Equity incentives, block ownership and returns to mergers

(1)

Equity Incentives, Block Ownership and Returns to Mergers

Tony Chau (10555269)

Abstract:

In mergers and acquisitions, the conflict of interest between shareholders and company executives often contributes to a lower return for the acquiring company (Jensen & Meckling, 1976). Although past research has proposed solutions including equity based compensation and blockholder ownership, this study finds that they correlate negatively with M&A returns when applied separately, possibly because executives are incentivized by equity based compensation to undertake excessive risk when making investments, while blockholder ownership demotivates executives from seeking new investment opportunities for fear of shareholder intervention. However, their combined effect limit the respective negative impact on M&A returns, since blockholders could then effectively intervene when excessive risk is undertaken, whereas demotivated executives would be incentivized with equity based compensation to seek new value-generating investments. While this study does show a positive effect of applying both solutions simultaneously on M&A returns, it is still insufficient to overcome the negative impact arising from excessive risk-taking behavior induced by equity based compensation.

Keywords: mergers and acquisitions, equity based compensation, blockholder ownership, conflict of interest, shareholder intervention

(2)

2

Master thesis Supervisor: Florian Peters

MSc Finance, Corporate Finance track Date: June 2018

Statement of originality

This document is written by Student Tony Chau who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document are 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.

(3)

3

Acknowledgements

First and foremost, I would like to express my sincere gratitude to my supervisor Dr. Florian Peters for his guidance, understanding, and advice.

Special thanks to my friends at the University of Amsterdam, who have been most generous in lending their help and offering insightful comments to facilitate the development of this thesis.

Last but not the least, my heartfelt thanks to my mother for her constant encouragement, and my loving girlfriend Ruby for her unfailing support every step of the way.

(4)

4

Shareholder value can be influenced by a variety of corporate investment decisions. Mergers and acquisitions (M&A), in particular, are especially significant since these high-profile decisions have a far greater impact on shareholder value than most other relatively smaller scale investments. It has been shown in various research studies, however, that mergers and acquisitions in general undermine shareholder value. Over a 20-year period from 1980 to 2001, M&A decisions have shown a negative impact on share price (Moeller et al., 2004). Similarly, earlier studies have also suggested negative price drifts following acquisitions (Jensen & Ruback, 1983; Agrawal et al., 1992).

A possible explanation to these observations relates to potential conflicts of interest between shareholders and company executives (Jensen & Meckling, 1976). This is because executives are incentivized to grow the company beyond its optimal size due to the prestige in managing a larger company (Jensen, 1986), and the higher compensation that comes with business growth (Murphy, 1985).

Equity based compensation (EBC) can potentially remediate this conflict by providing executives with greater incentives to act in shareholders’ interests through creating a direct link between compensation and share performance. On the flip side, this can also induce managers to take on excessive risk in making investment decisions (Humphery-Jenner et al., 2016).

On the other hand, blockholder ownership enables more effective monitoring of business activities to prevent company executives from pursuing acquisitions that are detrimental to the company’s value. However, this can also demotivate executives from seeking new and potentially profitable investment opportunities for fear of excessive shareholder intervention (Burkart et al., 1997).

In exploring the effectiveness of these potential solutions to the agency problem, past research has mainly focused on the unilateral relationships of equity based compensation and blockholder ownership on shareholder value respectively. There has, however, only been limited literature on their combined effect, which can potentially mitigate their respective disadvantages when applied separately. Although equity based compensation can incentivise executives to become excessively risk-seeking, effective monitoring by blockholders could deter executives from proceeding with sub-optimal investments. At the same time, the negative effects of lower managerial discretion due to blockholder ownership may be alleviated by the application of equity based compensation as an incentive.

Considering the potential complementary effects of the two measures, this thesis aims to address the research question on how the combined effect of equity based compensation and blockholder ownership may impact shareholder return in acquisitions.

(6)

two-6

step procedure is employed to determine this combined effect. First, by the use of an event study, the level of abnormal return around the time of M&A announcement can be estimated, after which it can be used as a dependent variable to conduct a multivariate analysis. The dataset under study is retrieved from a range of databases including Thomson One, Compustat, Execucomp, Center for Research in Security Prices (CRSP), Institutional Shareholder services (ISS), and the Hoberg-Philips Data Library, to acquire a sample comprising a total of 252 transactions over the period 2006 – 2016.

From the results, it is found that equity based compensation correlates negatively with M&A returns, possibly because equity based compensation induces executive management to undertake excessive risk and hence to pursue sub-optimal investment opportunities. (Malmendier & Tate, 2005, 2008, 2009; Humphery-Jenner et al., 2016). On the other hand, a positive relationship is identified between executive ownership and M&A returns, suggesting that share ownership disincentivises executives from pursuing risky investments, since it effectively aligns the interests of executives with that of shareholders.

On the contrary, no significant relationship is found between blockholder ownership and M&A returns. Prior research has suggested several reasons for this observation. In Li, Liu, & Wu (2017), it is found that some CEOs were more likely to finance acquisitions with cash rather than equity in order to bypass shareholder voting and hence avoid shareholder intervention. Another reason relates to managerial discretion. While shareholder intervention could deter executives from proceeding with value-destroying mergers or acquisitions, this can also demotivate executives from seeking new investment opportunities for fear of intervention (Burkart et al., 1997).

Regarding the main variable of the study, the interaction term between blockholder ownership and equity based compensation showed a positive and significant result, suggesting a positive combined effect of blockholder ownership and equity based compensation on M&A returns. However, results also showed that this positive combined effect is insufficient to cover the negative impact of equity based compensation.

Through investigating the separate and combined effects of equity based compensation and blockholder ownership on M&A performance, this thesis provides better insights into the effectiveness of simultaneously implementing both measures to alleviate the agency problem between company executives and shareholders. This study also produces insights on how shareholders should incentivize their executives. Rather than employing equity based compensation, shareholders should ensure that executives hold a substantial ownership position to align the interests of shareholders and company executives, which will prevent executives from being incentivised to pursue sub-optimal M&A investments. This is especially important for companies that adopt a buy and build strategy where mergers and acquisitions play an important role in their business growth (e.g. Refresco).

(7)

7

and defines the hypotheses. Section 3 discusses the regression models and variables employed in this study. Section 4 describes the data sample and descriptive statistics. Section 5 reports the empirical results. Section 6 concludes.

2. Literature review and hypotheses

This section discusses existing literature regarding the effect of equity based compensation and blockholder ownership on M&A and firm performance, and is split into four parts: (i) the sole effect of equity compensation on performance, (ii) the sole effect of blockholder ownership on performance, (iii) the combined effect of both factors on company performance, and (iv) the influence of transaction and firm characteristics on M&A returns. In addition, four hypotheses are developed in the first three subsections.

2.1 Equity compensation

In a corporate context, especially in larger scale companies, it is not surprising to find executives making company decisions with their personal interests in mind. Eventually, these actions come at the cost of shareholders and are termed “agency conflicts” (Jensen & Meckling, 1976). A typical example of an agency conflict could be illustrated in the acquisition of another company, which would increase the power and size of the acquiring company, making it appear more prestigious to manage (empire building), despite not being necessarily value-enhancing in the long run.

