“CEO reputation and its effects on the investment decisions of the firm:
evidence from Mergers & Acquisitions”
Student Name: Nikoleta Koimtzi
Student Number: 11374845
MSc. Business Administration, Strategy Track
University of Amsterdam, Faculty of Economics and Business
Supervisor: Hesam Fasaei
Erasmus University Rotterdam, Rotterdam School of Management Date of submission: 22nd of June 2017
STATEMENT
OF
ORIGINALITY
This document is written by student Nikoleta Koimtzi who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
TABLEOFCONTENTS
STATEMENT OF ORIGINALITY ... 2
ABSTRACT ... 4
1. INTRODUCTION... 5
2. LITERATURE REVIEW ... 9
2.1. WHAT CEO REPUTATION IS ... 9
2.2 REPUTABLE CEOs CHARACTERISTICS AND INVESTMENT DECISIONS ... 10
2.3 THEORETICAL FRAMEWORK ... 12
2.3.1. Reputed CEO’s Overdominance & M&A: Reputed CEOs Excessive Control Over Free Cash Flow... 12
2.3.2 Reputable CEOs Personal - Social Characteristics & M&A ... 14
3. DATA AND METHOD ... 20
3.1 SAMPLE AND DATA COLLECTION ... 20
3.2 MEASURES ... 22
3.2.1 Independent Variable: CEO reputation: The media-coverage proxy (rpCEO) ... 22
3.2.2 Dependent Variable: Mergers and Acquisitions... 23
3.2.3 Control Variables... 24
4. RESULTS ... 25
4.1 TESTS FOR CEO REPUTATION’s PROXY ... 25
4.1.1 A four-year proxy of CEO reputation ... 25
4.1.2 Validity tests ... 26
4.2 DESCRIPTIVE STATISTICS – CORRELATION ... 30
4.2.1 Descriptive Statistics and Correlation analysis – rpCEO, SumUSnews and SumINTnews ... 31
4.2.2 Descriptive Statistics and Correlation analysis among the variables of this study 31 4.3 REGRESSION ANALYSIS ... 36
5. DISCUSSION - CONCLUSION ... 40
5.1 ACADEMIC RELEVANCE ... 41
5.2 MANAGERIAL IMPLICATIONS ... 43
5.3 LIMITATIONS AND FUTURE RESEARCH ... 44
6. REFERENCES ... 46
LIST OF FIGURES & TABLES Figure 1. Conceptual Model………..…..12
Table 1. Correlation between rpCEO and 4rpCEO….………26
Table 2. Descriptive statistics and correlations for Tenure, Prior Position, Age …………...28
Table 3. Descriptive statistics for positive and negative tone of media coverage proxy...29
Table 4. Recognition of CEO in annual lists (rpCEOLists)……...………..30
Table 5. Descriptive statistics and correlations among media coverage proxies…...…..31
Table 6. Means, Standard Deviations & correlations among variables…………...……...32
Table 7. Hierarchical Regression Model for ACQ…………..………37
Table 8. Hierarchical Regression Model for CASH………38
ABSTRACT
This study examines the relation between reputed Chief Executive Officers’ (CEO) characteristics and Mergers and Acquisitions (M&A) decisions. It provides evidence that CEO reputation has an effect on the number, the method of payment and the type of M&A conducted by the same company. Reputed CEOs’ overdominance and personal and social characteristics such as overconfidence, risk-taking behavior, social comparison and visibility bias are the underlying mechanisms that affect reputed CEOs’ M&A decisions. The study is conducted for the period of 2002-2010 and it includes 244 CEOs in 196 US companies. The major finding of this study is that CEO reputation has a positive relation with the number of M&A, all-cash M&A and diversifying M&A that a company conducts. It might explain some motives behind M&A decisions, even though there are other factors that can affect more these decisions. This study contributes to the existing literature that examines reputed CEOs’ attributes and corporate decisions and aims to provide more evidence about CEO’s reputation-building motives and their impact on the investment decisions of a company. The findings of this study can also raise the awareness of reputed CEOs and the board of directors about their investment decisions and governance processes.
Key words: CEO reputation; reputed CEO’s characteristics; Mergers and Acquisitions; all-cash acquisitions; diversifying acquisitions
1. INTRODUCTION
“Who makes the decision matters”-Reger (1997: 221). As Ranft, Zinko, Ferris, and Buckley (2006) mention, the role of Chief Executive Officer (CEO) is more than supervising. It is the embodiment of the company itself. A CEO is responsible for the successes and failures of the firm and has long lasting, but at the same time, immediate and tangible effects on the firm. CEOs are the key individuals presiding over investment, financing, and operating decisions (Bertrand and Schoar, 2003).
An emerging stream of research has been examining the impact of CEO reputation on corporate decisions of a firm, such as the association between high CEO reputation and earnings quality (Francis, Huang, Rajgopal, and Zang, 2008), corporate risk taking (Liu, Zhang and Jiraporn, 2011), capital investments (Jian and Lee, 2011) and dividend payouts (Likitratcharoen, 2012). Additionally, studies have found that CEOs who receive prestigious awards, as the reputable ones, are more prone to manage earnings in order to meet performance expectations, and, more controversially, are also prone to develop hubris, to underperform and to overinvest (Malmendier and Tate, 2008; 2009).
According to prior studies, there are some attributes of reputable CEOs that affect their investment decisions and might lead them to overinvest (Cartwright and Cooper, 1992; Gamache and Steinbach, 2015; Hirshleifer, 1993; Karuna, 2010; Liu, Zhang and Jiraporn, 2011). Firstly, reputable CEOs have been associated with overdominance (Karuna, 2010). Overdominance has a significant impact on the probability of a CEO taking more opportunistic decisions that benefit his position but worsen the firm’s value (Karuna, 2010). Secondly, reputed CEOs have been associated with risk-taking behavior (Liu, Zhang and Jiraporn, 2011), “empire-building” motives (Cartwright and Cooper, 1992; Hirshleifer, 1993; Jian and Lee, 2011), overconfidence (Gamache and Steinbach, 2015; Liu, Zhang and
Jiraporn, 2011), social comparison (Cartwright and Cooper, 1992; Hirshleifer, 1993) and visibility bias (Hirshleifer, 1993).
These studies have investigated, to a certain extent, the relationship between CEO reputation and investment decisions, even though they have given barely any attention to the impact of reputable CEOs’ characteristics on Mergers and Acquisitions (M&A) (Gong and Guo, 2014; Mohamed, Souissi, Baccar and Bouri, 2014). On the basis of this research, it is possible that such personality characteristics of reputed CEOs might affect their M&A decisions.
Although, M&A are among the most important investment decisions of a company (Chikh and Filbien, 2011; Malmendier and Tate, 2008), taking into account several studies (Andrade, Mitchell, and Stafford 2001; Fuller, Netter, and Stegemoller 2002; Jensen and Ruback, 1983; Mulherin and Boone, 2000) a great percentage of M&A do not provide any value-created for shareholders (Chikh and Filbien, 2011) and have a neutral or negative effect on shareholders’ wealth of acquiring firms (Doukas and Petmezas 2007; Wuebker, 2015). Behavioral corporate finance gives a possible explanation. It supports that managers are not fully rational and they have psychological and emotional biases (Mohamed et al., 2014) that affect their investment decisions. These biases are intensified when CEOs have a strong media profile (Gao and Sudarsanam, 2005) and when they are subject to social comparison, as reputable CEOs tend to be (Gamache and Steinbach, 2015). Thus, further research on the relation between CEO reputation and M&A is important, as it will provide some more evidence about M&A motives and it will shed more light on the conflict of interest between managers and shareholders in M&A decisions (Brown and Sarma, 2007; Jensen, 1986).
