_AWARDS_FV). Because the new FASB123 sets new reporting requirements in 2006, I use two variables to identify one item at separately. In addition to the compensation components, Annual CEO Flag and Executive’s ID number information are used to identify CEO among top executives. Datta et al. (2001) define equity-based compensation “as the Black-Scholes value of new options granted to the top five executives in the year preceding the acquisition divided by their total compensation”, while this paper

(16)

14 defines equity-based compensation fraction as the sum of value of stock option and the value of restricted stock divided by total compensation preceding the acquisition year, the formula is

Panel B documents the structure of CEO compensation in the year preceding the announcement year. The first two columns indicate that the firm grants more stock option of 2,576,770 dollars than restricted stock of 1,753,320 dollars to their CEOs on average, even twice larger if looking at median number. Total compensation is the sum of salary, bonus, restricted stock, stock option, long term incentive plan and others. Although the median total compensation is 3,809,950 dollars, the maximum compensation is 134,457,900 which indicates that firms grant their CEO compensation variously. The last column of Panel B shows that about 50 percent of the compensation are paid in the form of restricted stock and stock option. The median EBC fraction is slightly larger than mean number which is 55 percent. The result is still larger than Datta et al. (2001) which could be explained by the increasing trend of EBC fraction.

3.3 CEO control variables

According to the previous studies, this paper considers CEO gender and CEO tenure as control variables into the regression. CEO gender is defined as the dummy variable which equals to 1 if CEO is a male, 0 inversely. CEO tenure is calculated from the year of announcement less the year when he or she became CEO (year information is from DatebecameCEO).

Panel A of Table II presents statistics for independent variables and dependent variables. It reports that CEOs take merge and acquisition investment decision when they are in their sixth

(17)

15 year in the company on average. The reason could be that the decision made by CEO is fully supported by the company after the CEO gets familiar with the business. The result of CEO gender provides information that among the full 2,552 sample, only 2 percent which is 44 CEOs are female.

3.4 Firm control variables

Firm size is defined as the natural logarithm of the total market value of equity. Total market value of equity is calculated from total market value minus total liabilities (MKVALT-Market value total minus LT-Liabilities total) in the year prior to the announcement. Leverage is defined as Debt-to-Equity ratio (LT-Liabilities total divided by the difference between AT-Assets total and LT-Liabilities total). The firm age information are partially collected from Datastream database, and those cannot be found in the database are hand-collected by searching Google website, Wikipedia, Bloomberg BusinessWeek, Reuters website and firm website. This paper uses the market value of assets divided by the book value of assets preceding the announcement year as a proxy for growth opportunity (MKVALT-Market value divided by AT-Assets). Annual stock return for each deal are collected from CRSP.

Panel A of Table II further reveals the statistic of independent variables and dependent variables. The mean leverage is 7.07 which is extremely high. Datta et al. (2001) examine that the payment of merge and acquisition are equity normally. The median firm age is 28. A firm is categorized in the young firm group if its age is at or below the median, otherwise the firm is in the established group. The median growth opportunity is 1.45. A firm is categorized in the low growth opportunity group if its growth opportunity is at or below the median, otherwise the firm is in the high growth opportunity group.

(18)

16 Panel A of Table III partitions that firm age by the equity-based compensation level. A firm is regarded as a low EBC firm if its EBC fraction is at or below the median, otherwise the firm is in the High EBC group. This paper finds that firms with low equity-based compensation tend to have higher age (48 to 42) than firms with high equity-based compensation. Panel B of Table III finds that firms with low equity-based compensation have less growth opportunity than firms with high equity-based compensation. This result is consistent with my hypothesis that CEOs of established firms tend to accept low-risk compensation because they may have many job options, while young firms tend to grant their CEOs more restricted stock and stock option because they have rare cash generated in the beginning period of time and wish to incent them with the potential increase of equity-based compensation. Further, in firms with high growth opportunity, restricted stock and stock option are more valuable than cash-based compensation, CEOs would like to receive high fraction of restricted stock and stock option to maximize their total wealth after the firms fulfill the potential performance.

3.5 Correlation matrix

Panel B of Table II presents correlation matrix for short run variables, and Panel C of Table II presents correlation matrix for long run variables. Two tables show that most coefficient are less than 0.1 which indicates that there is no close correlative relationship which exists among short run or long run variables.

IV Methodology

4.1 Cumulative Abnormal Return

Like Datta et al. (2001) and many other papers, this paper also collects cumulative abnormal return from Cross-Sectional Daily Analysis of Eventus to appraise the event’s short run

(19)

17 performance. The abnormal return is defined as actual return less normal return. The normal return is the expected stock return estimated by market model without an announcement happening. The market model predicts a stable linear relation between the return of any given stock and the market return. For any stock , the formula is

Where is the period-t return on stock , is the market portfolio return which I choose CRSP value Weighted Index for market index as a common choice, is the zero mean

disturbance term. and are two parameters to be estimated.