A possible solution to the many challenges brought about by agency conflicts could be equity based compensation, which encourages managers to act in the interest of shareholders. For example, equity based compensation incentivises executives to pursue riskier projects with long term payoffs (Guay, 1999; Datta et al., 2001; Williams & Rao, 2006; Sudarsanam, 2017). This is related to the fact that executives face firm-specific risk given that their human capital depends heavily on the company, whereas investors (shareholders) are in a better position to diversify the risks (Feito-Ruiz & Renneboog, 2017). As a result, equity based compensation would discourage executives from pursuing corporate decisions with risk levels lower than what is optimal, and instead encourage them to pursue risky but value-enhancing projects (Smith and Stulz, 1985).

The benefits of equity based compensation extends also to the context of mergers and acquisitions. Using 485 company acquisitions that occurred between 1993 – 1996, Datta et al. (2001) established a positive relationship between equity based compensation and stock price performance around the time of acquisitions announcement. In this study, it is found that firms with a higher level of equity based compensation experienced positive stock price returns whereas firms with lower

(8)

8

equity based compensation experienced significant losses. The research also suggests that firms with higher equity based compensation in general offered lower premiums for their acquisitions and made riskier investments.

Hypothesis 1:

Equity based compensation for the executive management of the acquiring company increases acquirers’ shareholder value in a takeover

While equity based compensation addresses issues associated with agency conflict, excessive equity based compensation could prove to be counterproductive, since this can induce management to become overly risk-seeking. This negative effect becomes even stronger for overconfident CEOs (Humphery-Jenner et al., 2016; Cooper et al., 2016), who are more likely to accept highly convex compensation contracts (such as options) (Gervais et al., 2011). This is because the convex payoff structure of stock options incentivises managers to take more risk. This may lead to overinvestment and in turn give rise to mergers and acquisitions that undermine company value (Malmendier & Tate, 2005, 2008, 2009).

Upon further research, it is revealed that excessive compensation is a common result of weak governance, possibly brought about by managerial power (Core et al., 1999; Bertrand & Mullainathan, 2001; Grinstein & Hribar, 2003). In fact, a study by Bebchuk & Fried (2006) suggests that an entrenched CEO can hold so much bargaining power over the board of directors that they would be able to influence his / her own compensation. To better illustrate this point, the board of directors of a typical public company should represent and act on behalf of shareholders, which includes monitoring the CEO to act in shareholders’ interests. However, if the positions of CEO and board chairman are occupied by the same person, the board loses its independence and hence its value in effective monitoring and control. In such a case, the CEO would have enormous power within the organisation and can easily over-compensate his / her own work.

It is important, therefore, to retain a level of independence in the board of directors to ensure alignment of shareholder objectives with executive incentives. This is further supported by Paul (2007), which suggests that a board of directors with higher independence decreases the likelihood of making sub-optimal acquiring bids, whereas Hardwick et al. (2011) reveals a positive relationship between the proportion of non-executive directors and profitability, especially in cases where the CEO and board chairman are occupied by different individuals.

(9)

9 Hypothesis 2:

Excessive equity based compensation for the executive management of the acquiring company is negatively correlated with shareholder value in a takeover

2.2 Blockholder ownership

The issue of agency conflict can also be addressed by way of shareholder intervention. Agency conflicts can be remediated by intervening actions such as preventing value-damaging decisions or relieving underperforming executives from their roles. These actions are, however, often time-consuming and costly for the investor, despite their positive effects on uplifting firm value (Edmans, 2014). This is because the process of gathering and analysing information, or influencing managers or other shareholders often proves to be challenging for an investor who does not normally actively engage himself in the operations of the company (Fich et al., 2015).

Interestingly, it is observed that in a company with a dispersed shareholder base, shareholders are less likely to intervene due to the free-rider problem, the situation where one or a small group of investors bear most of the costs whereas all shareholders receive the benefits. Still, some major shareholders (blockholders) are willing to internalise the costs of monitoring, especially on significant corporate investment decisions such as mergers and acquisitions, which would have a significant bearing on the company’s share price (Gillan & Starks, 2000).

The likelihood of intervention also depends on the nature of the investors themselves. In fact, Gillan & Starks (2007) has found that mutual funds and pension funds intervention are less likely to drive company performance, whereas hedge funds are better able to influence corporate boards and management through employing highly incentivised fund managers (Brav et al., 2008).

The effect of blockholder ownership also extends to the context of mergers and acquisitions. Fich et al. (2015) has found that institutional ownership correlates positively with the return of the company being acquired (i.e. target company), and that it results in a higher chance of deal completion as well as higher premiums for the target company. In turn, this translates to a lower return for the acquiring firm and a lesser number of bids since the terms are less favourable to bidders. These relationships were shown to be more significant in cases where the target firm represented a substantial allocation of funds in the investor’s portfolio. Similarly, Boyson et al., (2017) has found a positive relationship between hedge funds and target return, primarily by influencing takeover outcomes.Rather than simply seeking to acquire undervalued companies or to sell portfolio companies for higher premiums, hedge funds add value by effective monitoring of executive management decisions.

(10)

10

On the downside, major blockholders may hold excessive bargaining power, and in effect disparage governance by extracting private benefits or pursuing objectives other than maximising firm value, such as implementing labour-friendly policies (unions and pension funds) (Edmans, 2014; Cohn et al., 2016). The presence of blockholders may also lower liquidity of shares, and erode managerial initiative due to the foresight of possible resistance from shareholders.

More recently, a study conducted by Harris et al. (2010) has found that bidding managers tend to make certain financing decisions in takeovers as an attempt to circumvent monitoring by shareholders. This can be achieved by utilising either cash or shares as payment in a takeover under different scenarios. To further illustrate this, where the acquiring company has aggressive blockholders, the bidding manager may attempt to dilute the ownership of the blockholders by financing the takeover with equity rather than cash. Following the same principle, where the target company has relatively more aggressive blockholders, the bidding manager would tend to acquire the target company with cash rather than equity to minimize influence from the target company’s blockholders. Interestingly, executive management are also observed to minimize shareholder intervention during the acquisition process by financing it in different ways (Li, Liu, & Wu, 2017). According to listing rules on NYSE, AMEX and NASDAQ, shareholder voting is mandatory when an acquiring company wishes to issue over 20% of new shares to finance an acquisition. Leveraging this policy, Li, Liu & Wu (2017) identified a sample of mixed payment deals where the acquiring company had financed the acquisitions with a substantial proportion of cash relative to equity, from which it was observed that acquirers with more severe agency problems and overconfident CEOs were more likely to bypass shareholder voting using cash rather than stock to avoid the 20% rule.