In this paper, I intend to investigate further this relation by examining the underlying mechanisms-characteristics of reputed CEOs that affect M&A decisions, such as the number, their method of payment (All-Cash M&A) and their type of M&A (Diversifying M&A). I
have classified these underlying mechanisms in two main categories: 1) reputed CEOs’ overdominance (greater power in corporate decisions and control over Free Cash Flow) and 2) reputed CEOs’ personal and social characteristics (risk-taking behavior, overconfidence, higher career concerns, social comparison and visibility bias). The aim of this thesis is to provide in-depth understanding about the role of reputed CEOs characteristics in M&A decisions by answering the following question: What is the effect of CEO reputation on the number, the type and the method of payment of M&A?
First of all, by bridging the gap between CEO reputation and its effects on M&A, this study aims to contribute to the ongoing discussion about how individual managers’ characteristics matter for corporate decisions (Brow and Sarma, 2007; Cartwright and Cooper, 1992; Croci and Petmezas, 2015; Francis et al. 2008; Hutton, Jiang and Kumar, 2010; Malmendier and Tate, 2008 etc).
Secondly, the study further aims to contribute to the existing literature that examines CEO reputation and investment decisions. As Lafond (2008: 150) emphasizes in the discussion of the paper “CEO Reputation and Earnings Quality” by Francis et al. (2008), there is a need for further investigation of the relation between CEO reputation and investment decisions such as M&A in order for other studies such as “CEO reputation and earnings quality” to reach further. More to the point, Hirshleifer (1993) in his article “Managerial reputation and Corporate Investment Decisions” starts a discussion about “manipulation” and “distortions” of investment decisions associated with reputation building. He analyses the fact that a large variety of managerial investment policies are driven by the managers’ desire to build and maintain their reputation. The aim of this study is to contribute to this discussion by providing more evidence.
Thirdly, the study effectively adds more insights to the existing literature that examines the relation between CEO reputation and corporate risk taking (Liu, Zhang and Jiraporn, 2011; 2016), as it investigates the impact of CEO reputation on the number of diversifying and all-cash acquisitions; investments that have a greater percentage of risk (Furfine and Rosen, 2009; Hagendorff and Vallascas, 2011; Park, 2002).
Fourthly, the study focuses on the agency problem in investment decisions, which is a matter of great concern in the existing literature (Jensen, 1986; Morck, Shleifer and Vishny, 1990). Given the well-documented presence of substantial agency conflicts in M&A, the closer examination of the relation between CEO reputation and M&A will add further insights to the agency problems and to the causes and motives of M&A (Amihud and Lev, 1981; Croci and Petmezas, 2015; Haleblian, Devers, McNamara, Carpenter and Davison, 2009; Jensen, 1986; Morck, Shleifer and Vishny, 1990).
Finally, considering the scope of its focus, the present study will be useful for managers and for the board of directors. On the one hand, it will raise the awareness of the board of directors with regard to the monitoring system and the overall control that is exercised over reputed CEOs. On the other hand, it will motivate reputed CEOs to re-examine their investment choices and understand better their own motives leading to their M&A decisions.
The structure of this study is organized as follows: The next section examines prior literature, constructs the theoretical framework and develops the hypotheses. Section 3 describes the sample, the data and the measurement of the variables used for testing the hypotheses. Section 4 reports the results of the analysis and, finally, Section 5 summarizes the major conclusions, limitations and contributions of the study, as well as suggesting some crucial points for future research.
2. LITERATURE REVIEW
An emerging stream of studies examines the effect of CEO reputation on firms’ investment decisions. This thesis falls within this category and particularly concerning the subject of reputed CEO characteristics and M&A. This section starts with the examination of what CEO reputation is and reviews prior literature that has associated CEO reputation with investment decisions. Afterwards, the theoretical framework is constructed and the hypotheses are formulated based on the impact of CEO reputation on the number of M&A, their method of payment and their type.
2.1. WHAT CEO REPUTATION IS
CEO reputation is a manager-specific attribute with an economically meaningful impact on corporate outcomes (Likitratcharoen, 2012). It constitutes a set of perceptions that major stakeholders form, based on perceived CEOs’ performance, their abilities and values. It captures the manner in which parties external to the firm view CEOs (Lafond, 2008). It is a shared perception by others of a person’s attributes or behavior (Ranft et al., 2006). Accounting, finance, management and marketing literature have cumulatively examined the definition of CEO reputation as an intangible asset for the firm (Liu, Zhang and Jiraporn, 2011). CEO reputation has a multidimensional nature and incorporates perceived characteristics that are difficult to measure such as credibility, charisma, integrity, honesty, and vision (Francis et al., 2008). CEOs’ tasks are characterized by ambiguity and complexity and, focusing on this, several studies have analysed the ways in which managerial perceptions, experiences, ego, biases and managerial style have an effect on the corporate decisions of firms (Bertrand and Schoar, 2003; Carpenter, Geletkanycz, and Sanders, 2004; Francis et al., 2008; Reger, 1997). CEO's positive reputation plays a major role in generating market capitalization for a company (Jin & Yeo, 2011), it can influence the
expectation of stakeholders about firm performance (Liu, Zhang, and Jiraporn 2011; 2016) and thus it should be treated as an important long-term asset (Jin and Yeo, 2011).
2.2 REPUTABLE CEOs CHARACTERISTICS AND INVESTMENT DECISIONS Despite the fact that ostensible CEO reputation has a positive impact on the firm, there is an emerging stream of studies that examines the negative effects of CEO reputation on corporate decisions (Francis et al. 2008; Liu, Zhang, and Jiraporn 2011; Karuna, 2010; Likitratcharoen, 2012). Prior literature has been examining the rent-extraction hypothesis, according to which reputable CEOs are more likely to use their discretion to act opportunistically and, contrary to the firm’s interests, they might possess personality characteristics that lead them to make decisions that do not have a positive effect on the firm (Francis et al., 2008; Liu, Zhang, and Jiraporn, 2011; Karuna, 2010; Likitratcharoen, 2012).
Firstly, recent studies have associated CEO reputation with lower monitoring and greater discretion on firm decisions (Karuna, 2010). Lower monitoring stems from shareholders’ perception that the higher the reputation of CEOs is, the greater their abilities are. Consequently to lower monitoring, reputable CEOs might have greater power in the process of decision-making of a firm (Karuna, 2010). That is, they can impose their will on others, as a result of exerting greater control than the less reputable CEOs (Brown and Sarma, 2007; Haleblian and Finkelstein, 1993; Karuna, 2010).
Another line of reasoning of the rent-extraction hypothesis is that reputable CEOs increase company risk-taking through investment decisions (Liu, Zhang, and Jiraporn, 2011). It is CEO visibility in the media that can increase CEO’s attribution bias and, as a consequence, reputable CEOs might underestimate the risks of projects. Studies have found that CEOs with flattering media profiles are more prone to developing managerial hubris (Gao and Sudarsanam, 2005) and becoming overconfident about the efficacy of their past
actions and future abilities (Hayward, Rindova, and Pollock 2004). More reputable CEOs might invest significantly more in R&D and adopt more leverage that leads to higher firm risk-taking (Liu, Zhang, and Jiraporn, 2011). This risk-taking behavior leads shareholders to require higher rates of return and as a result, firms experience higher costs of debt (Liu, Zhang, and Jiraporn, 2011).