The abnormal return and the cumulative abnormal return, AR and CAR, are calculated as

Where is the actual return on stock on day , is the normal return on stock on day estimated by market model without announcement. The estimation period is from 255 days to 46 days prior to the announcement date. The cumulative abnormal return is the sum of abnormal return from to . To test the trend of effect around the announcement date, two event windows, CAR(-1,+30) and CAR(-1,+60), are chosen to be observeed. The intuition behind this is that, I assume that the EBC fraction is not significantly related to CAR(-1,+30), but is positively significant and related to CAR(-1,+60). The reason could be market inefficiency, the stock price could not fully reveal firm value immediately and investors hesitate to make judgment to the merge and acquisition. The fear of inside trading reduces the stock purchased by investors than expected. The price goes up because the

(20)

18 uncertainty are resolved. People believe that CEO made non-value reducing deals to maximize shareholder’s wealth.

4.2 Buy-and-Hold Return

Many prior studies, e.g., Loughran and Vijh (1997), Spiess and Affleck-Graved (1999), use BHR approach to examine event’s long run performance. Moreover, Barber and Lyon (1997) examine the effectiveness between cumulative abnormal return and buy-and-hold abnormal return, where they conclude that CAR is a biased predictor of BHAR. Consequently, I favor the use of buy-and-hold return to detect long run performance in this paper. I collect daily buy-and-hold stock return data from CRSP. The buy-and-hold return, BHR, is calculated as

Where day t=1 is the first trading day following the announcement date, is the return on

stock on day , and is the one-year or three-year anniversary date of the announcement date. The reason to measure one-year BHR and three-year BHR at the same time is similar to short run performance, the trend of effect enables a specification change which could be observed in this paper. The reason why many papers mention using BHR instead of BHAR is because the selection of a benchmark is always problematic. Even though, the delisting problem of matching firm could be treated by choosing the next closest sample, the benchmark is still skewed. The effect on actual return is available to check in this paper, it provides more flexibility for firms to calculate abnormal return by testing various benchmark.

4.3 Multivariate Regression Analysis: Short run

This paper uses cross-sectional regression analysis to test whether the equity-based compensation paid to acquiring company CEO has any relation to the stock price change

(21)

19 around merge and acquisition decision announcement. The dependent variables includes three window times which are 3 days 1,+1) CAR, 32 days 1,+30) CAR and 62 days (-1,+60) CAR. The regression model are estimated as follows:

EBC fraction is defined as the sum of value of stock option and the value of restricted stock divided by total compensation preceding the acquisition. CEO gender is defined as the dummy variable which equals to 1 if CEO is a male, 0 inversely. CEO tenure is calculated from the year of announcement less the year when he or she became CEO (year information is from DatebecameCEO). Firm size is defined as the natural logarithm of total market value of equity preceding the announcement year. Firm Leverage is defined as Debt-to-Equity ratio preceding the announcement year. Firm age is defined as the year of announcement less the year the firm is established. Firm growth opportunity is defined as market value of assets divided by the book value of assets preceding the announcement year. This paper also includes industry dummies which are based on two-digit SIC codes to control fixed effect and year dummies which are based the announcement year to control time trends effect.

Firstly, three window times are chosen to test the trend of effect during different periods. Then this paper will pay much attention on 62 days (-1,+60) CAR, because most result from previous papers using 3 days (-1,+1) CAR or 32 days (-1,+60) CAR prove that equity-based compensation are positively related to the cumulative abnormal return around the announcement period. This paper provides an extra prospective to see if the equity-based compensation could affect cumulative abnormal return in sixty days, I also believe it is the best window time to examine the sustainable short run effect.

(22)

20 If the coefficient of EBC fraction is significantly positive, it means equity-based compensation including restricted stock and stock option is positively related to the merge and acquisition performance, increasing restricted stock and stock option in the total compensation can provides CEO with right incentives and maximum shareholder value in short run perspective. On the other hand, if the coefficient of EBC fraction is insignificant or significantly negative, it means that the equity-based compensation including restricted stock and stock option has negative or no effect on merge and acquisition performance in short run perspective, restricted stock and stock option cannot prevent agency cost but decrease shareholder value.

4.4 Multivariate Regression Analysis: Long run

As Datta et al. (2001) did, this paper uses cross-sectional regression analysis to test whether the equity-based compensation paid to acquiring company CEO has any relation to the post-event stock price change. The dependent variable, BHR, is defined as three-year buy-and-hold return after the announcement which is slightly different with Datta et al. (2001) (they use three-year buy-and-hold return following the effective date). The regression model is estimated as follows:

Where EBC fraction, CEO gender, CEO tenure, Firm size, Firm leverage, Firm age, Firm growth opportunity are as defined earlier. Annual return is defined as the stock return of the announcement year. Moreover, coefficients of firm growth opportunity and annual return are expected to be significantly positive. This paper also includes industry dummies which are based on two-digit SIC codes to control fixed effect and year dummies which are based on the announcement year to control time trends effect.

(23)

21 This paper also uses clustered standard errors (GVKEY) because the samples may not be independent but correlated across some known groups. For example, if a merge is announced by acquirer ABC in 2003, its three-year buy-and-hold return could be affected by merge if a second merge announced by acquirer ABC in 2004. This clustered procedure requires only that samples are independent in the clusters (GVEKY) and not the entire sample

The significantly positive coefficient of EBC fraction reveals that increasing restricted stock and stock option in the total compensation can provide CEO with right incentives and maximum shareholder value in long run perspective. On the other hand, the significantly negative or insignificant coefficient of EBC fraction means that restricted stock and stock option cannot prevent agency cost or even decrease shareholder value.