This, however, proves detrimental to shareholders. From a second sample comprising all stock deals where the acquisitions were financed only by equity, it was found that the return of deals that just bypassed the shareholder voting (i.e. those issuing slightly below 20% of shares) were 4.3% lower than those with shareholder voting (i.e. those issuing slightly over 20% of shares). By use of this regression discontinuity design, the study suggests a positive causal effect of shareholder voting on M&A returns, a relationship that is found prevalent amongst companies with high institutional ownership (Li, Liu & Wu, 2017).

Hypothesis 3

(11)

11 2.3 Combined effect

A study conducted by Burkart et al. (1997) explores the relationship between executive management and monitoring institutions. The study suggests that constraints imposed by major shareholders by monitoring company decisions of the executive management can be costly, since in certain circumstances managerial discretion comes with benefits. This is because when shareholders are likely to intervene, executive management is often less inclined to search for new investment opportunities.

Although extensive research has already been conducted on the sole effects of equity based compensation and blockholder ownership on M&A returns, studies on their combined effect are limited. However, based on Burkart et al. (1997), and other past research relating to equity based compensation and blockholder ownership, it can be argued that an appropriate complementary balance between monitoring and equity based compensation could result in more superior outcomes in aligning the interests of the executive management with that of shareholders.

The disadvantage of equity based compensation is that executives can become too risk-seeking, pursuing excessively risky M&A investments that are detrimental to value maximization (Malmendier & Tate, 2005, 2008, 2009). On the other hand, the prevalence of blockholders can limit managerial discretion from fear of intervention (Burkart et al. 1997). The combination of both effects, however, could have a synergising effect in mitigating the explained disadvantages. Although highly incentivized executives can become excessively risk-seeking, under effective monitoring by blockholders, executives will be deterred from proceeding with investments that do not align with shareholder interest. At the same time, the negative impact driven by limited managerial discretion due to the prevalence of a blockholder may be mitigated by incentivizing executive management through the use of equity based compensation.

Hypothesis 4

The interaction of blockholder ownership and equity based compensation for the executive management of the acquiring company has a positive effect on shareholder value during acquisitions; i.e. the effect of equity based compensation is more positive when blockholder ownership is larger 2.4 Firm and transaction characteristics

Taking reference to Jensen’s free cash flow theory (1986), executive managers in companies with a high level of free cash flow are more likely to engage in investments detrimental to shareholder value. This is likely because managers in control of a higher level of free cash flow essentially have access to more resources to pursue their own objectives. For example, executive management are often keen to pursue acquisitions to manage a larger-sized company (empire building). Although such acquisitions do not necessarily result in negative share price movements, an acquisition driven by personal

(12)

12

objective is not likely to be value-maximising. Based on the same theory, debt would have the opposite effect due to mandatory repayments and interest costs, which would in turn reduce free cash flow and hence limit resources available to exercise managerial discretion (Masulius et al., 2007).

Takeover return is also influenced by the nature of the firm being acquired, in terms of whether it is publicly listed or private company. The process of acquiring a public firm tends to be more complicated considering the higher number of parties involved, whereas takeovers involving private target firms are generally less challenging because of limited bidder competition (Feito-Ruiz & Renneboog, 2017). Therefore, the costs associated with acquiring public companies are often greater as compared with acquiring private companies (Officer, 2007). Aside from the higher costs arising from stronger bidder competition, public companies also require higher offers than do unlisted firms. This is because shares of listed firms have liquidity premium since they are publicly traded on the stock market and could therefore be easily sold if investors chose to sell their stake, as opposed to the shares of unlisted companies (Fuller et al., 2002).

Aside from the acquirer’s level of free cash flow and nature of the target (i.e. private vs. publicly listed), the method of payment is also found to affect the announcement return of bidding companies. In general, the majority of acquisitions are financed by either equity or cash. Usually, equity involves higher costs since investors would require a higher level of return for the higher risks and opportunity costs involved. In comparison, free cash involves much lower opportunity costs (i.e. foregone interest from bank deposits). This is why a bidding company’s decision to fund an acquisition with equity is often perceived negatively by the market (Sudarsanam & Mahatem, 2003; Moeller et al., 2004). Since a company’s executive management is endowed with more knowledge on the company’s value than the potential acquirer due to information asymmetry in the market (Myers and Majluf, 1984), the use of equity as a means of financing the acquisition can signal that the company’s shares have been overvalued (Myers and Majluf, 1984). In contrast, an acquiring company making a cash offer signals confidence in its own company performance given its reluctance to issue potentially undervalued shares.

On a more global level, cross border M&A is observed to have a mixed effect on share price. In a study conducted by Francis et al. (2008), positive results were found for cross-border M&A due to the expanded access to international capital markets. Since it is in general more difficult for firms based in underdeveloped financial markets to access funding for making investments, these companies are often forced to forgo value-enhancing projects. However, in the event that they are acquired by multinational companies, they would be empowered with better access to the international financial markets to finance projects. Similarly, Bris & Cabolis (2008) also found a positive relationship between M&A returns and cross border transactions, but for a different reason. This study found that better protection and more stringent accounting practices in the country of the acquiring company tends to

(13)

13

increase the return around the time of announcement. Better governance is beneficial to the target company since it eliminates managerial opportunism (Hagendorff et al., 2007; Martynova & Renneboog, 2008). On the other hand, Moeller & Shlingemann (2005) suggested a negative relationship between cross-border M&A and share price performance. This can be explained by asymmetric information, as well as agency conflicts in the foreign target company due to limited managerial influence over the acquired company relative to the holding company where most managers are based.

Another characteristic determining M&A returns would be the level of similarities between the acquirer and target company (Hoberg & Philips, 2010). This is because of the potential synergy effects resulting from a more streamlined approach to integration, and higher efficiency in managing new assets from the target company. By use of the text-based analysis of 10-K product descriptions, Hoberg & Philips (2010) developed a new measure that identifies and compares similarities across companies, and found a higher profitability for merging firms with higher ex ante similarity. This new measure is unique in the sense that it is able to capture the relatedness of firms in the product market space rather than only in terms of industry categories, and therefore more effectively captures similarities both within and across industries. In contrast, neither the North American Industry Classification System (NAICS) nor the Standard Industry Classification (SIC) are able to provide such spatial or continuous representation of pairwise similarity of any two firms. An example would be the merger of Pixar and Disney, which were classified in different SIC codes. However, Hoberg & Philips’ similarity measure showed a rich array of relatedness. Therefore, the Pixar Disney merger was expected to be positive based on Hoberg & Philips’ theory since both firms were similar enough to facilitate asset complementarities and new product synergies. Following the merger, many Disney movies (e.g., Toy Story, Cars) have been produced leveraging Pixar’s technology and distributed through Disney’s promotion and retail channels.

3. Methodology

Taking reference to studies conducted by Moeller et al. (2004) and Masulis et al. (2007), a two-step procedure is employed to determine the effect of blockholder ownership and equity based compensation on abnormal returns during acquisitions. First, by use of an event study, the abnormal return of the acquiring company around the time of M&A announcement is estimated. The second step involves the use of cumulative abnormal returns (CARs) of these firms over a five-day event window (-2, +2) as dependent variable to conduct a multivariate analysis using an OLS regression model with robust standard errors. The dependent and independent variables will be discussed in more detail below and are summarised in Table 1.

