To what extent does CEO overconfidence effect M&As and how is this moderated by board gender diversity and CEO duality: Evidence from the United States of America

(1)

To what extent does CEO overconfidence effect M&As and how is

this moderated by board gender diversity and CEO duality:

Evidence from the United States of America

Rick Heitkönig

S3466477 – R.J.Heitkonig@student.rug.nl B190215123 – R.Heitonig2@newcastle.ac.uk

Dissertation

MSc Advanced International Business Management & Marketing University of Groningen

Newcastle University Business School

Supervisors Dr J. Shin Dr D. Gregory-Smith

(2)

1

Abstract

The purpose of the study was to investigate the relationship between CEO overconfidence and risk-taking in M&As. Previous research on the upper echelons theory showed that characteristics of the CEO influence organisational outcomes such as decision making and risk-taking. Therefore, this study utilised the upper echelons theory to enrich the literature and further establish the relationship between CEO overconfidence and risk-taking in M&As. This study used sources of overconfidence (self-importance, recent organisational performance & frequency of M&As) as proxies to identify CEO overconfidence and tested the hypothesized relationship with risk-taking in M&As. Previous research indicates that board gender diversity and CEO duality most likely affect the relationship. Therefore, their moderating effects were examined. Based on previous research on board gender diversity and CEO duality, it was expected that their effect on the proposed relation were opposites. The results of this study are obtained through a quantitative research design with the usage of secondary data. The sample contained US publicly listed firms as acquiring firms, and all the target firms globally that were acquired to construct an international perspective to this study. In total, the sample of this study consisted of 878 M&As. The results confirmed the hypothesized relationship between CEO overconfidence source “self-importance” and risk-taking in M&As. However, the results on the other hypothesized relationships are inconclusive. This indicates that further research is necessary on this topic to further establish the relationship. Nevertheless, this study and its results provide valuable insights and contribute to the literature.

(3)

2

Acknowledgements

(4)

3

List of tables

Table I: Summary of variables and data sources ……….. 33

Table II: Descriptive statistics ………. 36

Table III: VIF values ……… 38

Table IV: Pearson correlation table ………....39

Table V: Overview of the hypotheses ………..42

Table VI: Logistic regression ………43/44 Table VII: Robustness test sample unrelated M&As ……… 46/47 Table VIII: Robustness test sample related M&As ………... 48/49

List of figures

Figure I: Mergers & Acquisitions Worldwide ………... 6

(6)

5

1. Introduction

In 2000, America Online acquired Time Warner for $165 billion, which was one of the most significant media mergers in history. The corporation operated after the merger as AOL Time Warner and great potential was expected. However, in 2002, AOL Time Warner reported a quarterly loss of $54 billion in shareholder value as investors pulled out (History, 2009; The New York Times; 2018). This is one of many examples and questions that have gained attention over the last decade are; why do mergers and acquisitions fail? And what is the role of the Chief Executive Officer regarding risk-taking in mergers and acquisitions? (Friedman, Carmeli, Tishler and Shimizu, 2016; Malmendier and Tate, 2005;2008)

(7)

6

Hambrick (2005) propose that “hubris” as well as “overconfidence” belong to the same overarching construct. Since they are effectively the same, this study will mainly use overconfidence to describe this phenomenon.

Figure I: Mergers & Acquisitions Worldwide

Source: IMAA Analysis, 2019

(8)

7

remain a popular strategy for strategic expansion, it is essential and interesting to examine whether CEO overconfidence is an explanation for risk-taking in M&As. This needs to be studied further to protect shareholders, organisations and CEOs in the future. Within business academia, attention has also been paid to diversity at the board of directors (BOD) level. Research has demonstrated how a greater proportion of gender diversity at the senior level can improve decision making (Ray, 2005; Lenard, Work and Yu, 2014). Furthermore, women have a risk perception which increases risk avoidant behaviour (Loukil and Yousfi, 2016). Although the link between improved business performance and diversity has been widely studied, the effect of gender diversity on risk-taking is unclear and requires further research (Loukil and Yousfi, 2016). Researchers refer to CEO duality when the CEO is also acting as chairman of the BOD (Boyd, 1995; Krause, Semadeni and Cannella, 2014), it is hypothesised that this enhances the CEO’s power. Therefore, it easier to pursue risky decisions (Crossland and Hambrick, 2007;Tang, 2017). Thus, it could prove to be an anomaly in this already established correlation, indeed, it may even have the opposite effect.