V Empirical Findings

5.1 Short run result

The main result from Panel A of Table IV shows that the coefficient of equity-based compensation is significantly negative in two months at 5 percent significance level(t-statistics = -2.40). This paper finds that increasing the portion of equity-based compensation in total compensation provides CEO with incorrect incentives in M&A deals in short run, which cumulative abnormal return decreases by 0.04 percent if increasing equity-based compensation fraction by 1 percent. The result is consistent with Cheng and Farber (2003) hypothesis that lower option-based compensation paid to CEO comes with a decrease in the riskiness of investment decision, which reflects in subsequent better operating performance and lower equity return volatility.

Aiming to check the trend of effect of equity-based compensation in short term, 32 days (-1,+30) CAR is examined in Model 1. Including control variables of CEO characteristics and

(24)

22 firm characteristics, the coefficient of equity-based compensation on CAR(-1,+30) is not significant (t-statistic = -1.35). This paper concludes that the stock price of acquiring firm starts to decrease due to CEO compensation structure after two months of the announcement date. The delay of market reacting, the opposite as expected before, could be explained by market inefficiency that the stock price could not fully reveal firm value immediately and investors hesitate to make judgment to the merge and acquisition. The fear of hiding information, probably inside trading or any other trap, reduces the stock purchased by investors than expected. The price goes down because the uncertainty are resolved, people believe that CEO made value-destroying deal to damage shareholder’s wealth.

The result is virtually the same when I control CEO characteristics in Model 3 and firm characteristics in Model 4. In the integrated regression of Model 5, besides equity-based compensation with a coefficient of -0.0324 (t-statistic = -2.40), the significantly negative coefficient of Leverage (t-statistic = -3.18) indicates that equity investors concern about overleveraged acquiring firm. The economic explanation of this concern is that an overleveraged firm needs to finance more money to acquire another company, the debt-to-equity ratio is higher than before the merge happens. The risk of making profit from new investment could lead a finance distress that acquiring firm cannot repay the loan interest on time. In the worst situation, the company goes bankrupt after acquisition because of high leverage. Size, another control variable is significantly negative in Model 4 which reveals that firm size is negatively related to the M&A performance. In other words, firms with small size make better performance than firms with big size in M&A deals. However in Model 5, the coefficient of Size becomes insignificant.

As shown in Model 5, CEO gender (t-statistic = 1.02) and CEO tenure (t-statistic = -0.27) are not significantly related to the acquisition’s short term performance. The possible explaining for the insignificance is that most CEOs are male (98 percent) so the difference could not be

(25)

23 observed completely and most CEOs have similar tenure (stand deviation = 6.97) so investors do not concern about his or her tenure. However, one expectation for CEO tenure could be that the longer the CEO tenure is, the worse the firm performs. Normally, experienced CEO has more control over the company which means that the CEO investment decision cannot be supervised efficiently.

To further study the effect of equity-based compensation on firm types, all samples are split by firm age and growth opportunity and the result is shown in Panel B and Panel C. The insignificant coefficients of low firm age (tstatistic = 1.6) and high firm age (tstatistic= -1.26) from Panel B reveal that the effect of equity-based compensation does not performss differently by firm age levels. Unlike firm age, the significantly negative coefficient (t statistic = -2.26) from Panel C indicates that firms with high growth opportunity perform worst, which cumulative abnormal return decreases by 0.03 percent if increasing equity-based compensation fraction by 1 percent. It indicates that the CEO in firm with more growth opportunity has more choices to make investment decision that in turn give more investment and leads to higher volatility, however it will incur a worse short run performance after acquisition. The coefficient of control variable Leverage is still significant in high growth opportunity. Moreover, the result also shows that the relation between equity-based compensation fraction and 62 days cumulative abnormal return in low growth opportunity firms is not significant (t-statistic = -1.00). This paper empirically reveals that equity-based compensation leads to the worst performance in firms with high growth opportunities.

5.2 Long run result

The main result from Panel A of Table V shows that the coefficient of equity-based compensation is significantly negative at 5 percent significance level (t-statistic = -2.17). This paper finds that equity-based compensation including restricted stock and stock option could

(26)

24 not aligns the CEO interest and shareholders’ wealth in M&A deals in long run, which buy-and-hold return decreases by 0.12 percent if increasing equity-based compensation fraction by 1 percent. This effect is larger than the effect in short run. The result is consistent with Cheng and Farber (2003) and Cai and Vijh (2007) hypotheses that acquirer CEOs with higher restricted stock and stock option pay a higher premium which lead to bad post-acquisition performance.