(14)

14 3.1 Dependent variable

As set out above, the cumulative abnormal return for shareholders of the acquiring firm is used as the dependent variable. The return of the company around the time of acquisition announcement consists of the normal return and the abnormal return.

R

it

= NR

it

+ AR

it (1)

To account for the normal return, the market-adjusted return model is employed. The market parameters will not be estimated based on a time period before each bid since some acquirers might have had previous takeover attempts which would have been included in the estimation period, making the beta estimates less meaningful (Fuller et al., 2002). Past research has also shown that in short-window event studies, weighting the market return with the firm's beta does not significantly improve the accuracy of the estimation (Fuller et al., 2002).

NR

it

= R

mt

+ ε

it (2)

The market value-weighted return is then used as a proxy for the normal return and subtracted from the actual return of the acquiring firm around the time of M&A announcement to determine the abnormal return in the event window (-2, +2).

AR

it

= R

it

- R

mt (3)

As a final step, the cumulative abnormal return for the firm is calculated by using the sum of the daily abnormal return over the event window.

CAR

i

= Σ AR

i, t-2 to t+2 (4)

3.2 Independent variables 3.2.1 Variables of interest

- Equity based compensation

“Equity based compensation” is calculated by dividing the stock-based pay and option-based pay of the executive management by their total compensation (Masulis et al., 2007; Harris et al., 2013). This measure is expected to be positively correlated to stock price performance around the time of M&A announcements (Datta et al., 2001) as equity based compensation

(15)

15

incentivizes managers to act in the interest of shareholders by pursuing value generating investments.

- Executive ownership

“Executive ownership” refers to the percentage of company shares owned by the executive management. Similar to equity based compensation, executive ownership essentially aligns the interests of the executive management with that of shareholders. Therefore, higher stock ownership is expected to have a positive effect on M&A returns (Shinn 1999; Walters et al., 2007; Kroll et al., 2008).

- Blockholder ownership

“Blockholder ownership” is measured by the percentage ownership of the largest blockholder (Andriosopoulos & Yang, 2015). The higher the ownership percentage, the more likely that blockholders would intervene and monitor firm activities. As such, blockholders are more likely to appeal if they believe that the M&A decision was driven by the executive management’s personal benefits rather than shareholders’ interests. This is further supported by Harris et al. (2013) and Fich et al., (2015) which found a positive effect of a large shareholder (i.e. blockholder) on a firm’s value.

- Multiple blockholders

“Multiple blockholders” is set as a dummy variable that equals one in the case where there are multiple blockholders (defined as shareholders with 5% ownership or more) and zero otherwise. Although a blockholder tends to influence shareholder value positively (Harris et al., 2013; Fich et al., 2015), the prevalence of multiple blockholders has a negative effect as it exacerbates the free-rider problem (Edmans, 2014).

3.2.2 Transaction characteristics - Payment method (Cash)

“Payment method” is measured by dividing the amount of cash used for the transaction by the deal value (Moeller et al., 2004; Masulis et al., 2007; Harris et al., 2013; Fich et al., 2015; Andriosopoulos & Yang, 2015; Feito-Ruiz & Renneboog, 2017). For deals funded only by cash, this variable would be one; whereas for deals funded by only equity, the measure would be zero. In effect, mixed cash equity deals would have a value between one and zero. Prior

(16)

16

research (Travlos, 1987; Sudarsanam & Mahatem, 2003; Moeller et al., 2004) showed a positive relationship between the performance of the acquiring firm and cash financing. - Transaction value

”Transaction value” is calculated as the natural logarithm of the deal value, and is expected to correlate negatively with the return of the acquiring company. This can be explained by the difficulties in integrating with larger sized companies as compared with smaller-sized companies, which would in turn lower the return on investments.

- Relative value

“Relative value” is measured by transaction value divided by the market value of the acquirer’s assets. By a similar logic as “transaction value” from the section above, integration with relatively larger sized companies is more time consuming and complex, which could result in a lower return.

- Tender Offer

“Tender offer” is set as a dummy variable that equals one if the bid consists of a tender offer, and zero otherwise (Moeller et al., 2004; Harris et al., 2010; Fich et al., 2015; Feito-Ruiz & Renneboog, 2017). When a tender offer is made, the board of the target company is often bypassed, which could negatively impact the bidder’s announcement returns due to limited discussions and negotiations (Moeller et al., 2004).

- Multiple Bidders

“Multiple Bidders” is set as a dummy variable, equalling one if multiple bidders are involved and zero otherwise (Harris et al., 2010; Feito-Ruiz & Renneboog, 2017). In the case of multiple bidders, competition may cause the winner to succumb to the “winner’s curse”, resulting in a higher price for the acquisition hence lowering the return for the acquiring firm (Moeller et al., 2004; Feito-Ruiz & Renneboog, 2017).

- Similarity between the acquirer and the target firm

Acquisitions of similar companies are expected to create more synergies than acquisitions of unrelated companies (Hoberg & Philips, 2010). To establish a measure of similarity between companies, the text-based measure of product market relatedness one year prior to the acquisition proposed by Hoberg and Phillips (2010) is adopted in this study. This measure

(17)

17

effectively captures the relatedness of firms in the product market space with the ability to measure firm similarities both within and across industries, rather than segregating them by industry.

3.2.3 Governance measures - Independent Board

“Independent board” is measured by the proportion of independent directors on the board in the year preceding the acquisition (Feito-Ruiz & Renneboog, 2017). Fama and Jensen (1983) have suggested that competition in the labour market for independent board members encourages them to build and maintain an image of professionalism. These board members would therefore be expected to exercise a greater degree of influence on strategic decisions compared to insiders who would be subjected to a higher degree of influence from the CEO. - Board size

“Board size” refers to the number of executive and non-executive directors on the board in the year preceding the acquisition (Masulis et al., 2007; Feito-Ruiz & Renneboog, 2017; Harris et al., 2013). Although increasing the size of a board is likely to increase its effectiveness, boards that comprise an excessive number of board members may in fact be counterproductive and reduce its capacity to adapt promptly to dynamic market changes (Jensen, 1993). Therefore, a larger sized board may be associated with greater risks from the lack of decisiveness, which could in turn affect firm value negatively (Yermack, 1996; Eisenberg et al., 1998).

- CEO Duality

“CEO duality” is a dummy variable that equals one if the positions of CEO and chairman are held by the same person in the year preceding the acquisition and zero otherwise (Moeller, 2004; Masulis et al., 2007; Harris et al., 2013; Feito-Ruiz & Renneboog, 2017). The agency theory proposed by Jensen & Meckling (1976) suggests that the simultaneous occupation of the positions of CEO and Chairman is detrimental to company performance due to the lack of separation between executive management and shareholder control. This view is further supported by other studies (Agrawal and Knoeber, 1996).