CEO overconfidence, which may lead managers to make suboptimal decisions, has been examined by prior literature (Karuna, 2010; Malmendier and Tate, 2005; 2008). Prior literature has associated overconfidence with overinvestment, focusing on practices such as the tendency for high managerial acquisitiveness or undertaking riskier investments, such as diversifying acquisitions (Brown and Sarma, 2007; Doukas and Petmezas, 2007; Malmendier and Tate, 2005).
Finally, some studies have found that reputable CEOs may use investment choices -that are detrimental for shareholders but beneficial for them- as a means for building and maintaining their personal reputation (Hirshleifer, 1993; Malmendier and Tate, 2009). A possible explanation is that by increasing their reputation, CEOs manage to increase their prestige and thus they can achieve better bargaining power for their salary (Hirshleifer, 1993).
As a result, some reputed CEOs’ characteristics might have a significant effect on the investment decisions of the company. In order to examine the effect of CEO reputation on M&A decisions, I classified reputed CEOs characteristics in two underlying mechanisms: 1. Reputable CEOs’ overdominance and 2. reputable CEOs’ personal and social characteristics. These are the underlying mechanisms of reputable CEOs that ultimately have a crucial role in M&A decisions and more precisely in the number, the method of payment and the type of M&A.
2.3 THEORETICAL FRAMEWORK Figure 1. Conceptual Model
H1 H2 H3
2.3.1. Reputed CEO’s Overdominance & M&A: Reputed CEOs Excessive Control Over Free Cash Flow
Reputed CEOs overdominance in corporate decisions is an aspect that offers us further insights about M&A decisions. In general, the board and the management team decide whether a company will conduct a M&A, which are the criteria of the target (related, unrelated) and which is the method of payment for a M&A. However, when a CEO has greater control over the firm, he also has greater impact on these decisions. Reputed CEOs have more control and autonomy in taking corporate decisions and are subject to lower monitoring due to weaker governance (Karuna, 2010). This might occur because reputed CEOs are associated with prior successes and these successes reinforce CEOs’ authority in the company (Gao and Sudarsanam, 2005). In actual fact, boards do not place too many restrictions to reputed CEOs as they overestimate their ability (Brau and Fawcett, 2006; Boot, Gopalan, and Thakor 2006; Karuna, 2010; Prendergast, 2002; Raith, 2008). Boards judge
CEO
reputation
M&A number All-cash M&A Unrelated M&A Underlying mechanisms▪ Reputed CEO’s dominance
CEOs abilities upon CEOs’ prior performance. Consequently, reputed CEOs with greater control and lower monitoring can more easily engage in opportunistic actions (Brown and Sarma, 2007; Karuna, 2010) and take decisions that are detrimental for shareholders’ interests (Karuna, 2010; Prendergast, 2002) but at the same time beneficial for them.
Another line of reasoning concerns the information asymmetry about CEOs’ ability affecting shareholders’ perceptions (Karuna, 2010). This might happen because shareholders rely on CEOs’ reputation in order to measure how their respective stakes are protected (Karuna, 2010). Intentionally or not, the overdominance of reputed CEOs and the excessive power they have, lead them to undertake more risks (Francis et al., 2008), and as a result they overinvest by making more and riskier acquisitions. If CEOs believe that a specific M&A will give them the opportunity to extract rents from shareholders through higher compensation and bonuses, they might choose this M&A that maximizes their own wealth rather than the shareholders’ value (Grinstein and Hribar, 2004). This is further emphasized given the fewer rejections or the less resilience that they face from the stakeholders.
Additionally, taking into consideration that reputed CEOs have assumed greater power, they consequently exert more control over the free cash flow. Prior studies mention that managers manipulate cash flows over time to improve their reputations (Hirshleifer, 1993). Overconfident CEOs as reputed CEOs might be (Liu, Zhang and Jiraporn, 2011) -have high acquisitiveness, especially given abundant internal resources such as high free cash flow (Malmendier and Tate, 2008). As Shleifer and Vishny (1988) mention “Making acquisitions is often just the quickest and easiest way for managers to expand the scope of their control by directing the firm's cash flows into new ventures”. As a consequence, the greater control reputed CEOs exercise on corporate liquidity (Pawlina and Renneboog, 2005), the more acquisitions they will make, in order to leverage personal benefits, such as prestige and visibility. Besides, in firms with high free cash flow, reputed CEOs might affect to a
greater extent the decisions concerning the deployment of cash and they might be more prone to pursuing an empire-building strategy (Croci and Petmezas, 2015; Grossman and Hart, 1983) and to overinvesting (Jensen, 1986).
2.3.2 Reputable CEOs Personal-Social Characteristics & M&A
CEO reputation, Risk taking & Overconfidence
Liu, Zhang and Jiraporn (2011) found that reputed CEOs adopt aggressive risk-taking policies and that they tend to increase firm risk through their investment policies. Furthermore, recent study adds that CEOs with risk-taking incentives are more likely to invest in acquisitions (Croci and Petmezas, 2015).
Reputed CEOs’ risk-taking behavior can be explained by focusing on the role of the managerial hubris, or otherwise self-attribution bias (Doukas and Petmezas, 2007; Liu, Zhang and Jiraporn, 2011; Roll, 1986). The recent success of CEOs, the media praise for them and their self-importance are the three sources of CEO hubris examined by Hambrick and Hayward (1997). That is, reputable CEOs are associated with a record of successes before becoming well-known. Media favors successful CEOs and often presents them as “heroic” (Gao and Sudarsanam, 2005). A high and flattering media profile influences reputable CEOs’ self-image but also their self-esteem (Gao and Sudarsanam, 2005). “The more that others provide an individual with attributional accounts, the more likely it is that the individual will adopt the view expressed by others” (Hayward, Rindova, and Pollock, 2004).
Meindl, Ehrlich, and Dukerich (1985) conducted a research in the Wall Street Journal for 10 years and found that journalists attribute firm and industry performance to CEOs’ ability (Hayward, Rindova, and Pollock, 2004). As a result, CEOs convince themselves of
being the architects of the firm’s success (Brown and Sarma, 2007; Gao and Sudarsanam, 2005; Malmendier and Tate, 2008). They have so much faith in the efficacy of their leadership that they overestimate their own ability to manage a risky project such as a diversifying acquisition (Brown and Sarma, 2007; Gao and Sudarsanam, 2005; Malmendier and Tate, 2008). In other words, they overestimate their ability to control the project outcome and they underestimate the odds of failure (Liu, Zhang and Jiraporn, 2011).
Additionally, by developing managerial hubris, reputed CEOs are more prone to insisting on past strategies that gave them their glory as, for example, a prior big M&A that resulted in great publicity for the firm and the CEO. Hayward, Rindova and Pollock (2004) argue that reputed CEOs are more likely to be committed to the strategic choices that turned them into a celebrity; an aspect that leads to managerial inertia. As a result, there might be a decrease in their managerial adaptability to recognize new competitive demands, so the firm relies on the strategic choices that had initially increased the reputation of their CEO (Hayward, Rindova and Pollock, 2004).
In general, overconfidence is positively related with overinvestment, such as multiple acquisitions conducted by the same firm (Doukas and Petmezas, 2007; Malmendier and Tate 2008). The high managerial acquisitiveness is explained by the cognitive biases mentioned above, which lead reputed CEOs to engage in more complicated tasks-investments (Doukas and Petmezas, 2007).