The result is virtually the same when I control CEO characteristics in Model 2 and firm characteristics in Model 3. Similar to short run regression result, coefficients of CEO gender (t-statistic = 1.58) and CEO tenure (t-statistic = 0.69) are not significant. This paper records that CEO characteristics are not the determiners of stock price change whether in short run or long run. Compared to the Datta et al. (2001), the mode of acquisition (Mergers or Tender offer) and the means of payment (Cash or Noncash) are more related to the stock price change in M&A deals. Unlike the result of integrated regression of short run, the coefficient of Size is significantly negative (t-statistic = -2.15) in all Models. It indicates that firms with small size perform better than firms with big size in the long run. Surprisingly, the coefficient of leverage became insignificant in the long run. One explanation for the different coefficients between short run and long run regression could be that investors worry about that acquiring firm goes into financial distress if the loan interest cannot be paid on time. However, debt-to-equity ratio became less important after three years that merge happens because the operating performance is shown to the public, the surviving company is proved to meet the loan interest instead of being bankrupt.

With respect to insignificant coefficient of firm age and growth opportunity, the effect of equity-based compensation fraction in the long run is examined by splitting firm types into firm age and growth opportunity levels. Same result with previous short run test reveals from Panel B that coefficients of EBC in high and low firm age levels are not significant. However

(27)

25 in Panel C, the coefficient of equity-based compensation in firms with low growth opportunity is significantly negative at 10 percent level (t-statistic = -1.78). This paper examines a slight difference between high growth opportunities in the short run and low growth opportunity in the long run, the difference could be explained that CEO has less chance to involve either value-enhancing investment or value-destroying investment in low growth opportunity. It is low efficiency to grant restricted stock and stock option to these CEOs, while agency cost happens more easily.

5.3 Robustness check

One possibility of the negative relation between equity-based compensation and M&A performance is that equity compensation leads to skewed incentives, in other words, owing equity is good generally, however, firms have been granting the wrong type of equity incentives. Perhaps stock options of equity-based compensation cause CEO to take much risky investment of merge and acquisition, for example, growth firms grant many options to CEO. To test the robustness of the above result and study the reason of skewed incentives, this paper splits EBC fraction into restricted stock fraction and stock option fraction. Moreover, because companies started to grant more restricted stock and less stock option in the past ten years (shown in Panel A of Table VI), they come up with an idea that the difference effect of restricted stock and stock option can be examined by splitting samples into before and after 2005.

In Table VI, Model 1 examines the effect of restricted stock option and stock option on M&A performance, respectively. The result indicates that stock option fraction has a significant negative effect on M&A short and long run performance at 5 percent significance level. In Panel A, 62 days cumulative abnormal return decreases by 0.0328 percent if stock option fraction increases by 1 percent (t-statistic = -2.13) which is even larger than the combined

(28)

26 equity-based compensation (-0.0324), while the restricted stock fraction is only significantly related to the cumulative abnormal return at 10 percent significance level (t-statistic = 1.84). In Panel B, three-year buy-and-hold return decreases by 0.13 percent if stock option fraction increases by 1 percent (t-statistics = -2.11), while the coefficient of restricted stock option is not significant. This paper examines the reason of skewed incentives that stock option in the EBC paid to CEO causes the performance to decline and suggests that firms should grant less options in their CEO compensation package.

Model 2 examines the effect of equity-based compensation before 2005 and Model 3 examines the effect of equity-based compensation after 2005. Consistent with former prediction, the coefficient of equity-based compensation before 2005 is significantly negative at 10 percent significance level (t-statistic = -1.92) in the short run that 62 days cumulative abnormal return decreases by 0.04 when increasing EBC fraction by 1 percent, while the coefficient of EBC fraction after 2005 is not significant(t-statistic = 0.88). In a long run test, the coefficient of EBC fraction before 2005 is significantly negative at 5 percent significance level (t-statistic = -2.41) that three-year buy-and-hold return decreases by 0.19 percent if equity-based compensation fraction increases by 1 percent, while the coefficient of EBC after 2005 is not. The results from Model 2 and Model 3 indicate that firms with more stock option in their CEO compensation perform worse in M&A deals. Both results from Panel A and Panel B support the idea that companies grant skewed incentives in the form of stock option.

VI Conclusion and Discussion

Using a sample of 2,552 merge and acquisition made within American firms during the period 2003 to 2013, this paper documents a significantly negative relation between equity-based compensation paid to acquiring firms’ CEO and stock price change around and

(29)

27 following the acquisition decision announcement. The 62 days cumulative abnormal return decreases by 0.03 percent if equity-based compensation fraction increases 1 percent. The three-year buy-and-hold return decreases by 0.11 percent if equity-based compensation fraction increases 1 percent. The negative relation, which is consistent with Cheng and Farber (2008), could be explained by that lower option-based compensation paid to CEO comes with a decrease in the riskiness of investment decision, which reflects in subsequent better operating performance and lower equity return volatility.

This paper further examines the effect on firm types by splitting samples into high and low firm age, high and low growth opportunity. The insignificant coefficient of equity-based compensation either in low firm age or high firm age indicates that the effect of equity-based compensation on M&A deals does not perform differently by firm age level. On the other hand, the equity-based compensation fraction leads to the worst M&A short run performance in firms with high growth opportunities. It reveals that a CEO in a firm with more growth opportunity has more choices to make investment decision which leads to higher investment volatility and worse short run performance around the acquisition announcement. However the coefficient of equity-based compensation is significantly negative in firms with low growth opportunity in the long run. When it comes to the insufficient efficiency of granting restricted stock and stock option to low growth opportunity firms, CEO incurs agency cost in the long run easily.