(18)

18 3.2.4 Acquirer characteristics

- Free cash flow

“Free cash flow” is measured by dividing operating cash flow by the book value of assets in the year preceding the acquisition (Fich et al., 2015). Companies with a high level of free cash flow are more likely to pursue M&A decisions that are detrimental to shareholder value, as managers with better access to financial resources are more likely to engage in projects benefiting their own self-interest rather than maximizing shareholder value (Jensen, 1986). - Cash Holding

“Cash holding” is measured by dividing the cash and equivalents by the book value of assets in the year preceding the merger or acquisition (Moeller et al., 2005; Ferreira et al., 2009; Harris et al., 2010; Andriosopoulos & Yang, 2015).

In this study, “cash holding” is set as the control variable since the method of payment has a significant effect on takeover return (Travlos, 1987; Sudarsanam & Mahatem, 2003; Moeller et al., 2004).

- Leverage

“Leverage” is calculated by dividing the total debt with the market value of assets in the year preceding the acquisition (Moeller et al., 2004; Masulis et al., 2007; Ferreira et al., 2009; Harris et al., 2013; Fich et al., 2015; Feito-Ruiz & Renneboog, 2017). A high level of debt lessens the availability of free cash flow, that might otherwise be used by managers to pursue investments or projects for their own personal interests rather than to maximize firm value (Jensen 1986). However, prior research has shown inconsistent results on the relationship between leverage and M&A returns. While Masulis et al. (2007) had found a positive relationship between leverage and short-term performance of the acquiring company, Moeller et al. (2004) found the opposite. Despite the inconsistencies, the debt level is included to account for the effect of the capital structure.

- Company size

“Company size” is measured as the natural logarithm of the book value of assets in the year preceding the acquisition (Moeller et al., 2005; Masulis et al., 2007; Harris et al., 2013). In Moeller et al. (2004), the size of the acquiring company is found to be negatively correlated with acquisition return, which supported the managerial hubris hypothesis (Roll, 1986), since larger acquirers pay higher premiums and pursue acquisitions that generate negative

(19)

19

synergies. Another explanation could relate to how a large firm size can serve as a defense mechanism for aggressive takeovers as it requires more resources to acquire a large company. Managers of larger companies are therefore more entrenched and are more likely to make value-reducing acquisitions (Masulis et al., 2007).

4. Data and descriptive statistics

This section serves to provide further details on the dataset and its characteristics. First, the variety of data sources and the sample formation process will be discussed. Then, the descriptive statistics would be analysed in detail, by comparing the treatment sample (M&A sample consisting of S&P 1500 companies that have engaged in M&A activities) with the control sample (full sample comprised of companies in the S&P 1500).

4.1 Data

The deal and transaction characteristics analysed in this study are retrieved from Thomson One, with a sample period starting from 2002 when the market was recovering from the equity crisis till 2016. To be included in the dataset, the transaction has to satisfy the following criteria:

(1) Transaction has to be completed;

(2) Transaction has to be conducted between two US-listed companies;

(3) The acquiring company has to acquire 50% or more of the target company’s shares; (4) The deal value has to be over $10 million; and

(5) The acquirer or target must not be a subsidiary.

In this study, company data such as debt and cash are collected from Compustat, while information on executive compensation is collected from the Execucomp database. The holding period returns of companies around the time of M&A announcement are obtained from the Center for Research in Security Prices (CRSP), whereas corporate governance measures and ownership information are retrieved from the Institutional Shareholder Services (ISS). Finally, the similarity measure between companies is collected from the Hoberg-Phillips Data Library.

In view of regulatory issues or unique capital structures, financial and utility firms (i.e. companies with SIC code 4900-4999 and 6000-6999) are excluded from the dataset. To account for data errors, unnatural values (e.g. negative debt or over 100% ownership for executives) are eliminated and outliers are trimmed at 1% depending on the variables. Since the focus of the paper lies on acquisitions of new companies, a number of selected deal types are also excluded, namely: minority stake purchases, acquisitions of remaining interest, spinoffs, recapitalizations, self-tenders, and repurchases. After applying the exclusions as set out above, 411 M&A deals are retained for

(20)

20

analysis in the descriptive statistics section, whereas 252 M&As are utilised in the regression analysis since Execucomp only provides information on equity based compensation from 2006 onwards.

4.2 Descriptive statistics

Table 2 summarises the descriptive statistics of the cumulative abnormal returns and transaction characteristics (including “payment method (Cash)”, “transaction value”, “relative size”, “tender offer”, “multiple bidders”, and “similarity”). Table 3 describes the same variables but on an annual basis.

In line with prior research (Jensen & Ruback, 1983; Agrawal et al., 1992; Moeller et al., 2004),

this study found a negative abnormal return for the acquiring company with a mean (median) of -0.41% (-0.41%). Given a mean (median) market value of the acquirer of $41.8 ($11.1) billion (see Table

8), acquirers’ shareholders lose approximately $171.4 ($45.5) million per transaction. When considering the annual distribution, a more negative CAR during the financial crisis (2008 & 2009) can be observed, with a median CAR of -1.22% and -2.04% respectively. After the financial crisis, M&A investments in general demonstrated better performance with a positive CAR between the period of 2011 and 2014.

On average, transactions in the M&A sample were financed with 81% cash and 19% equity. In the period after the equity crisis (2002 - 2006), the average cash financing for deals fell between 60% and 80%, whereas during and after the financial crisis (2007 - 2016), the average deal was financed by cash between 80% and 90%, out of which 29% were completed following a tender offer, and 5% included multiple bidders.

As illustrated in Table 2, the relative sizes of M&A deals are observed to be relatively small. Although the mean of “relative size” lies at 15%, the median is only 6%, which implies that most transactions pursued by acquiring companies represent less than 10% of their own company size. As for “similarity”, the measure provided by Hoberg and Philips shows a mean (median) similarity of 3.28% (0.66%) between acquirers and targets.

(21)

Table 1: Description of variables and data source

Variable name Definition Data source

Equity based compensation Stock-based pay and option-based pay of the executive management divided by total compensation Execucomp

Executive ownership Percentage of company shares owned by the executives Execucomp and Compustat

Blockholder ownership The percentage ownership of the largest Blockholder ISS

Multiple blockholders Dummy variable that equals one when there are multiple blockholders ISS

Payment method (Cash) Proportion of cash used for the transaction Thomson One

Transaction value Natural logarithm of the deal value Thomson One

Relative size Transaction value divided by the market value of assets of the acquirer Thomson One and Compustat

Tender offer Dummy variable that equals one if the bid consists of a tender offer Thomson One

Multiple bidders Dummy variable that equals one if multiple bidders were involved Thomson One

Similarity Text-based measure of product market relatedness proposed by Hoberg and Phillips Hoberg and Philips Library

Independent board Independent board members divided by number of directors in the board ISS

Board size Number of directors in the board ISS

CEO duality Dummy variable that equals one if the positions of CEO and chairman are held by the same person ISS