CEO reputation, Career concerns & empire building motives
Reputable CEOs have more incentives to make a greater number of acquisitions or acquisitions that are unrelated due to their high career concerns about their professional reputation. More specifically, higher CEO reputation is positively associated with larger firms or with firms that have a wide range of businesses. As Hirshleifer (1993) argues,
managers that care about their reputation have incentives to use investments as a tool for building their personal reputation. These “reputation-building” motives (Gamache and Steinbach, 2015; Hirshleifer, 1993; Jian and Lee, 2011) lead managers to focus on short-term profit and distort investment decisions such as myopic investments (Narayanan 1985; Jian and Lee, 2011) or herding of investments that lead to overinvestment (Holmstrom and Costa, 1986; Jian and Lee, 2011; Trueman, 1986; Scharfstein and Stein, 1990). More precisely, CEOs who care about the increase of their reputation, follow an “empire building” strategy (Pawlina and Renneboog, 2005) and might increase the size and scope of the company beyond the point that maximizes shareholder wealth (Avery, Chevalier, and Schaefer 1998; Grossman and Hart, 1983; Pawlina and Renneboog, 2005). Behind this agency problem lies the fact that CEOs in large firms receive higher pecuniary and nonpecuniary managerial benefits such as higher compensation, and of course, prestige (Conyon and Murphy, 2000; Dyck and Zingales; 2004, Jensen, 1986; Renneboog and Trojanowski, 2004; Pawlina and Renneboog, 2005). As a result, their corporate objective focuses on growth rather than on the value maximization (Pawlina and Renneboog, 2005).
CEO reputation, Social comparison & shareholders’ expectations
Due to the fact that reputed CEOs are persons with a well-known media profile, they tend to engage more frequently in social comparison (Gamache and Steinbach, 2015). Social comparison with other CEOs’ standing in the business community in turn affects their investment decisions (Avery, Chevalier, and Schaefer, 1998; Gamache and Steinbach, 2015). A principal reason for that is their fear of losing their reputation. As a result, they try to remain ahead of their peers by growing their firm and follow “empire building” strategies (Pawlina and Renneboog, 2005). This is consistent with the concepts of mimicry and avoidance (Hirshleifer, 1993). Hirshleifer (1993) argues that mimicry leads to investment biases and distortions - such as overinvestment. In other words, firms with high quality,
having a good reputation, overinvest in order to avoid being characterized as lower-quality firms and essentially mimic the investment strategy of their peers in order to maintain their reputation (Hirshleifer, 1993). Consistent with the above statements is CEO’s fear of obsolescence (Cartwright and Cooper, 1992; Levinson, 1970). According to some CEOs’ perspective, CEOs who survive are “highly visible men of action-people that are recognized because they always look for new opportunities” (Cartwright and Cooper, 1992). They use M&A as a means to overcome their fear for continuous survival and as a way to enhance their credibility and confidence (Cartwright and Cooper, 1992). By doing M&A, they show to the business community that they are able to discover new opportunities and to exploit adequately the investment potential of their company.
Additionally, shareholders’ expectations play a significant role (Hirshleifer, 1993). If a firm invests at a lower scale, investors will assume that the firm has poor investment opportunities (Hirshleifer, 1993). Through acquisitions, reputable CEOs can prove their managerial abilities and meet shareholders’ expectations. For example, an acquisition that yields short-term profits improves the perception of shareholders about CEOs’ abilities, reputation and compensation (Narayanan, 1985) and about CEOs’ skills at identifying new investment opportunities.
CEO reputation & Visibility bias
Weber Shandwick, in partnership with KRC Research (2015) conducted a survey of more than 1.700 executives and found that 81% of executives reported that it was highly significant for CEOs to have a visible public profile. One of the incentives of CEOs to build and maintain their reputation is the visibility bias (Hirshleifer, 1993). “Visibility bias is defined as improving what is immediately visible, at the expense of what is not immediately visible” (Hirshleifer, 1993). Visibility of an investment gives an early reputational benefit to
CEOs who improve their reputation now (Hirshleifer, 1993). A company that often announces acquisitions or invests in new markets through diversification will have a great percentage of visibility both for the company and for the CEO. This aspect gives additional motives to reputed CEOs, who care about the maintenance of their reputation, to increase the number and the scope of acquisitions and exploit immediately this visibility.
Through the examination of reputed CEOs characteristics and their impact on M&A decisions this paper examines the following three hypotheses. Starting with the first hypothesis, as I mentioned above, reputable CEOs’ overdominance, their empire building motives, which stem from their high career concerns, and their risk taking behavior might lead the reputed CEOs to conduct a greater number of acquisitions than those CEOs with a lesser reputation. As a result:
Hypothesis 1: CEO reputation has a positive effect on the number of acquisitions that a company conducts
Additionally, taking into account that reputable CEOs are greater risk takers (Liu, Zhang and Jiraporn, 2011), I analyzed further their risk-taking behavior by examining the method of payment (cash or stock) and the type of M&A (related or unrelated) that reputable CEOs conduct.
During the period for which I am conducting this research (2002 - 2010), which includes the sixth merger wave (2003-2006), a much smaller percentage of M&A has been financed by stock as opposed to prior merger waves (Alexandridis, Mavrovitis, and Travlos, 2012). In general, all-cash M&A shareholders of the acquiring firm undertake the entire risk of the synergy value, which is embedded in the bid price and might not be materialized (Furfine and Rosen, 2009; Hagendorff and Vallascas, 2011). In an all-stock acquisition, the risk is shared with selling shareholders. The synergy risk is allocated in accordance with the
percentage of stake that the acquiring and selling shareholders have upon the new entity (Rappaport and Sirower, 1999). Subsequently, all-cash acquisitions are riskier investments than the all-stock ones. Having already analyzed the risk-taking behavior of reputed CEOs and their greater control in company’s liquidity / cash flows, I continue with the second hypothesis:
Hypothesis 2: CEO reputation has a positive effect on the number of all-cash acquisitions in a company
Finally, prior studies (Bergh, 1997; Croci and Petmezas, 2015; Park, 2002) argue that diversified M&A incorporate diversification risks. These risks make diversifying M&A riskier investment choices than the related acquisitions. They are more likely to be outside the company’s area of expertise and as a result managers might have less knowledge, information and network about the industry of the target (Croci and Petmezas, 2015). Additionally, future outcomes are less predictable and have greater uncertainty, since unrelated industries have deviations concerning the products, the markets and technologies (Park, 2002). Reputed CEOs’ risk-taking behavior and their overconfidence may lead them to underestimate the risks of an unrelated M&A and to conduct a greater number of unrelated acquisitions (Doukas and Petmezas, 2007; Liu, Zhang, and Jiraporn 2011; 2016). Moreover, the greater media visibility of these acquisitions, in combination with the enlargement of scope and size of the firm, constitute motives for reputed CEOs who want to maintain and increase their reputation.