This paper also test the reason of skewed incentives in two ways: dividing equity-based compensation fraction into restricted stock fraction and stock option fraction; splitting samples into before and after 2005. The significantly negative relation between stock option fraction and M&A performance and the significantly negative relation between EBC fraction and M&A performance suggest that firms should grant more restricted stock rather than stock option in a CEO compensation package.

(30)

28 Nevertheless, the result can still be improved by looking at more omitted variables, for example, three-year buy-and-hold return includes annual return of the announcement year as an additional control variable. The significantly positive coefficient (t-statistic = 24.61) suggests that firms with better performance contribute a good post-acquisition performance. Further studies can research on the effect of stock return on equity-based compensation because there is a possibility that good stock return increases the value of equity-based compensation hold by CEO.

This paper improves the study of the effect of equity-based compensation on M&A performance creatively. By examining the trend of short run cumulative abnormal return, this paper conclude a market inefficiency that lead delayed reaction to the announcement. Moreover, a new perspective of the efficiency of equity-based compensation is put forward by splitting samples into firm age and growth opportunity. Finally, the reason of skewed incentive is examined that a firm grants wrong type of compensation.

(31)

29

References

Bebchuk, Lucian A. and Fried, Jesse M., 2003, Executive compensation as an agency problem, Journal of Economic Perspectives, Vol. 17, Number 3, Pages 71–92.

Bebchuk, Lucian A. and Fried, Jesse M., 2004, Pay without performance, the unfulfilled promise of executive compensation, Part II: Power and Pay, Harvard University Press. Bliss, Richard T. and Rosen, Richard J., 2001, CEO compensation and bank mergers. Journal

of Financial Economics, 61 (2001) 107–138.

Cai, Jie and Vijh, Anand M., 2007, Incentive effect of stock and option holdings of target and acquirer CEOs, The Journal of Finance, Vol. LXII, No. 4. 1891-1933

Cheng, Qiang and Farber, David B., 2008, Earning restatement, changes in CEO compensation, and firm performance, The Accounting Review, Vol. 83, No. 5. 1217-1250

Core, John and Guay, Wayne, 1999, The use of equity grants to manage optimal equity incentive levels, Journal of Accounting and Economics, 28 (1999) 151-184

Dataa, Sudip, Iskandar-Datta, Mai and Raman, Kartik, 2001, Executive compensation and corporate acquisition decisions, The Journal of finance, Vol. LVI, No.6

Duttaa, Shantanu, MacAulay, Kenneth and Saadi, Samir, 2011, CEO power, M&A decisions, and market reactions, Journal of Multinational Financial Management, 21(5), 257-278 Gozalez, Tanja Artiga and Calluzzo, Paul, 2013, Coordinated shareholder activism

Guest, Paul M., 2009, The impact of mergers and acquisitions on executive pay in the united kingdom, Economica, Vol. 76, Issue 301, pp149-175

(32)

30 Hall, Brian J. and Murphy, Kevin J., 2003, The trouble with stock options, Journal of

Economic Perspectives, Volume 17, Number 3, Summer 2003, Pages 49–70

Jensen, Michael C. and Meckling, William H., 1976, Theory of the firm: managerial behavior, agency costs and ownership structure, Journal of Financial Economics, October, V. 3, No. 4, pp. 305-360.

Jensen, Michael C. and Murphy, Kevin J., 1990, Performance pay and top-management incentives, Journal of Political Economy, Vol. 98, No. 2. 225-264

Jensen, Michael C. and Ruback, Richard, 1983, The market for corporate control: the scientific evidence, Journal of Financial Economics, 11, 5-50

Lewellen, Wilbur G., Loderer, Claudio and Martin, Kenneth, 1987, Executive compensation and executive incentive problems: An empirical analysis, Journal of Accounting and

Economics 9, 287-310.

Loughran, Timothy, and Vijh, Anand, 1997, Do long-term shareholders benefit from corporate acquisitions? Journal of Finance 52, 1765-1790

Mahoney, Lois Schafer and Thorn, Linda, 2006, An examination of the structure of executive compensation and corporate social responsibility: A Canadian investigation, Journal of

Business Ethics, (2006) 69:149-162

Minnick, Kristina, Unal, Haluk and Yang, Liu, 2010, Pay for performance? CEO compensation and acquirer returns in BHCs

Murphy, Kevin J., 1985, Corporate performance and managerial remuneration: An empirical analysis, Journal of Accounting and Economics 7, 1142.

(33)

31 Murphy, Kevin J., 1999, Executive compensation, Handbook of Labor Economics, Volume 3,

Chapter 38, 2485-2563.

Shleifer, Andrei, and Vishny, Robert, 1988, Value maximization and the acquisition process,

Journal of Economic Perspective 2, 7-20

Smith, Clifford, and Stulz, Rene, 1985, The determinants of firms’ hedging policies, Journal

of Financial and Quantitative Analysis 20, 391-406.