(9)

8

argues that CEO psychological and demographical characteristics to firm outcomes and performance (Hambrick and Mason, 1984; see also, Johnson and Rudy, 2019; Wang, Holmes, Oh and Zhu, 2015; Yim, 2013). Their perception of the complex situation provides the basis for engaging in strategic decisions (Arena et al., 2018). In addition, the behavioural decisions theory argues that it is likely that an overconfident CEO overestimates their own problem-solving capabilities (Kahneman and Lovallo, 1993), underestimates the resource intensiveness (Li and Tang, 2010), and undervalues the challenges and uncertainties involved when performing cross-border M&As (Shimizu, Hitt, Vaidyanath, and Pisano, 2004). This study integrates these two theories to explain the connection between CEO overconfidence and risk-taking within M&A’s. Therefore, the following research question is formulated.

“To what extent does CEO overconfidence effect M&As and how is this moderated by board gender diversity and CEO duality?”

(10)

9

2. Literature review

This section explains the theoretical concepts that form the bases of this study. It is necessary in order to explain the effect of CEO overconfidence on risk-taking in M&A and the moderating role of board gender diversity and CEO duality. First, the literature on risk-taking in M&As is reviewed. Second, this work examines studies on CEO overconfidence. Finally, this thesis will investigate the moderating variables on board gender diversity and CEO duality. This chapter will conclude by establishing a conceptual framework for this study.

2.1 Risk-taking in M&As

(11)

10

(12)

11

changing the organization (Chin, Hambrick, and Treviño, 2013; Wang et al., 2015). So, both types of characteristics, psychological as well as observable, influence strategic decision-making (Hambrick, 2007; Yim, 2013). As the values and cognitions of the managers are not directly observable, upper echelons theory scholars argue that observable managerial characteristics can be used as proxies (Arena et al., 2018).

(13)

12

characteristics of the executive. Research has shown high levels of support for the theory that was introduced in 1984 (Bromily and Rau, 2018).

(14)

13

desist (Roll, 1986; see also, Berkovitch and Narayanan, 1993; Seth et al., 2000). Seth et al., (2000) find in their research that the agency motive is a more likely an explanation for these irrational choices then hubris. However, their findings are not explicitly based on examining the individuals at the executive level. As CEO overconfidence is a personality trait, and Seth et al., (2000) use the decline in shareholder wealth as an indicator, it is making their studies conclusions questionable. Building on the literature, all three factors have an underlying relation to the other as they ultimately lead to the belief that the acquiring firm can do better.

The motives for completing an M&A are clear, and risk-taking is, to some degree, essential to the development and survival of the organisation (Li and Tang, 2010). CEOs engaging in M&As are certainly exposed to a certain degree of risk, as an M&A is an economic event involving resources (Anand, Capron and Mitchell, 2005). Consequently, due to resources intensiveness of M&As, the risk for acquiring shareholders and the firm increases (Anand, Capron and Mitchell, 2005). The executives pursuing these expansions expect positive returns, but more commonly the effect of M&A’s is neutral, or even negative, on shareholder wealth (Hayward and Hambrick, 1997; Malmendier and Tate, 2008). Therefore, M&As often result in value-destruction (Malmendier and Tate, 2008; Yim 2013).