Free cash flow Operating cash flow divided by book value of assets Compustat

Cash holding Cash and equivalents divided by book value of assets Compustat

Leverage Debt (short term and long term) divided by market value of assets Compustat

Company size Natural logarithm of the book value of assets Compustat

Note: All variables are measured for the year preceding the acquisition exept for the transaction characteristics

Panel A: Variables of interest

Panel B: Transaction characteristics

Panel C: Governance measures

(22)

Table 2. Descriptive Statistics of CAR and Transaction Characteristics

The table reports the mean, standard deviation, minimum, median, and maximum of the dependent variable “cumulative abnormal return (CAR)”, and transaction characteristics including “payment method”, “transaction value”, “relative size”, “tender offer”, “multiple bidders”, and “similarity” for the period 2002 - 2016. To be included in the dataset, the transaction has to be (1) completed, (2) between two companies listed in the U.S., (3) the bidding companies acquires more than 50% of the target’s shares, (4) the deal value is $10 million or more (5) the acquirer or target is not a subsidiary. The dataset will exclude financial and utility firms due to regulatory issues or uncommon capital structure in these industries. It is also required that the acquirers have all the dependent and independent variables available. The sample consists of 411 completed M&A transactions. To account for outliers, the sample is winsorized at the 1% level depending on the variables. Variable definitions, construction, and sources can be found in the Table 1.

Varriable name Mean Std. Dev. Min Median Max

CAR (-2, +2) (0.41%) 5.83% (29.19%) (0.41%) 19.07%

Payment method (Cash) 81% 34% 0% 100% 100%

Transaction value12 ₂₃₉₈ ₆₆₂₉ ₁₀ ₅₄₀ ₆₇₂₈₆

Relative size 15% 24% 0% 6% 162%

Tender offer 29% 46% 0.00 0.00 1.00

Multiple bidders 5% 22% 0.00 0.00 1.00

Similarity 3.28% 4.45% 0.00% 0.66% 24.10%

(23)

Table 3. Annual Statistics of CAR and Transaction Characteristics

The table reports the numbers of deals, the percentage of the total deals, annual median statistic of the dependent variable “cumulative abnormal return (CAR)” and the median of the transaction characteristics “payment method”, “transaction value”, “relative size”, “tender offer”, “multiple bidder”, and “similarity” for the period 2002 - 2016. If it involves an indicator variable, mean statistics will be used instead of the median. To be included in the dataset, the transaction has to be (1) completed, (2) between two companies listed in the U.S., (3) the bidding companies acquires more than 50% of the target’s shares, (4) the deal value is $10 million or more (5) the acquirer or target is not a subsidiary. The dataset will exclude financial and utility firms due to regulatory issues or uncommon capital structure in these industries. It is also required that the acquirer have all the dependent and independent variables available (except for equity based compensation between 2002 and 2006). The sample consists of 411 completed M&A transactions. To account for outliers, the sample is winsorized at the 1% level depending on the variables. Variable definitions, construction, and sources can be found in the Table 1.

Year N % CAR (-2, +2) Payment method Transaction value Relative size Tender offer Multiple bidders Similarity

2002 31 7.54% (1.82%) 61% 203 6% 32% 6.45% 0.00% 2003 27 6.57% 0.00% 73% 142 7% 37% 7.41% 0.00% 2004 34 8.27% (1.21%) 63% 644 13% 12% 2.94% 1.31% 2005 32 7.79% (0.62%) 71% 917 4% 13% 15.63% 1.45% 2006 27 6.57% (0.47%) 82% 766 4% 7% 3.70% 1.45% 2007 38 9.25% 0.32% 98% 637 6% 34% 0.00% 0.00% 2008 25 6.08% (1.22%) 92% 234 4% 60% 0.00% 1.09% 2009 24 5.84% (2.04%) 82% 1354 3% 54% 12.50% 3.83% 2010 31 7.54% (0.30%) 88% 442 5% 35% 0.00% 0.00% 2011 17 4.14% 0.17% 88% 522 3% 41% 5.88% 0.02% 2012 25 6.08% 0.08% 91% 618 3% 24% 0.00% 0.64% 2013 15 3.65% 0.80% 84% 944 10% 13% 0.00% 1.01% 2014 24 5.84% 0.22% 81% 769 6% 29% 8.33% 0.39% 2015 34 8.27% (0.53%) 79% 1413 9% 26% 5.88% 2.00% 2016 27 6.57% 0.05% 94% 888 6% 30% 3.70% 1.22%

(24)

Table 4 illustrates the descriptive statistics of the variables of interest (i.e. “equity based compensation”, “executive ownership”, “blockholder ownership”, and “multiple blockholders”) for the “M&A sample” and “full sample”. The full sample covers all S&P 1500 companies with non-missing variables of interest, governance measures, and acquirer characteristics, whereas the M&A sample covers the S&P 1500 that have engaged in M&A activities with non-missing dependent and independent variables. Table 5 covers the same variables and samples but on an annual basis.

From the data results, the mean (median) of “equity based compensation” for the full sample lies at 47% (50%), whereas that for the M&A sample is 54% (57%). As illustrated in Table 5 and Graph 1, “equity based compensation” for the M&A sample appears to be higher than that of the full sample over 2007 - 2016. Therefore, it is observed that companies offering higher equity based compensation for its executive management are more likely to pursue M&A investments.

Executive ownership, however, seems to have a different effect on the likelihood of pursuing M&A investments. The mean (median) of “executive ownership” for the full sample is 5.61% (3.14%), whereas that for the M&A sample is 3.72% (2.13%). As can be seen in Table 5 and Graph 2, “executive ownership” is found to be lower for the M&A sample compared to the full sample over 2002 - 2016 with the exception of 2013. Another interesting observation from Graph 2 lies in the decreasing trend of “executive ownership” during the sample period.

Combining the observations of “equity based compensation” and “executive ownership”, it would appear that equity based compensation has a stimulating effect on executives to pursue sub-optimal M&A investments while executive ownership does the opposite. This could potentially be explained by the inducing effect that equity based compensation has on executive management to become excessively risk-seeking. Executive ownership on the other hand, lowers incentives for executives to pursue risky investments due to their stake on the company’s value and performance, which aligns their interests with that of shareholders.

The mean (median) of the biggest blockholder’s ownership in the full sample is 6.92% (6.00%), out of which 44% of companies have multiple blockholders, whereas that in the M&A sample is 7.16% (7.20%) and 55% respectively. From the results, it might appear that companies with larger shareholders and multiple blockholders favour M&A investments. However, this trend is not consistent over the entire sample period. As illustrated in Table 5 and Graph 3, the medians of “blockholder ownership” in 2003 and 2009 are observed to be lower in the M&A sample as compared to the full sample, whereas in 2005, 2010, and 2016, the medians are found equal in both samples. Therefore, a clear relationship between blockholder ownership and the likelihood of pursuing M&A investments cannot be established from the descriptive statistics.