Spiess, D. Katherine, and Affleck-Graves, John, 1999, The long-run performance of stock returns following debt offerings, Journal of Financial Economics 54, 45-73

Weisbach, Michael S., 2007, Optimal executive compensation versus managerial power: A Review of Lucian Bebchuk and Jesse Fried’s Pay without Performance: The Unfulfilled Promise of Executive Compensation, Journal of Economic Literature, Vol. XLV (June 2007), pp.419–428

Yermack, David, 1995, Do corporations award CEO stock and options effectively?, Journal

(34)

32

Table I

Deal and CEO compensation statistics

Panel A of Table I presents descriptive statistics of 2,552 completed acquisitions between 2003 and 2013. I collect CEO compensation information using the ExecuComp database, including total compensation (TDC1), stock option (1992 format: OPTION_AWARDS_BLK_VALUE1992 format; 2006 format: OPTION_AWARDS_FV), restricted stock (1992 format: RSTKGRNT; 2006 format: STOCK_AWARDS_FV). This paper defines equity-based compensation fraction as the sum of value of stock option and the value of restricted stock divided by total compensation preceding the acquisition year.

Panel A: Distribution by year

Year Number of M&A % of Sample Avg. Deal Value ($ Millions) 2003 232 9.09 209.42 2004 290 11.36 292.85 2005 308 12.07 398.64 2006 303 11.87 254.58 2007 345 13.52 368.45 2008 243 9.52 345.58 2009 170 6.66 379.47 2010 186 7.29 397.89 2011 214 8.39 312.34 2012 205 8.03 385.88 2013 56 2.19 389.49 Total 2552 100.00 333.38

Panel B: Compensation to CEO ($000)

Mean Median

Standard

Deviation Minimum Maximum Restricted stock 1753.32 382.42 3563.78 0.00 30900.18

Stock option 2576.77 774.20 6975.18 0.00 119458.80

EBC 4330.09 1959.77 8102.44 0.00 119458.80

Total compensation 6852.34 3809.85 10164.25 0.00 134457.90

(35)

33

Table II

Independent variables and dependent variables statistics

This paper defines equity-based compensation fraction as the sum of value of stock option and the value of restricted stock divided by total compensation preceding the acquisition year. CEO gender is defined as the dummy variable which equals to 1 if CEO is a male, 0 inversely. CEO tenure is calculated from the year of announcement less the year when he or she became CEO (year information is from DatebecameCEO). Firm size is defined as the natural logarithm of total market value of equity. Total market value of equity is calculated from total market value minus total liabilities (MKVALT-Market value total minus LT-Liabilities total) in the year prior to the announcement. Leverage is defined as Debt-to-Equity ratio (LT-Liabilities total divided by the difference between AT-Assets total and LT-Liabilities total). Panel A of Table II further reveals the statistic of independent variables and dependent variables. Panel B of Table II presents correlation matrix for short run and long run variables, and Panel C of Table II presents correlation matrix for long run variables.

Panel A : Independent variables and dependent variables

Mean Median

Standard

Deviation Minimum Maximum Observation EBC fraction (%) 49.59 54.95 27.36 0.00 100.00 2552 CEO tenure 8.31 6.00 6.97 0.00 62.00 2552 Size 7.35 7.07 1.83 1.61 12.29 2552 Leverage 0.90 0.75 3.52 -52.28 85.59 2552 Firm age 45 28 40 1 252 2552 Growth opportunity 1.84 1.45 1.34 0.38 15.53 2552 Annual return 0.1435 0.0915 0.4477 -0.9313 4.1601 2070 CAR(-1,+1) 0.0058 0.0040 0.0525 -0.4150 0.3120 2552 CAR(-1,+30) 0.0008 0.0040 0.1209 -0.8080 0.5180 2552 CAR(-1,+60) -0.0084 -0.0040 0.1743 0.1743 0.7910 2552 Three-year BHR 0.3004 0.1316 0.8295 -0.9650 8.2231 2070

(36)

34 Panel B: Correlation matrix for short run variables, N=2552

EBC fraction CEO gender CEO tenure Size Leverage Firm age Growth opportunity

EBC fraction 1 CEO gender 0.0117 1 CEO tenure -0.0969 0.0231 1 Size 0.2574 0.0331 -0.0121 1 Leverage -0.0154 0.0163 -0.0097 -0.0228 1 Firm age -0.0467 0.0190 -0.1313 0.1159 0.0267 1 Growth opportunity 0.1464 -0.0371 -0.0318 0.3896 -0.0503 -0.1647 1 Panel C: Correlation matrix for long run variables, N=2070

EBC fraction CEO gender CEO tenure Size Leverage Firm age Growth opportunity Annual return

EBC fraction 1 CEO gender 0.0134 1 CEO tenure -0.0723 0.0217 1 Size 0.2495 0.0330 -0.0203 1 Leverage -0.0219 0.0223 -0.0052 -0.0276 1 Firm age -0.0472 0.028 -0.1477 0.1208 0.0204 1 Growth opportunity 0.1434 -0.0509 -0.0331 0.3904 -0.0522 -0.1684 1 Annual return 0.0053 0.0197 -0.0007 -0.1031 -0.0279 -0.0409 -0.0539 1

(37)

35

Table III

Firm Age and Growth opportunity

Panel A of Table III partitions that firm age by equity-based compensation levels. Panel B of Table III partitions that growth opportunity by equity-based compensation levels.