Hypothesis 3: CEO reputation has a positive effect on the number of unrelated acquisitions in a company
3. DATA AND METHOD
3.1 SAMPLE AND DATA COLLECTIONThis study used secondary data gathered from Wharton CRSP, Thomson One SDC and Bloomberg databases. First of all, the study uses Standard and Poor’s (S&P) 500 companies over the period 2002-2010, as identified from the ExecuComp database. The selected period is based on statistics from the Institute of Mergers, Acquisitions and Alliances that show an increasing activity of M&A starting from 2002 until 2007, which is the peak and then a slight decrease until 2010. Alexandridis, Mavrovitis, and Travlos (2012) mention that from 2003 to 2006 the sixth merger wave took place. By starting my research in 2002 until 2010, my purpose is to examine this “M&A cycle”. The final sample consists of 1255 year-observations from 244 CEOs in 196 unique companies covering the period of these 9 years. (In detail, the range of sample observations is 156 for 2002, 154 for 2003, 150 for 2004, 147 for 2005, 140 for 2006, 145 for 2007, 132 for 2008, 125 for 2009 and 106 for 2010). To be included in the sample, data must be completed in the Execucomp database including the name of CEO, the company name and the ticker. Also the date that CEOs joined the company has to be mentioned (as employees or immediately as CEOs), the date that the person became a CEO and the date that this person left this position. This information was used to measure CEO tenure and CEO prior position in the company (Brown and Sarma, 2007; Francis et al., 2008; Milbourn, 2003 etc). Moreover, financial and utility companies were excluded from the sample. These companies are regulated and CEO’s investment biases are less likely to be as pronounced as in non-regulated firms (Doukas and Petmezas, 2007; Likitratcharoen, 2012). Similarly, CEOs with less than three years of tenure in the same company were excluded from the sample. An individual year of CEO reputation can be a noisy measure, since the reputation of a person is developing over several years in the same company (Francis et al., 2008). As a result, the minimum period of CEO tenure in the sample is three years.
Furthermore, I used Securities Data Corporation's (SDC) Mergers and Acquisitions Database, Bloomberg Database and Capital IQ to find the exact number of acquisitions that the 196 companies of the sample have conducted during the given period. Additionally, I gathered data about the method of payment (cash or stock) and whether acquisitions were related or unrelated to the core business of the company. I selected 1346 acquisitions from SDC and then I added 36 public acquisitions from Bloomberg database that were missing in SDC. As a result, the final number of acquisitions was 1382. The criteria to include an acquisition in the sample are: acquisitions have to be completed, the target is either a private or public firm (Brown and Sarma, 2007) and the acquirer has to purchase at least 50% of the target’s value as a result of the takeover (Doukas and Petmezas, 2007).
In order to test the three hypotheses of this study:
I used linear regression analysis in SPSS. The independent variable is CEO reputation (media-coverage proxy: rpCEO) and the dependent variables are: the number of M&A (ACQ) (hypothesis 1), the number of all-cash M&A (CASH) (hypothesis 2) and the number of diversifying M&A (UN_ACQ) in a given year-observation (hypothesis 3). Moreover, in order to isolate the effects of CEO reputation, I control for the compounding effects of company characteristics and other factors that have an effect on managers’ decision to acquire (Brown and Sarma, 2007; Malmedier and Tate, 2008). I include the number of employees for each year-observation as a control for firm size (Audia and Greve, 2006), Total q at the beginning of the year as a control for growth and investment opportunities, an indicator for efficient board size as a measure of corporate governance, cash flows normalized by total assets as a measure of different levels of available internal resources and the number of shares that the CEO holds in the given company divided by the total number of shares outstanding in the given year-observation to control for ownership incentive effects
(Brown and Sarma, 2007; Croci and Petmezas, 2015; Gan, 2015; Malmendier and Tate, 2008).
3.2 MEASURES
3.2.1 Independent Variable: CEO reputation - The media-coverage proxy (rpCEO)
According to prior studies (e.g Francis et al., 2008; Karuna, 2010; Milbourn, 2003) the use of media coverage can be a proxy for reputation. The greater the media prominence for the CEO, the more the CEO is perceived as a successful leader in the media. That is, CEO reputation is measured as a media-coverage proxy based on how parties external to the firm view the CEO (Francis et al., 2008; Milbourn, 2003). The media-coverage proxy is reflected in the number of articles containing CEO’s full name and company affiliation that appeared in major U.S. and global business newspapers (Francis et al., 2008; Jian and Lee, 2011; Milbourn, 2003). Following former studies (Francis et al., 2008; Jian and Lee, 2011; Karuna, 2010; Liu, Zhang and Jiraporn, 2011; 2016; Milbourn, 2003), I hand-collected all the related articles as returned from the Lexis-Nexis database. In this respect, an article is included if it has once the CEO’s full name and company name (Francis et al., 2008; Milbourn, 2003).
First of all, I searched related articles in the best-known US newspapers for each CEO, each year-observation. My search was carried out in the following newspapers: the New York Times, the Washington Post and USA today. Then, all the US articles summed-up for each year-observation (SumUSnews).
SumUSnews(t) = Number of articles in the New York Times(t) + Number of articles in the
Washington Post(t) + Number of articles in USA today(t)
Where (t) is the year-observation from 2002-2010
Afterwards, I searched related articles in International newspapers for each CEO and each year-observation. My search was carried out in the following newspapers: Financial
Times, the Guardian, the International Herald Tribune. Then, all the articles from international newspapers summed-up for each year-observation (SumINTnews).
SumINTnews (t) = Number of articles in Financial Times(t) + Number of articles in
Guardian(t) + Number of articles in the International Herald Tribune(t)
Where (t) is the year-observation from 2002-2010
Finally, the total number of articles (rpCEO) is the proxy of CEO reputation and it is the sum of all the US and International news for each year-observation.
rpCEO(t) = SumUSnews(t) + SumINTnews (t)
Where (t) is the year-observation from 2002-2010
3.2.2 Dependent Variable (s): Mergers and Acquisitions
Number of M&A (ACQ): To test the first hypothesis, the dependent variable is the number of acquisitions (ACQ). This variable is a continuous variable that indicates how many public and private acquisitions have been conducted in a given year-observation for each company.
Method of Payment (CASH): In order to examine whether CEO reputation positively affects the number of all-cash acquisitions, I collected from SDC all the acquisitions that were paid exclusively in cash. The variable cash is a continuous variable that indicates how many acquisitions, from the total number of acquisitions, were paid all-cash in a given year- observation.
Unrelated M&A (UN_ACQ): In order to examine the third hypothesis, I classified all acquisitions as unrelated and related. I hand-collected the SIC code of the main business for
each company - both for acquirers and targets. If the acquirer and the target had the same four-digit SIC code, the acquisition was classified as related (Croci and Petmezas, 2015; Graham, Lemmon and Wolf, 2002). In case that there is no common SIC code, the acquisition is classified as unrelated (UN_ACQ)(Graham, Lemmon and Wolf, 2002).
3.2.3 Control Variables
Apart from the dependent and independent variables, in order to isolate the effects of CEO reputation, I control for confounding effects of firms and CEOs characteristics. I gathered data for the control variables from the Wharton and SDC Databases.
Size (Size): As the size of the firm might affect the investment decisions within a firm (Brown and Sarma, 2007; Malmendier and Tate, 2008), I control for firm size. This control variable is constructed as the number of employees in a company in a given year-observation (Audia and Greve, 2006).
Peters and Taylor’s Total q as a control for growth (Total Q): Former studies (Brown and Sarma, 2007; Malmendier and Tate, 2008; 2009) measure Tobin’s Q as a control for growth. However, according to the latest reports of the Wharton Database, Peters and Taylor’s Q is an improved Tobin’s Q proxy that includes intangible capital in the denominator. Peters and Taylor (2016) show that Total q captures firms’ investment opportunities better than other popular Tobin’s q proxies. Peters and Taylor estimate the replacement cost of firms’ intangible capital by accumulating past investments in R&D and SG&A (Wharton Database).