(15)

14

(16)

15

asymmetry. Finally, due to the increase in challenges, uncertainty and information asymmetry, it is more challenging to learn, and shift knowledge between the home and local market (Baik et al., 2015; Boeh, 2011; Shimizu et al., 2004), and therefore it is harder to accomplish synergistic desires. Therefore, these unique risks which enhance “liability of foreignness and “double-layered acculturation’’ serve as barriers in M&As (Shimizu et al., 2004). On the reasons mentioned above, it is assumed that with cross-border M&As, more risk is involved compared to domestic M&As. Therefore, this study defines risk-taking as the differentiation between choosing for cross-border M&As (more risk) over domestic M&A (less risk).

2.2 CEO Overconfidence / hubris

(17)

16

internal “illusion of control”, implicating that overconfident managers believe that actions and outcomes are more determined by factors within their control rather than outside their control (Hiller and Hambrick, 2005; Hribar and Yang, 2016).

(18)

17

and Tate (2008) elaborated further on their research in 2005, and they confirmed that overconfident CEO’s, overestimate their ability to generate returns. The CEO’s overestimations result in overpaying a target firm by the company looking to acquire them. Overconfidence CEOs expect positive outcomes, even in environments with high uncertainty (Hribar and Yang, 2016). Overall, this leads to “value-destructing M&As”. Even when the CEO is non-overconfident, it is also possible that due to risk-taking incentives, he/she becomes more confident. As a result, he/she is more likely to invest in risky acquisitions (Croci and Petzemas, 2015). Brown and Sarma (2007) have further examined ‘value destruction’ within acquisitions of the United States and argue that overconfident managers are more likely to destroy value compared to other managers. Overall, it is reasonable to believe that overconfident managers will misinterpret the situation regarding M&As. It is likely that their own abilities are overestimated and that they are willing to take the extra risk that is involved in performing cross-border M&As, which enhances the chances of value destruction.

2.3 Sources of CEO overconfidence

CEO overconfidence is not directly observable, as it is a personality trait (Hayward and Hambrick, 1997; Sadler-Smith, 2015). This study used sources as proxies to identify possible CEO overconfidence. Following sub-sections describe the sources that are used to identify CEO overconfidence.

2.3.1 CEO self-importance

(19)

18

CEOs self-importance may be revealed by the centralisation of structural powers in the organisation, or the aggregation in different titles. But more likely, the CEO relative compensation (Hayward and Hambrick 1997). Lee, Cho, Arthurs and Lee (2019) find that overpaid CEOs choose to increase their efforts to compensate for the overpayment. Overpaid CEOs focus their efforts on lowering the premiums that are paid for the acquisitions. Therefore, Lee et al., (2019) show empirically that positive pay deviation influences a CEOs’ decision and behaviour. Van Essen, Otten and Carberry (2015) find that higher levels of compensation is positively related to the power of the CEO, and thus, it influences the power of the CEO when decisions are made. Croci and Petzemas (2015) argue that non-overconfident can become more confident due to risk-taking incentives related to M&As. Thus, the pay of the CEO influences decision-making and risk-taking. CEO self-importance is seen as a stable personality trait by scholars and measured by relative compensation to second-best paid officer (Hiller and Hambrick, 2005; Hayward and Hambrick 1997). It is reasonable to believe that CEO self-importance influences CEO overconfidence. Therefore, this study expects that the higher his/her self-importance is, the more likely it is that the CEO is overconfident. Consequently, it is expected that CEO self-importance is positively related to risk-taking in M&As. Therefore, the first hypothesis is formulated as:

“Hypothesis 1: CEO self-importance encourages the CEO to complete cross-border M&As rather than completing domestic M&As.”