(25)

25 Table 4. Descriptive Statistics of Variables of Interest

The table reports the mean, standard deviation, minimum, median, and maximum of the variables of interest including “equity based compensation” (EBC), “executive ownership”, “blockholder ownership”, and “multiple blockholders” for the period 2002 - 2016. All variables are recorded at the year-end preceding the acquisition announcement. To be included in the dataset, the transaction has to be (1) completed, (2) between two companies listed in the U.S., (3) the bidding companies acquires more than 50% of the target’s shares, (4) the deal value is $10 million or more (5) the acquirer or target is not a subsidiary. This dataset will exclude financial and utility firms due to regulatory issues or uncommon capital structure in these industries. It is also required that the acquirer have all the dependent and independent variables available. The M&A sample consists of 411 completed M&A transactions. The full sample covers all S&P 1500 companies with 14086 observations having non-missing variables of interest, governance measures, and acquirer characteristics. To account for outliers, the sample is winsorized at the 1% level depending on the variables. Variable definitions, construction, and sources can be found in the Table 1.

Table 5. Annual Median Statistic of Variables of Interest

The table reports the annual median statistic of the variables of interest including “equity based compensation” (EBC), “executive ownership”, and “blockholder ownership” for the period 2002 - 2016. If it involves an indicator variable, mean statistics will be used instead of the median. All variables are recorded at the year-end preceding the acquisition announcement. To be included in the dataset, the transaction has to be (1) completed, (2) between two companies listed in the U.S., (3) the bidding companies acquires more than 50% of the target’s shares, (4) the deal value is $10 million or more (5) the acquirer or target is not a subsidiary. The dataset will exclude financial and utility firms due to regulatory issues or uncommon capital structure in these industries. It is also required that the acquirer have all the dependent and independent variables available. The M&A sample consists of 411 completed M&A transactions. The full sample covers all S&P 1500 companies with 14086 observations having non-missing variables of interest, governance measures, and acquirer characteristics. To account for outliers, the sample is winsorized at the 1% level depending on the variables. Variable definitions, construction, and sources can be found in the Table 1.

Varriable name Mean Std. Dev. Min Median Max Mean Std. Dev. Min Median Max

Equity based compensation1 _54% _19% _0% _57% _98% _47% _21% _0% _50% _98%

Executive ownership 3.72% 5.18% 0.01% 2.13% 34.23% 5.61% 7.77% 0.00% 3.14% 89.35% Blockholder ownerhsip 7.16% 6.80% 0.00% 7.20% 63% 6.92% 11.34% 0.00% 6.00% 100% Multiple blockholders 55% 50% 0.00 1.00 1.00 44% 50% 0.00 0.00 1.00 Note (1) sample period is 2007 - 2015 due to data availability in Execucomp (N=252 in M&A sample and N=9242 in Full sample)

Full sample (N = 14086) M&A sample (N = 411)

Year EBC1 _{Exec. Owners.} _{Block. Owners.} _EBC1 _{Exec. Owners.} _{Block. Owners.}

2002 n.a. 3.05% 7.9% n.a. 4.53% 5.3% 2003 n.a. 3.89% 0.0% n.a. 4.55% 5.2% 2004 n.a. 3.89% 9.4% n.a. 4.44% 5.6% 2005 n.a. 2.27% 6.0% n.a. 4.16% 6.0% 2006 n.a. 1.45% 9.7% n.a. 3.62% 6.0% 2007 47% 2.49% 7.4% 46% 3.17% 6.3% 2008 55% 2.29% 9.3% 48% 3.09% 5.6% 2009 61% 0.72% 0.0% 49% 3.17% 5.7% 2010 54% 2.49% 5.9% 45% 2.99% 5.9% 2011 56% 1.25% 9.6% 46% 2.89% 5.9% 2012 60% 2.45% 7.3% 50% 2.88% 5.6% 2013 56% 2.96% 8.9% 50% 2.57% 6.2% 2014 54% 1.37% 7.1% 53% 2.30% 6.1% 2015 62% 1.35% 7.3% 53% 2.00% 6.3% 2016 62% 0.89% 6.6% 55% 1.90% 6.6%

Note (1) sample period is 2007 - 2015 due to data availability in Execucomp (N=255 in M&A sample and N=9242 in Full sample)

Full sample (N = 14086) M&A sample (N = 411)

(26)

26

Graph 1. Annual Median Statistic Equity Based Compensation (EBC)

This graph shows the annual median “equity based compensation”. Specific information can be found in Table 5. Variable definitions, construction, and sources can be found in Table 1.

Graph 2. Annual Median Statistic Executive Ownership

This graph shows the annual median “executive ownership”. Specific information can be found in Table 5. Variable definitions, construction, and sources can be found in Table 1.

Graph 3. Annual Mean Statistic Blockholder Ownership

This graph shows the annual median statistic of “blockholder ownership”. Specific information can be found in Table 5. Variable definitions, construction, and sources can be found in Table 1.

30% 40% 50% 60% 70% 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016

Equity Based Compensation

M&A sample Full sample

0% 1% 2% 3% 4% 5% 2002 2004 2006 2008 2010 2012 2014 2016 Executive Ownership

0.0% 3.0% 6.0% 9.0% 12.0% 2002 2004 2006 2008 2010 2012 2014 2016 Blockholder ownership

(27)

27

Table 6 reports the descriptive statistics of the governance measures (i.e. “independent board”, “board size”, and “CEO duality”) for the “M&A sample” and “full sample” respectively. The full sample covers all S&P 1500 companies with non-missing variables of interest, governance measures, and acquirer characteristics, whereas the M&A sample covers the S&P 1500 that have engaged in M&A activities with non-missing dependent and independent variables. Table 7 covers the same variables and samples but on an annual basis.

The mean (median) of the “independent board” for the full sample is 75% (78%), similar to that of the M&A sample with 76% (80%). Interestingly, as can be observed in Graph 4, the proportion of independent directors in the board show an increasing trend over the entire sample period for both samples respectively, showing that companies in general are increasing the independence of their boards.

The measure “board size” is also found to exhibit similar characteristics in both the M&A sample and the full sample. The median of the full sample and that of the M&A sample are both observed to be 9, while the mean of the full sample is 9, only slightly lower than that of the M&A sample at 9.5. As illustrated in Table 7, the board size is in general stable for both samples – while the median of the full sample stays constant at 9 over the entire sample period, the median of the M&A sample remains similarly steady with slight variations between 8 and 10 and an outlier of 11 in 2009.

The final governance measure is “CEO Duality”. The positions of CEO and chairman are held by the same person in 55% of the companies in the S&P 1500 (i.e. full sample), which is similar to that of the M&A sample with 54%. This similarity is consistent over the sample period with the exception of 2008, the year of the financial crisis, when both samples showed a significant decrease in “CEO duality“ (see Graph 5). Interestingly, a general decreasing trend in “CEO Duality” is observed in both samples. In the M&A sample, the percentage of companies with “CEO duality” was originally around 60% in 2002 which had dropped to 30% by 2016, whereas the full sample had a percentage of approximately 65% in 2002 which had decreased to around 40% by 2016.