Cash flow as a control for different levels of internal resources (Cash Flow): According to the Wharton Database, a researcher will gather information for the cash flow of a company with the variable (OANCFY). Cash flow represents the net profit after tax before abnormal items plus depreciation. The cash flow then is normalized by assets (divided by the book value of assets) (Brown and Sarma, 2007; Gan, 2015; Malmendier and Tate, 2008).
Shares held by CEOs as a control for ownership incentive effects (Ownership): CEOs’ stock and option ownership is included as a control for its incentive effects on managers. This variable is measured as the number of ordinary shares of the company in which the CEO has a beneficial interest, divided by the total number of shares outstanding (Brown and Sarma, 2007).
Efficient Corporate Governance (Governance): The number of directors on the board that indicates an efficient board size (Malmendier and Tate, 2008). It is a binary variable where 1 indicates that the board has between four and twelve members and 0 if the board has more than 12 members (Croci and Petmezas, 2015; Malmendier and Tate, 2008). It is possible that boards with fewer members have stronger monitoring mechanisms and board strength whereas larger boards are less efficient due to the higher coordination costs and free rider problems (Karuna, 2010).
4. RESULTS
The following section reports the results of this study. First of all, I conducted some tests in order to examine whether the media-coverage proxy (rpCEO) reflected significant aspects of CEO reputation. Secondly, the descriptive statistics for all the variables of this research are presented providing an overview of the collected data and then the important correlations are reported. Finally, the linear regression analyses were performed in order to test the three hypotheses.
4.1 TESTS FOR CEO REPUTATION’s PROXY
4.1.1 A four-year proxy of CEO reputation
As CEO reputation develops over in several years, individual year-observations of the press-coverage proxy may not reflect CEOs’ true reputation (Francis et al., 2008). In order to
see if my media-coverage proxy for CEO reputation for an individual year-observation (rpCEOt) captures CEOs accumulated stock of reputation, I followed the Francis et al, (2008) method and I constructed a four-year proxy for CEO reputation. I hand-collected all the related articles for each CEO for t-3 to t (t-1, t-2 and t-3) from each newspaper (New York Times, USA Today, the Washington Post, Financial Times, The Guardian, International Herald Tribune), which were in total 22.590 additional observations. Finally, by summing-up all the articles from these four years, I constructed the variable 4rpCEO. By testing the correlation between the individual year measurement (rpCEO) and the multiyear measurement (4rpCEO), I found a strong positive correlation, which is statistically significant between the two media coverage proxies (Table 1). More precisely, according to Pearson test the correlation between rpCEO and 4rpCEO is r = 0.898** p < 0.01. Similarly, according to the Spearman test the correlation is rs = 0.840** p < 0.01. As a result, the individual-year measurement (rpCEO) captures significantly a CEO’s reputation, which is consistent with the findings of Francis et al. study (2008).
Table 1. Correlation between rpCEO and 4rpCEO
Pearson rpCEO 4rpCEO
rpCEO 1 0.898**, p = 0.000
4rpCEO 0.898**, p = 0.000 1
Spearman rpCEO 4rpCEO
rpCEO 1 0.840**, p = 0.000
4rpCEO 0.840**, p = 0.000 1
**Correlation is significant at the 0.01 level (2-tailed) * Correlation is significant at the 0.05 level (2-tailed)
4.1.2 Validity tests
Additionally, I conducted three validity tests in order to confirm that the media coverage proxy (rpCEO) captures significant aspects of the reputation of CEOs. The first is based on
the examination of the correlation between the media-coverage proxy (rpCEO) and a CEO’s tenure, age and prior position in the company (Francis et al., 2008, Milbourn, 2003). The second test investigates whether the content of media coverage is favorable or not towards the CEO (Francis et al., 2008; Karuna, 2010). That is, as CEO reputation proxy is measured by the number of business related articles, it is necessary to be examined whether these articles are favorable with respect to the CEO (Francis et al., 2008). The third validity test examines whether press coverage has a positive correlation with the likelihood that the CEO names are mentioned in one of the “top CEO lists” (such as Worth’s list, Fortune’s list etc) in a given year t (Francis et al., 2008).
1st VALIDITY TEST: CEO tenure, Age and Prior Position
This validity test examines the correlation between the proxy of CEO reputation (rpCEO) and the tenure, age and prior position of the CEO. According to Milbourn, (2003) tenure, age and prior position are positively related to the market’s perception of the CEO ability.
CEO Tenure (Tenure) is a continuous variable that indicates the number of years that an executive is the CEO in a firm (Francis et al., 2008; Milbourn, 2003). Longer tenure means higher assessment of CEOs’ ability from the board, since CEOs will have remained in the company and they will have “survived” from previous dismissal decisions (Francis et al., 2008; Milbourn, 2003). Table 2 (A) shows that the mean of a CEO’s tenure is 7.07 years and the median is 5 years. Pearson tests show (Table 2 (B)) that there is positive and highly significant association between the media-coverage proxy for CEO reputation (rpCEO) and CEO tenure (r = 0.087, p = 0.002), whereas Spearman tests shows negative association that is not highly significant (rs = -0.009, p = 0.748).
CEO age (Age) measures the age of CEO for each year-observation t. I collected the CEO age from the ExecuComp in order to examine the correlation between the media coverage
proxy (rpCEO) and the age of CEO. According to descriptive statistics, shown in Table 2 (A), the average and median CEO age is 57 years. According to both tests, the correlation, shown in Table 2 (B), between CEO age and the media-coverage proxy (rpCEO) is positive but not statistically significant (Pearson: r = 0.046, p-value of 0.101, Spearman: rs = 0.038 with p-value 0.181).
Prior Position shows whether the CEO was appointed from within or outside of the firm. Prior position indicates that a CEO who is appointed from outside the firm is more likely to have higher perceived ability than an inside candidate (Francis et al., 2008; Jian and Lee, 2011; Milbourn, 2003). Prior Position is a dummy variable that equals to one if the executive became CEO at the same time he joined the company and equals to zero if he was a former employee. According to Table 2 (B) Pearson test shows significant association between the proxy of CEO reputation (rpCEO) and CEO’s prior position (r = 0.061, p = 0.031). Spearman test shows a positive but not significant association between them (rs = 0.046, p = 0.1).
Table 2 (A). Descriptive data for CEO Tenure, Age, Prior Position
Number of
observations Mean Median
St.
Deviation Minimum Maximum
Tenure 1255 7.07 5.00 6.48 0 39
Age 1255 57.16 57.00 6.61 41 89
Prior
Position 1255 0.27 0 0.443 0 1
Table 2 (B). Correlations among Tenure, Age, Prior Position & rpCEO
Number of observations rpCEO Pearson Spearman Tenure 1255 0.087** (p = 0.002) -0.009 (p = 0.748) Age 1255 0.046 (p = 0.101) 0.038 (p = 0.181) Prior Position 1255 0.061* (p = 0.031) 0.046 (p = 0.1)
**Correlation is significant at the 0.01 level (2-tailed) * Correlation is significant at the 0.05 level (2-tailed)
2nd VALIDITY TEST: Positive and negative tone of media coverage proxy
Studies have shown that press coverage is generally associated with favorable mentions of the CEO (Karuna, 2010; Milbourn, 2003). That is, CEOs are mentioned for their positive characteristics / actions in the press more often than they are mentioned for undesirable actions e.g., scandals (Francis, et al. 2008; Karuna, 2010; Milbourn 2003). In order to confirm if the articles that I have gathered to construct the media-coverage proxy have a positive tone towards CEOs, I followed the method of Francis et al. (2008). For each year-observation t, I chose randomly five CEOs and for each CEO I chose to analyze ten articles. I found the tone of each article by using Linguistic Inquiry and Word Count program (LIWC). LIWC provides coefficients for positive and negative emotions giving an indication whether the tone of the articles is positive or negative. In total, I analyzed 450 articles. The results, mentioned in Table 3, show that the majority of articles have a neutral to positive tone in respect to the CEO.