2.3.2 Recent organisational performance

(20)

19

basis for future decisions and behaviour (Kelly, 1971). When recent organisational performance and the attributions a CEO receives are positive, it boosts the CEO’s self-esteem (Tourish, 2019). Adding to that, Lane, Lane and Kyprianou (2004) find that judgements about recent performance directly influences motivation for future efforts. Conversely, when the organisational performance is negative, the poor performance is attributed to the CEO and the top management team, resulting in a decrease in their power and even possible dismissals (Boeker, 1992). A danger to recent organisational success is that past achievement and recognition, motivate CEOs to gain power and stay in current positions (Park et al., 2015). When the organisation is more powerful, it is likely to develop specific patterns of belief and justification of the actions of the CEO. This can make a CEO more self-serving than self-critical (Hayward and Hambrick, 1997), as the CEO is not critical anymore, the CEO tends to narrow his/her focus to own perspective and attitudes. Therefore, limiting their view and only see the perspectives, attitudes and styles that relate to personal aspects (Kanadlı et al., 2018). When the CEO had recent successes with their company, he/she will likely receive positive attributions by others. In turn, the attributions influence and stimulate overconfidence, and therefore, their expectations of their own abilities are influenced (Tourish, 2019). Thus, when recent organisational performances are high, it is expected that the CEO is overconfident about their own abilities. Based on these arguments, it is predicted that success in recent organisational performance will foster CEO overconfidence. Therefore, the following hypothesis is formulated:

(21)

20

2.3.3 Frequency of M&As

(22)

21

identify CEO overconfidence. This study elaborates further upon previous research and uses “more than five M&As in five years” as a proxy (frequency of M&As) to identify overconfidence. Accordingly, the third hypothesis of this study is formulated as:

“Hypothesis 3: CEOs who completed more than 5 M&As in 5 years are encouraged to complete cross-border M&A’s rather than completing domestic M&As.”

2.4 Board gender diversity

(23)

22

literature shows that a diverse team brings new perspectives. In turn, new perspectives lead to an enriched amount of group knowledge (Kanadlı et al., 2018). Nevertheless, it could also hamper the effectiveness of the boardroom by increasing indecisiveness (Gul et al., 2011). Kravitz (2003) suggests that gender diversity in teams is beneficial when the task itself is creative and complex, and therefore many insights are essential. However, when the task is simple, structured and does not request many insights, it could hinder the team due to the multiple points of view. Decisions that are made by the BOD are typically unstructured and request multiple insights, in order to assess the consequences and make higher quality decisions, diversity in the BOD would likely, as a result, help (Gul et al., 2011). Adding to previous statements, Ray (2005) argues that gender diversity is associated with better managerial decision-making, as a wide range of perspectives are present, and consequently, it helps to accelerate complex decision making which requires creativity and judgement (Gul et al., 2011; Ray, 2005). Accordingly, it is more likely that they examine each other’s’ viewpoints more thoroughly, and thus, better vigilance is realised. Also, according to Loukil and Yousfi (2016), women do have a higher risk perception, which results in risk avoidant behaviour. Consequently, it would be assumed that a diverse board would exhibit less overconfidence and are less likely to take either or both risky and extreme positions (Ray, 2005). Which was confirmed by Lenard et al., (2014) as in their empirical research is found that by having a gender diverse board it takes more time to reach consensus, and less risky decisions are made. However, as the literature review demonstrates, what is missing is the effect risk-taking has (Loukil and Yousfi, 2016). Nevertheless, after considering the literature, it is likely that gender diversity reduces CEO overconfidence, and subsequently risk-taking in M&A’s.

(24)

23

2.5 CEO Duality

(25)

24

Li and Tang (2010) argue that when the positions of CEO and chair of the BOD are consolidated, overconfidence is stimulated. When the CEO is confident, he/she is likely to stay in the occupied prestigious position (Park et al., 2015). Ultimately, allowing for CEO overconfidence to influence decisions, and as discussed previously, the company to engage in more risky behaviour. In contrast, when the roles of CEO and chair of the BOD are separated, the BOD is described as an independent structure (Boyd, 1995). This independent structure is perhaps beneficial as it has an opposing effect which enhances monitoring, and thus affecting the decision-making process (Crossland and Hambrick, 2007; Park et al., 2015). Therefore, it is expected that CEO duality increases power and decreases board vigilance, and thus, CEO duality strengthens the amount of risk-taking in M&As. Therefore, the following hypothesis is formulated:

“Hypothesis 5: CEO duality strengthens the relationship between CEO sources of overconfidence and risk-taking in M&As.”