Full sample Low EBC High EBC

Panel A: Firm age

Mean 45 48 42

Median 28 33 25

Observation 2552 1276 1276

Panel B: Growth opportunity

Mean 1.84 1.62 2.07

Median 1.45 1.26 1.64

(38)

36

Table IV

Multiple Regression on Cumulative Abnormal Return

The sample consists of 2,552 completed acquisition from 2003 to 2013. The abnormal return is defined as actual return less normal return. The normal return is the expected stock return estimated by market model without announcement happens. EBC fraction is defined as the sum of value of stock option and the value of restricted stock divided by total compensation preceding the acquisition. CEO gender is defined as the dummy variable which equals to 1 if CEO is a male, 0 inversely. CEO tenure is calculated from the year of announcement less the year when he or she became CEO (year information is from DatebecameCEO). Firm size is defined as the natural logarithm of total market value of equity preceding the announcement year. Firm Leverage is defined as Debt-to-Equity ratio preceding the announcement year. Firm age is defined as the year of announcement less the year the firm is established. Firm growth opportunity is defined as market value of assets divided by the book value of assets preceding the announcement year. This paper also includes industry dummies which are based on two-digit SIC codes to control fixed effect and year dummies which are based on the announcement year to control time trends effect. *,**, and *** indicate significance at 10%, 5% and 1%, respectively.

Panel A: Multiple regression on cumulative abnormal return

Model 1 Model 2 Model 3 Model 4 Model 5

CAR (-1,+30) CAR (-1,+60) CAR (-1,+60) CAR (-1,+60) CAR (-1,+60) Intercept 0.0303 -0.0089 -0.0359 -0.0012 -0.0171 (0.71) (-0.17) (-0.61) (-0.02) (-0.28) EBC fraction -0.0126 -0.0403*** -0.0410*** -0.0356*** -0.0324** (-1.35) (-3.08) (-3.11) (-2.64) (-2.40) CEO gender 0.0112 0.0282 0.0313 0.0278 (0.60) (1.03) (1.15) (1.02) CEO tenure -0.0002 -0.0002 -0.0002 -0.0001 (-0.50) (-0.45) (-0.45) (-0.27) Size -0.0016 -0.0043** -0.0023 (-0.98) (-2.07) (-1.03) Leverage -0.0026*** -0.0031*** -0.0031*** (-3.84) (-3.14) (-3.18) Firm age 0.0001 0.0001 (1.38) (1.28) Growth opportunity -0.0063*** -0.0083*** (-2.92) (-2.67) R2-adjusted 0.0189 0.0061 0.0058 0.0105 0.0139 Observation 2552 2552 2552 2552 2552

(39)

37 Panel B: Multiple regression on 62 days (-1,60) CAR by firm age level

Low firm age High firm age

Intercept 0.4631** 0.0187 (2.21) (0.30) EBC fraction -0.0318 -0.0239 (-1.60) (-1.26) CEO gender 0.0286 0.0210 (0.72) (0.56) CEO tenure -0.0011 0.0001 (-1.19) (0.19) Size 0.0005 -0.0046 (0.13) (-1.56) Leverage -0.0042** -0.0020* (-2.57) (-1.66) Growth opportunity -0.0129*** -0.0003 (-2.98) (-0.05) R2-adjusted 0.0180 0.0102 Observation 1282 1270

Panel C: Multiple regression on 62 days (-1,60) CAR by growth opportunity level

Low growth opportunity High growth opportunity

Intercept -0.0556 -0.0490 (-0.59) (-0.56) EBC fraction -0.0199 -0.0434** (-1.00) (-2.26) CEO gender 0.0803** 0.0019 (2.19) (0.05) CEO tenure -0.0004 0.0002 (-0.50) (0.25) Size -0.0090*** 0.0025 (-2.74) (0.76) Leverage -0.0037 -0.0036*** (-0.74) (-3.32) Firm age 0.0001 0.0003 (0.54) (1.60) R2-adjusted 0.0144 0.0063 Observation 1276 1276

(40)

38

Table V

Multiple Regression on Three-year BHR

The sample consists of 2,070 completed acquisition from 2003 to 2013. Where EBC fraction, CEO gender, CEO tenure, Firm size, Firm leverage, Firm age, Firm growth opportunity are as defined earlier. Annual return is defined as the stock return of the announcement year. Moreover, coefficients of firm growth opportunity and annual return are expected to be significantly positive. This paper also includes industry dummies which are based on two-digit SIC codes to control fixed effect and year dummies which are based on the announcement year to control time trends effect. *,**, and *** indicate significance at 10%, 5% and 1%, respectively.