Considering the combination of all governance measures, a positive trend in corporate governance for public companies can be observed, with a rising proportion of independent directors, along with a decrease in likelihood of CEO Duality. Since governance measures and trends are in general similar across both samples, a clear relationship between governance measures and the likelihood of pursuing M&A cannot be established from the descriptive statistics.

(28)

28 Table 6. Descriptive Statistics of Governance Measures

The table reports the mean, standard deviation, minimum, median, and maximum of the governance measures ‘independent board”, “board size”, and “CEO duality” for the period 2002 - 2016. All variables are recorded at the year-end preceding the acquisition announcement. The M&A sample covers companies who engaged in M&A activities. To be included in the dataset, the transaction has to be (1) completed, (2) between two companies listed in the U.S., (3) the bidding companies acquires more than 50% of the target’s shares, (4) the deal value is $10 million or more (5) the acquirer or target is not a subsidiary. This dataset will exclude financial and utility firms due to regulatory issues or uncommon capital structure in these industries. It is also required that the acquirer have all the dependent and independent variables available. The M&A sample consists of 411 completed M&A transactions. The full sample covers all S&P 1500 companies with 14086 observations having non-missing variables of interest, governance measures, and acquirer characteristics. To account for outliers, the sample is winsorized at the 1% level depending on the variables. Variable definitions, construction, and sources can be found in the Table 1.

Table 7. Annual Statistics of Governance Measures

The table reports the annual median of the governance measures “independent board”, “board size”, and “CEO duality” for the period 2002 - 2016. If it involves an indicator variable, mean statistics will be used instead of the median. All variables are recorded at the year-end preceding the acquisition announcement. The M&A sample covers companies who engaged in M&A activities. To be included in the dataset, the transaction has to be (1) completed, (2) between two companies listed in the U.S., (3) the bidding companies acquires more than 50% of the target’s shares, (4) the deal value is $10 million or more (5) the acquirer or target is not a subsidiary. The dataset will exclude financial and utility firms due to regulatory issues or uncommon capital structure in these industries. It is also required that the acquirer have all the dependent and independent variables available. The M&A sample consists of 411 completed M&A transactions. The full sample covers all S&P 1500 companies with 14086 observations having non-missing variables of interest, governance measures, and acquirer characteristics. To account for outliers, the sample is winsorized at the 1% level depending on the variables. Variable definitions, construction, and sources can be found in the Table 1.

Varriable name Mean Std. Dev. Min Median Max Mean Std. Dev. Min Median Max

Independent board 76% 13% 10% 80% 93% 75% 14% 0% 78% 95%

Board size 9.5 2.3 5 9 19 9.0 2.2 3 9 21

CEO duality 54% 50% 0.00 1.00 1.00 55% 50% 0.00 1.00 1.00

M&A sample (N = 411) Full sample (N = 14086)

Year Indep. Board Board Size CEO Duality Indep. Board Board Size CEO Duality

2002 71% 8 58% 67% 9 66% 2003 70% 8 63% 67% 9 67% 2004 75% 8 82% 71% 9 65% 2005 75% 10 69% 71% 9 65% 2006 73% 10 56% 73% 9 62% 2007 71% 10 58% 75% 9 57% 2008 82% 9 24% 80% 9 43% 2009 83% 11 58% 80% 9 54% 2010 83% 9 65% 80% 9 53% 2011 88% 9 65% 80% 9 53% 2012 86% 9 48% 82% 9 51% 2013 79% 10 47% 82% 9 49% 2014 88% 10 42% 83% 9 48% 2015 87% 10 32% 83% 9 45% 2016 82% 10 30% 83% 9 42%

(29)

29 Graph 4. Annual Median Statistic Independent Board

This graph shows the annual median statistics of “independent board”. Specific information can be found in Table 7. Variable definitions, construction, and sources can be found in Table 1.

Graph 5. Annual Mean Statistic CEO Duality

This graph shows the annual mean of “CEO duality”. Specific information can be found in Table 7. Variable definitions, construction, and sources can be found in Table 1.

Table 8 shows the descriptive statistics of the acquiring company’s characteristics (i.e. “free cash flow”, “cash holding”, “leverage”, and “market value") for the “M&A sample” and “full sample”. The full sample covers all S&P 1500 companies with non-missing variables of interest, governance measures, and acquirer characteristics, whereas the M&A sample covers the S&P 1500 that have engaged in M&A activities with non-missing dependent and independent variables. Table 9 covers the same variables and samples but on an annual basis.

“Free cash flow” is observed to be higher for the M&A sample as compared to the full sample. The mean (median) “free cash flow” for companies engaged in M&A activities is found to be 13% (13%), and 11% (11%) for that of the S&P 1500 companies (i.e. full sample). This observation is consistent over the entire sample period with the exception of 2011 (see Graph 6). Therefore, it can be suggested that companies pursuing mergers or acquisitions tend to have higher free cash flow as compared to a

50% 60% 70% 80% 90% 100% 2002 2004 2006 2008 2010 2012 2014 2016 Independent Board

0% 15% 30% 45% 60% 75% 90% 2002 2004 2006 2008 2010 2012 2014 2016 CEO Duality

(30)

30

typical company in the S&P 1500. This is in line with the Jensen’s free cash flow theory (1986), which suggests that managers in companies with a high level of free cash flow are more likely to pursue their own objectives which could be detrimental to shareholder value. In this context, executives with access to a higher level of free cash flow are more likely to pursue mergers or acquisitions in order to manage a bigger and therefore more prestigious company (empire building).

Based on the same theory on free cash flow (Jensen, 1986), debt will have the opposite effect, such that company executives would refrain from overinvesting in mergers and acquisitions. As debt involves mandatory repayments and interest costs, free cash flow is reduced, which in effect limits managerial discretion (Masulius et al., 2007). In line with this theory, Table 8 shows a lower leverage position for the M&A sample as compared with that of the full sample. An alternative explanation could be that firms with a lower leverage position are better able to pursue acquisitions due to their financial capacity to take on additional debt.

The variable “cash holding” is observed to be higher for the M&A sample as compared with that of the full sample, with a mean (median) cash holding of 19% (14%) for the former, versus 16% (11%) for the latter. Such an observation relates to the method of payment for an acquisition. From Table 2, it is observed that companies are more likely to utilise cash rather than equity to finance M&A transactions. Therefore, it can be said that for the M&A sample, a higher cash holding in the year preceding the acquisition can be explained by the use of cash to finance the acquisition.

The final variable is the “market value of the acquiring company”. The mean (median) of the “market value for companies engaged in M&A activities” is $41.8 billion ($11.1 billion), whereas that for the S&P 1500 companies (i.e. full sample) is $14.2 billion ($2.9 billion). From the table, the size of the acquiring companies is observed to be significantly higher than that of the typical S&P 1500 company. Although this difference fluctuates over the sample period (see Graph 9), the size of the average S&P 1500 company does not exceed that of the typical acquiring company.

Equity incentives, block ownership and returns to mergers