2.6 Theoretical framework

(26)

25

5 years. Together form these sources, the independent variable CEO overconfidence. Based on the literature review, it was expected that CEO overconfidence influences risk-taking in M&As. Risk-taking in M&As is defined in this study as the differentiation between domestic (less risk) and cross-border (more risk), which is the dependent variable displayed in the right box. This study expected that two moderators would influence the proposed relationship. The box on the top of the model is the first moderator, board gender diversity. It was expected that board gender diversity would negatively influence the relationship. Which meant that by having a gender diverse board, it would be more likely that domestic M&As would be preferred over cross-border M&As. The box on the bottom of the model is the second moderator, duality. It was expected that by having CEO duality, the CEO would have more power, and thus, the ability to pursue risky decisions. Therefore, it was expected that duality would strengthen the relationship.

(27)

26

3. Methodology

The methodology section discusses the methodology that is used to test the hypotheses from this study. This section details the sample and data collection used in the study. In addition, it lists all the variables included in the rational and the description. Finally, it explains the techniques used in this study to examine the resultant data set.

3.1 Research philosophy

According to the French Philosopher August Comte, the best means of understanding human behaviour are observation and reason. The positivism ideology assumes that knowledge is objective and observable using the instruments of the independent researcher. Positivism reveals the truths with the usage of empirical means. This study is based on the positivism philosophy. Therefore this study is structured, using a relatively large sample and using a quantitative measurement (Hirschman, 1986). This study uses a deductive approach, and upon existing literature are hypotheses developed. Accordingly, the deductive design is used to test the formulated hypotheses (Hirschman, 1986). This study used a mono method quantitative approach using secondary data to test the formulated hypotheses.

3.2 Sample and data collection

(28)

27

December 2017 as the ending point of the period. Then, to capture an international perspective, the whole range of ‘target firms’, regardless of if they were domestic or cross-border were considered. The US is the largest country by number of deals. However, acquiring companies from different regions of the world might act differently. Therefore, by picking the US as a base country, this study has its limitations.

The sample was narrowed down as some requirements had to be met to make the analysis feasible and obtain reliable results. The first requirement was that the CEO had been acting as CEO in the entire period between 2013 and 2017. All the temporary CEOs who joined at a later stage or left early were excluded as these would negatively influence the results. Secondly, after extensive research in multiple databases, and performing cross-searches, it occurred that data was missing. Accordingly, the missing data was gathered manually when possible. However, if the data was not available, then the M&As were excluded from this study.

(29)

28

became apparent that not all information needed was retrievable through the various databases. For this reason, some information was searched for manually.

The dataset collected for this study is a composition of all of the data collected from the previously mentioned data sources. Additionally, the International Security Identification Number (ISIN) and Committee on Uniform Security Identification Procedures (CUSIP) of the companies were used to ensure that the collected data belonged to the right company. Furthermore, different types of functions and equations were utilized in Microsoft Excel to compose the dataset. The data collection started with a total of 1058 M&As, after the previously described division, the number was cut down to the final sample of 878 M&As. In total were the 878 observed M&As completed by 364 unique CEO’s, of these 878 M&As, a total of 261 of these were cross border.

3.3 Dependent variable

(30)

29

3.4 Independent variable

The independent variable in this study is defined as CEO overconfidence. Top executives of public companies are cautious about answering survey questions partially about personality traits, as that is viewed as sensitive information (Chatterjee and Hambrick, 2007). Therefore, a direct measure for CEO Overconfidence was complex. However, observable managerial characteristics can be used to identify CEO overconfidence (Arena et al., 2018). In previous research, there are several sources of information used as proxies to identify possible overconfidence (e.g. Hribar and Yang, 2016; McManus, 2018). This study elaborates further on previous research and combines measures to identify overconfidence.