Table 3. Descriptive data for positive and negative tone of media coverage proxy
No of articles Positive Tone Negative Tone Neutral Total
Non-negative
450 75% 19% 6% 81%
3rd VALIDITY TEST: Recognition of the CEO in annual lists
The 3rd validity test examines whether the media coverage proxy for CEO reputation is consistent with the likelihood that the CEO is recognized as “top CEO” in calendar year t (Francis et al., 2008). The lists are: Top 25 Influential Business leaders (CNN), the Best Performing CEOs in the World (Harvard Business Review) and Forbes lists (2003-2009). The variable named rpCEOLists is a binary variable where 1 means that the CEO in the year t was at least in one list and 0 if he/she was not in any of these lists. Table 4 shows the results
from the logistic regression. The dependent variable is rpCEOLists and the independent is the media coverage proxy (rpCEO) (Francis et al., 2008). I controlled also for the size of the company (number of employees) and firm’s profitability (ROA: return on assets), as business press lists rate CEOs according to the size and the profitability of the firm (Francis et, al. 2008). According to the Table 4, there is a statistically significant positive association (chi-square = 10.66, p < 0.01) between rpCEOLists and rpCEO. Also, CEO’s appearance in these lists is associated with firms that are profitable (chi-square = 8.67, p < 0.01).
Table 4. Binary Regression - Recognition of CEO in annual lists (rpCEOLists)
B Chi-Square Pr > Chi-square
ROA 0.09 8.67 0.003
SIZE 0.00 0.17 0.68
rpCEO 0.03 10.66 0.001
Likelihood ratio 145.99 0.000
All in all, the results from the validity tests are consistent with Francis et al. (2008) research. The media-coverage proxy (rpCEO) can capture significant aspects of CEO reputation and as a result it can be used in the analysis of the three hypotheses.
4.2 DESCRIPTIVE STATISTICS – CORRELATION
Before doing my analysis, I tried to eliminate missing data. For the dependent variable of each hypothesis and the independent variable rpCEO, I did not have missing data, as I hand collected all the values. Moreover, in Wharton database there was complete information about the size of companies (number of employees) and the corporate governance (size of board). However, there was missing data for the normalized cash flows by assets, for Peters and Tailor’s q and for shares held by CEOs for some year observations. I tried to eliminate this missing data by hand-collecting information. For the data that I could not find information, I left the cell blank.
4.2.1 Descriptive Statistics and Correlation analysis – rpCEO, SumUSnews and SumINTnews
According to the Table 5 (A) that shows the descriptive statistics for the media proxy of CEO reputation, a CEO receives 5.78 mentions in the press (rpCEO). Of these, 3.01 are initiated by the US press and 2.77 are initiated by the International press. Table 5 (B) shows the pairwise correlations among rpCEO, SumUSnews and SumINTnews. There are high pairwise correlations (significant at 0.01 level (2-tailed)) among the three variables. Following Francis et al. 2008, I focus my attention only to the summary measure (rpCEO).
Table 5 (A). Descriptive statistics on proxy variable for CEO reputation
N Minimum Maximum Mean St.
Deviation Variance
rpCEO 1255 0 608 5.78 21.97 482.62
SumUSnews 1255 0 219 3.01 10.08 101.56
SumINTnews 1255 0 389 2.77 13.06 170.62
Table 5 (B). Correlations among rpCEO, SumUSnews and SumINTnews Pearson
rpCEO SumUSnews SumINTnews
rpCEO 1 0.934** (p = 0.000) 0.961** (p = 0.000)
SumUSnews 0.934** (p = 0.000) 1 0.799** (p = 0.000)
SumINTnews 0.961** (p = 0.000) 0.799** (p = 0.000) 1
Spearman
rpCEO SumUSnews SumINTnews
rpCEO 1 0.904** (p = 0.000) 0.823** (p = 0.000)
SumUSnews 0.904** (p = 0.000) 1 0.595** (p = 0.000)
SumINTnews 0.823** (p = 0.000) 0.595** (p = 0.000) 1
**Correlation is significant at the 0.01 level (2-tailed) * Correlation is significant at the 0.05 level (2-tailed)
4.2.2 Descriptive Statistics and Correlation analysis among the variables of this study
Table 6 (A) provides descriptive statistics and Table 6 (B, C) the bivariate correlation analysis of the variables of this study. The correlations were investigated using Pearson correlation coefficient (Table 6, Panel B) and Spearman correlation coefficient (Table 6, Panel C).
According to the Table 6 (Panel A), the mean for the media-coverage proxy of CEO reputation equals 5.78, which means that a CEO receives approximately 5.78 mentions in the press (rpCEO). Also, the mean for acquisitions equals 1.10, which means that a company is conducting approximately one acquisition for a given year-observation.
Table 6 (Panel A). Means, Standard Deviations
N Mean S.D. 1. rpCEO 1255 5.78 21.97 2. ACQ 1255 1.10 1.84 3. CASH 1255 0.35 0.77 4. UN_ACQ 1255 0.76 1.58 5. Cash flow 1162 0.13 0.073 6. Ownership 1183 1.29 3.94 7. Total Q 1183 1.64 2.27 8. Governance 1255 0.86 0.35 9. Size 1255 57.44 170.36
Starting with the dependent variable (ACQ) of the first hypothesis, both Pearson and Spearman tests (Table 6, Panel B & C) showed a statistically significant correlation between the number of acquisitions and the media-coverage proxy of CEO reputation (r = 0.120, p = 0.000; rs = 0.176, p = 0.000). Furthermore, both tests report a statistically significant positive correlation between the number of acquisitions (ACQ) and internal resources (Cash flow) (r = 0.095, p = 0.001, rs = 0.107, p = 0.000), which is consistent with prior studies that have found that cash-rich firms are positively associated with a greater number of acquisitions (Croci and Petmezas, 2015; Faccio and Masulis, 2005; Harford, 1999; Jensen, 1986). Furthermore, Spearman test showed a statistically significant negative correlation between CEO reputation and ownership incentives (rs = -0.185, p = 0.000). A possible explanation is that reputed CEOs might take more opportunistic decisions and thus they have lower ownership incentives towards the company. As concerns the size of the company, the correlation with the number of acquisitions (ACQ) is positive but not significant according to the Pearson
Test (r = 0.17, p = 0.545) and statistically significant according to the Spearman test (rs = 0.158, p = 0.000). The same happened with the correlation between Total Q and the number of acquisitions, where according to Pearson test there is a positive but not significant correlation (r = 0.031, p = 0.289) and according to Spearman test the correlation is both positive and statistically significant (rs = 0.094, p = 0.001). Variety in the results of the tests can be explained since the Pearson correlation examines the linear relationship between the two variables whereas the Spearman correlation examines the monotonic relation. This means that there is no significant linear correlation; the variables change together but not in a constant rate.