Panel A: Multiple regression on three-year BHR

Model 1 Model 2 Model 3 Model 4 Model 5

BHR3 BHR3 BHR3 BHR3 BHR3 Intercept 2.3558*** 2.1372*** 2.4717*** 2.4903*** 1.7841*** (8.27) (6.79) (7.70) (7.69) (6.26) EBC fraction -0.1448** -0.1417** -0.0815 -0.0849 -0.1176** (-2.40) (-2.34) (-1.32) (-1.37) (-2.17) CEO gender 0.2081 0.2371* 0.2409 0.1562 (1.58) (1.81) (0.83) (1.36) CEO tenure 0.0017 0.0017 0.0017 0.0032 (0.69) (0.70) (0.67) (1.45) Size -0.0450*** -0.0477*** -0.0200** (-4.66) (-4.52) (-2.15) Leverage -0.0050 -0.0050 -0.0008 (-1.18) (-1.17) (-0.22) Firm age -0.0001 0.0001 (-0.24) (0.19) Growth opportunity 0.0102 0.0092 (0.74) (0.76) Annual return 0.9233*** (24.61) R2-adjusted 0.2141 0.2145 0.2225 0.222 0.4026 Observation 2070 2070 2070 2070 2070

(41)

39 Panel B: Multiple regression on three-year BHR by firm age level

Low firm age High firm age

Intercept -0.1335 1.8145*** (-0.17) (6.48) EBC fraction -0.1136 -0.0815 (-1.41) (-1.12) CEO gender 0.1888 0.1176 (1.19) (0.70) CEO tenure 0.0088** -0.0027 (2.39) (-1.06) Size -0.0124 -0.0210* (-0.82) (-1.82) Leverage -0.0021 -0.0022 (-0.33) (-0.51) Growth opportunity 0.0179 -0.0246 (1.06) (-1.35) Annual return 0.7824*** 1.2580*** (14.97) (22.67) R2-adjusted 0.3559 0.5302 Observation 1063 1007

Panel C: Multiple regression on three-year BHR by growth opportunity level

Low growth opportunity High growth opportunity

Intercept 2.4519*** 0.7561 (6.98) (1.64) EBC fraction -0.1410* -0.1076 (-1.78) (-1.42) CEO gender 0.1714 0.0967 (1.10) (0.56) CEO tenure 0.0008 0.0044 (0.27) (1.38) Size 0.0089 -0.0336** (0.65) (-2.57) Leverage 0.0094 -0.0026 (0.50) (-0.65) Firm age 0.0008 -0.0005 (1.41) (-0.72) Annual return 1.0202*** 0.7786*** (21.06) (13.35) R2-adjusted 0.5242 0.3074 Observation 997 1073

(42)

40

Table VI Robustness check

In Table VI, Model 1 examines the effect of restricted stock option and stock option on M&A performance, respectively. The Model 2 examines the effect of equity-based compensation before 2005 and Model 3 examines the effect of equity-based compensation after 2005.

Panel A: Mean fraction Distribution by year N=2,552

Year Restricted stock fraction Stock option fraction

2003 6.81% 43.42% 2004 10.30% 35.85% 2005 18.18% 20.76% 2006 21.47% 20.54% 2007 23.24% 22.43% 2008 28.31% 21.34% 2009 30.36% 16.88% 2010 26.82% 22.44% 2011 32.70% 20.33% 2012 33.63% 19.31% 2013 33.60% 15.58% Panel B: CAR(-1,+60)

Model 1 Model 2 Model 3

Full sample Before 2005 After 2005

Intercept -0.0168 -0.0830 -0.0089 (-0.27) (-0.83) (-0.11) EBC fraction -0.0414* -0.0156 (-1.92) (0.88) Stock fraction -0.0317* (-1.84) Option fraction -0.0328** (-2.13) CEO gender 0.0278 0.0508 0.0260 (1.02) (0.99) (0.79) CEO tenure -0.0001 0.0004 -0.0005 (-0.27) (0.44) (-0.77) Size -0.0023 -0.0033 -0.0022 (-1.03) (-0.79) (-0.78) Leverage -0.0031*** -0.0024 -0.0035*** (-3.17) (-1.04) (-3.13) Firm age 0.0001 0.0003* 0.0001 (1.28) (1.65) (0.51) Growth opportunity -0.0082*** -0.0096** -0.0068 (-2.64) (-2.03) (-1.62) R2-adjusted 0.0135 0.0299 0.0176 Observation 2552 830 1722

(43)

41 Panel C: Three-year BHR

Model 1 Model 2 Model 3

Full sample Before 2005 After 2005

Intercept 1.7919*** 2.2646*** 0.5624 (6.27) (5.98) (1.37) EBC fraction -0.1917** 0.0210 (-2.41) (0.30) Stock fraction -0.0980 (-1.39) Option fraction -0.1307** (-2.11) CEO gender 0.1554 0.2374 0.1218 (1.35) (1.32) (0.86) CEO tenure 0.0033 0.0051 0.0044 (1.48) (1.49) (1.60) Size -0.0199** 0.0287* -0.0418*** (-2.14) (1.89) (-3.66) Leverage -0.0008 0.0039 -0.0040 (-0.21) (0.49) (-0.97) Firm age 0.0001 0.0015** 0.0001 (0.15) (2.09) (0.25) Growth opportunity 0.0099 -0.0198 0.0360** (0.82) (-1.18) (2.21) Annual Return 0.9242*** 0.9211*** 0.9278*** (24.59) (14.71) (20.12) R2-adjusted 0.4024 0.4527 0.4184 Observation 2070 726 1344

The effect of CEO compensation structure on M&A performance : evidence from American acquiring companies