Firstly, the CEO’s self-importance is an indication of overconfidence. Therefore, a calculation was performed to measure the salary comparison between the CEO and highest-paid officer within the company at the year of acquisition (Hayward and Hambrick, 1997; Jiang, Stone, Sun and Zhang, 2011; McManus, 2018). The second source used as a proxy for measuring CEO overconfidence was recent organisational performance (Hayward & Hambrick, 1997). Recent organisational performance was measured in this study by dividing the pre-deal net profits by the pre-deal total assets, which results in the return on assets. Lastly, based upon the research of Doukas and Petzemas (2007) frequency of M&As (undertaken more than five M&As in 5 years) proxy was established. Accordingly, a dummy variable was created, splitting the CEOs in more than or less than five M&As in the period between 2013-2017.

(31)

30

M&As). However, after performing a Cronbach’s alpha analysis and factor analysis, it was concluded that this was not possible.

3.5 Moderating variables

The moderating variables, as described the previous chapters, were “board gender diversity” and “CEO duality”. It was expected that their effects would be opposite to each other.

Board Gender Diversity. The exact number of males compared to females being active in the board at the end of the year measured board gender diversity. After that, the total amount of male board members was divided by the total amount of board members what created the board gender diversity ratios.

CEO duality. CEO duality was measured by collecting data on determining who was acting as chair of the BOD. Consequently, the name of the director acting as the chair was compared to the name of the CEO. If the CEO was occupying both positions, at the time of completing the M&A, it was marked as duality.

3.6 Control variables

In this study were control variables included to diminish alternative explanations. The control variables used were on individual-level, board-level, firm-level and deal-level. These variables were controlled for as these might explain risk-taking in M&As. In total were seven control variables included; CEO age, CEO gender, CEO tenure, board size, firm size, firm age and deal size.

(32)

31

al., 2015; Yim; 2013). Therefore, age is the first variable on the individual CEO level that was controlled. Moreover, men and women differ significantly in the way they make decisions and pursue goals. According to Jianakoplos and Bernasek (1998), men are more willing to take risk compared to women. Adding to that, Loukil and Yousfi (2016) find that the risk perception of women leads to risk avoidant behaviour. Therefore, the second variable that was controlled is gender. The third and final control variable regarding individual CEO’s was tenure. Tenure has been previously widely used by scholars as a control variable. When a CEO’s tenure increases, experience also increases. As a result, the CEO possess more knowledge and experience of the market and company. It is assumed that with more considerable experience would come better judgement of the market, and therefore, the risk would decrease because of this. Thus, the third variable that was controlled for was the number of years acting as CEO of the acquiring company.

(33)

32

Control variables were also considered for the company, not just at board-level. Variables were controlled regarding the number of employees and the age of the business. Scholars widely use firm size as a control variable. Studies will often take the size into account by using the number of employees as a control variable. Size influences the risk a company is facing. Larger companies will commonly have more funding available which often results in less vulnerability (Anand, Capron and Mitchell, 2005). In his research, Evans (1987) has demonstrated that older firms are more likely to survive in competing markets. It can be assumed that as experience and knowledge increased, so does the competency to handle uncertainties. The greater competency, in turn, lowers the amount of risk. Regarding the age of a company, however, work has demonstrated that young firms can take advantage of the absence of constraining routines (Bruneel, Yli-Renko and Clarysse, 2010). Therefore, young firms find it easier to adapt to the dynamic business environment (Autio, Sapienza and Almeida, 2000). The above reasoning is why age has been added as a control variable. It has been measured by the number of years since the firm has been reported as being incorporation.