In regard to the dependent variable of the second hypothesis (CASH), the correlation between the number of all-cash acquisitions and CEO reputation’s proxy (rpCEO) is positive and statistically significant in both tests (Pearson: r = 0.064, p = 0.024; Spearman: rs = 0.125, p = 0.000), which means that CEO reputation has a positive effect on the number of all-cash acquisitions. Also, both tests report in Table 6 (Panel B, C) a statistically significant positive correlation between the number of all-cash acquisitions (CASH) and cash flows (r = 0.096, p = 0.001, rs = 0.101, p = 0.001). Additionally, the Spearman test shows a positive significant correlation between all-cash acquisitions and Total Q (rs = 0.069, p = 0.017). This means that growth opportunities have a positive impact on the number of all-cash acquisitions. Similar were the results between all-cash acquisitions and the size of company (rs = 0.07, p = 0.013).
Concerning the dependent variable of the third and last hypothesis (UN_ACQ) shown in Table 6 (Panel B, C), both tests report a statistically significant correlation between the number of unrelated acquisitions (UN_ACQ) and the proxy for CEO reputation (Pearson: r = 0.075, p = 0.008; Spearman: r = 0.134, p = 0.000). According to the Spearman test, there is a statistically significant positive correlation between the unrelated number of acquisitions and growth opportunities (Total Q) (rs = 0.087, p = 0.003) and with the size of the company (rs =
0.182 p = 0.000). Both tests report a negative but not statistically significant relation between the number of unrelated acquisitions and ownership, which means that managers might have some concerns for the efficacy of unrelated acquisitions.
All in all, both Pearson and Spearman tests report statistically significant correlations between the dependent variable of each hypothesis and the proxy for CEO reputation. The correlations in both tests indicate that reputed CEOs work for larger firms and for firms with higher investment opportunities. Also, both tests suggest that less efficient governance is associated with larger firms (Pearson: r = -0.270, p = 0.000, Spearman: rs = -0.268, p = 0.000) and with firms with more reputable CEOs (Pearson r = -0.122, p = 0.000; Spearman: rs = -0.238, p= 0.000), which is consistent with the argument that reputable CEOs have greater discretion and autonomy in decisions, as was discussed in the theory development (Karuna, 2010). Another explanation for this result is that CEO press coverage might act as an external monitoring system and thus it substitutes an important internal monitoring role for companies (Karuna, 2010; Miller, 2006)
Even though both tests showed some statistically significant results, the strength of correlations appear low in general. A possible explanation is that CEO reputation is not directly observable and is measured as a media-coverage proxy, as well as the fact that efficient governance and ownership are formulated from different data sources. As a result the measures report a significant but low correlation (Malmendier and Tate, 2008).
Table 6 (Panel B). Pearson Correlation 1. 2. 3. 4. 5. 6. 7. 8. 9. 1. rpCEO 1 2. ACQ 0.120** 1 3. CASH 0.064* 0.639** 1 4. UN_ACQ 0.075** 0.852** 0.570** 1 5. Cash flow 0.021 0.095** 0.096** 0.054 1 6. Ownership 0.062* 0.082** 0.019 -0.013 0.028 1 7. Total Q 0.000 0.031 0.050 0.035 0.257** 0.040 1 8. Governance -0.122** 0.001 0.017 0.033 0.017 0.031 0.067* 1 9. Size 0.251** 0.017 0.008 0.016 0.000 -0.041 -0.065* -0.270** 1
Table 6 (Panel C). Spearman Correlation
1. 2. 3. 4. 5. 6. 7. 8. 9. 1. rpCEO 1 2. ACQ 0.176** 1 3. CASH 0.125** 0.643** 1 4. UN_ACQ 0.134** 0.815** 0.549** 1 5. Cash flow 0.080** 0.107** 0.101** 0.055 1 6. Ownership -0.185** -0.016 0.032 -0.018 -0.014 1 7. Total Q 0.048 0.094** 0.069* 0.087** 0.396** 0.073* 1 8. Governance -0.238** 0.006 0.008 0.021 0.013 0.218** 0.071* 1 9. Size 0.469** 0.158** 0.070* 0.182** -0.061* -0.252** -0.180** -0.268** 1 **Correlation is significant at the 0.01 level (2-tailed)
4.3 REGRESSION ANALYSIS
In order to test the hypotheses, linear (hierarchical) regressions were performed. For each hypothesis I constructed a plot involving both variables (dependent and independent) in SPSS in order to make sure that the model fits the data and there is no curvilinear relationship.
Hypothesis 1 states that the greater the reputation of a CEO in a company, the greater the number of acquisitions that this company will conduct. Table 7 (Model 1, 2), shows the results of the two steps of the regression. In the first step I entered the control variables. The model is statistically significant F (5, 1.097) = 3.582; p < 0.01 and explained 2% (R-Square = 0.02) of the variance in the number of acquisitions. After the introduction of the independent variable a total 3% of the number of acquisitions is explained, F (6, 1.096) = 5.361; p < 0.001. In the final model three out of six predictor variables are statistically significant.
The proxy of CEO reputation has a higher beta value (β = 0.12, p < 0.001), than the cash flows (β = 0.09, p < 0.01) and shares held by CEOs (β = 0.07, p < 0.05). So a company’s internal resources and ownership incentive effects have a lower impact on the acquisitiveness of the company than CEO reputation. The results show statistical significance but weak support for hypothesis 1. The low R-square shows that the media-coverage proxy of CEO reputation can explain a small percentage of the variance in the number of acquisitions. A possible reason for low R-square is that the media-coverage proxy has high-variability data and as I mentioned in the correlation analysis is a proxy of CEO reputation and not CEO reputation per se. However, there is a significant trend. This trend shows that CEO reputation proxy can provide some information about the number of acquisitions even though data points fall further from the regression line. So there is a small but reliable relationship. Taking these aspects into account, the hierarchical regression showed a statistically significant positive relation between the number of acquisitions and the reputation of CEOs,
but CEO reputation can only explain a small percentage of the increase in the number of acquisitions. This means that there are other factors that affect the variance in the number of acquisitions, which play a greater role for the number of acquisitions than CEO reputation does.
Table 7. Hierarchical Regression Model for ACQ
R R2 R2 Ch. B SE β t Model 1 0.13 0.02** Cash Flow 2.32 0.78 0.09** 2.96 Ownership 0.04 0.01 0.08* 2.58 Total Q 0.02 0.03 0.02 0.50 Governance -0.01 0.17 -0.002 -0.05 Size 0.000 0.00 0.02 0.55 Model 2 0.17 0.03*** 0.012*** Cash Flow 2.25 0.78 0.09** 2.90 Ownership 0.03 0.01 0.07* 2.30 Total Q 0.01 0.03 0.01 0.45 Governance 0.02 0.17 0.00 0.13 Size 0.00 0.00 -0.01 -0.39 rpCEO 0.01 0.00 0.12*** 3.75 Statistical significance: *p<0.05, **p<0.01, ***p<0.001
Hypothesis 2 states that the greater the reputation of CEO, the greater is the number of all-cash acquisitions in a company. Table 8 (Model 1) shows the introduction of the control variables. In Table 8 (Model 2) CEO reputation proxy has been included. Model (1) is statistically significant F (5, 1.097) = 2.937; p < 0.05. After the introduction of the proxy for CEO reputation (rpCEO) a total 2% of the number of all-cash acquisitions (CASH) is explained, F (6, 1.096) = 5.036; p < 0.05 (Table 8, Model 2). CEO reputation proxy has a small but positive beta (β = 0.07, p < 0.05) - as was also shown in the correlation results - is statistically significant at 5% and normalized cash flows by assets at 1%. Both results indicate that CEO reputation and internal resources have almost the same effect in the