(34)

33

Table I: Summary of variables and data sources

Variables Measured By Source Of Data

Main Variables CEO Overconfidence

Risk-Taking in M&As

Proxies; CEO self-importance, recent organisational

performance, Frequency of M&As

Total number of cross-border M&As compared to domestic

Orbis/BoardEx/Execucomp Zephyr

Moderating Variables Board Gender Diversity CEO Duality

Number of women compared to the number of men

Dual-position; CEO & Chair BOD Execucomp/Boardex Orbis/BoardEx/Execucomp Control Variables CEO Age CEO Gender CEO Tenure Board Size Firm Size Firm Age Deal Size

Age as an absolute number Male or female

Years acting as CEO of the firm Total number of board members Total number of employees Number of years incorporation Value of the deal

Execucomp/Boardex Execucomp/Boardex Execucomp/Boardex Execucomp/Boardex Orbis Execucomp/Boardex Zephyr 3.7 Statistical technique

Various steps have been conducted to examine the data from this study. The use of the statistical program STATA has permitted this study to analyse the data and retrieve results. Firstly, were the descriptive statistics and the correlations analysed to get an understanding of the dataset. This study contains a binary dependent variable. Accordingly, after researching the possibilities for regression, it was decided that a logistic regression was the most suitable to test the relationship. Before testing the relationship, assumptions had to be checked. These assumptions included: the types of measurement scales of dependent and independent variable, the distribution of the data (Appendix C) and multicollinearity.

(35)

34

if there were values above the threshold of 0.700 or below -0.700. In addition, the variance inflation factors (VIF) were checked if there were values present exceeding the threshold of 10. Values exceeding one of the thresholds indicate possible multicollinearity.

(36)

35

4. Findings

This section presents the results of the analysis. It first presents the descriptive statics. Then this work analyses the correlation matrix, and this is followed by an analysis of variance inflation factors to measure multicollinearity. Finally, it will examine the logistic regression with analysis and hypotheses testing.

4.1 Descriptive statistics

(37)

36

size was approximately 3 members with a minimum of 3 members and a maximum of 17 members. The size of the firm was measured in thousands, and the average was 9.318902. The mean of the year of incorporation of the firms was 38.52, with the oldest firm being 132 years old. Finally, analysis of the size of the deal. The sizes among the deals differ significantly, with the smallest deal being 0.1043 and the largest being 94748.67 measured in millions.

Table II: Descriptive statistics

Variable OBS Mean Std. Dev. Min Max. Dependent variable Domestic / Cross-border 878 .2972665 .4573154 0 1 Independent variable Self-importance 878 1.524629 .672914 0 10.59 Recent performance 878 .0580193 .0820431 -.9894396 .2959853 Frequency of M&As 878 .2403281 0.4275206 0 1 Moderating variables 878

BOD gender diversity 878 .8447107 .1391809 0 3

Duality 878 .5410023 .4986 0 1 Control Variables CEO age 878 55.76879 6.899182 29 79 CEO gender 878 .964692 .1846611 0 1 CEO tenure 878 9.374854 6.992582 0 45 Board size 878 9.38041 2.105795 3 17

Firm size (by thousands) 878 43.7720 99.7996 .031 647.5

Firm age 878 38.5262 28.5297 5 132

(38)

37

4.2 Correlations and multicollinearity

In table III is the correlation matrix displayed. It shows the correlations between all the variables in the dataset. CEO self-importance, recent organisational performance and frequency of M&As are the sources for CEO overconfidence and are significantly correlating with each other. To determine whether it was an option to add these sources up and generate one “overconfidence factor,” a Cronbach’s Alpha reliability test was performed. Self-importance and recent performance were positively correlating, but frequency was negatively correlating. This resulted in a reversed. Therefore, the frequency of M&As was excluded before performing another Cronbach’s Alpha reliability test. Unfortunately, the test results (α=0.2013) show that the threshold of 0.7 was not exceeded. Therefore, this study cannot add the sources of overconfidence up and create one variable “overconfidence factor”. Accordingly, this study focuses its understanding upon the effect the sources of overconfidence have on risk-taking in M&A’s.

(39)

38

Table III: VIF values

Variable VIF 1/VIF

Self-importance 1.10 0.9100

Recent performance 1.05 0.9505

Frequency of M&As 1.13 0.8841

BOD Gender Diversity 1.18 0.8502

Duality 1.18 0.8486

CEO Age 1.22 0.8216

CEO Gender 1.07 0.9344

CEO tenure 1.31 0.7647

Board size 1.31 0.7649

Firm size employees 1.27 0.7895

Firm age 1.21 0.8284

(40)

39

Table IV: Pearson correlation table

Note: * p <.10; ** p<.05; *** p<0.01 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 1. Domestic/ Cross-border 1.000 2. Self-importance 0.0623* (0.0649) 1.000 3. Recent performance 0.0579* (0.0866) 0.1119*** (0.0009) 1.000 4.Frequency of M&As 0.0366 (0.2786) -0.1472*** (0.0000) -0.008 (0.9822) 1.000 5. BOD Gender Div. -0.0663**

(0.0495) 0.0143 (0.6718) -0.0470 (0.1637) -0.0107 (0.7526) 1.000 6. Duality -0.0160 (0.6367) 0.0126 (0.7103) -0.0546 (0.1060) 0.1008*** (0.0028) -0.0928*** (0.0059) 1.000 7. CEO Age 0.0348 (0.3027) 0.1320*** (0.0001) 0.0945*** (0.0051) -0.0894*** (0.0080) -0.0788** (0.0195) 0.1710*** (0.0000) 1.000 8. CEO Gender 0.0839** (0.0129) -0.0670** 0.0472 -0.0073 (0.7838) 0.0209 (0.5355) 0.2025*** (0.0000) -0.0400 0.2365 -0.0046 0.8912 1.000 9. CEO Tenure 0.0271 (0.4232) -0.0233 (0.4903) -0.0093 (0.7838) 0.0380 (0.1246) 0.0288 0.3938 0.3141*** (0.0000) 0.3136*** (0.0000) 0.1030*** (0.0022) 1.000 10. Board size 0.0293 (0.3865) -0.0666** (0.0484) 0.0206 (0.5431) -0.0054 (0.8730) -0.2928*** (0.0000) 0.0758** (0.0200) 0.1233*** (0.0002) -0.1208*** (0.0003) -0.1074*** (0.0014) 1.000 11. Firm size employees 0.0924***

(41)

40

4.3 Results logistic regression

The first hypothesis states that when a CEO has a high salary compared to their second rank officer, it will encourage the CEO to take risks in M&As. A logistic regression is performed to test this hypothesis. The data shows that there is a significant positive relationship (β=.261, p=0.033) between self-importance and risk-taking in M&A, see model 2 in Table VI. Also, in model 5 Table in VI, all sources of CEO overconfidence are included. In addition, Model 5 also shows a significant positive relationship (β=.260, p=0.033) between self-importance and risk-taking in M&As. Therefore, hypothesis 1 is supported by the data.

Hypothesis 2 suggested that a recent organizational performance encourages the CEO to take risks in M&As. Model 3 in Table VI tests the effect of recent performance on risk-taking, it shows that there is an insignificant positive relation (β =1.344, p=0.228). Moreover, model 5 in Table VI confirms that the relationship is not significant. Therefore, hypothesis 2 is not supported.

Hypothesis 3 states that when CEOs have completed more than 5 M&As in the last five years, CEOs are more likely to engage in risky M&As. Model 4 in Table VI tests this relationship. However, it shows that the relationship is negative and insignificant (β=-0.0254, p=0.892). Therefore, hypothesis 3 is also not supported.

(42)

41

negative (β= -4,685) but insignificant ( p=0.665). Model 8 in Table VI shows the interaction between frequency of M&As and BOD gender diversity. The effect is positive (β=0.029) but also insignificant (p=0.986). Thus, hypothesis 4 is also unsupported.

Hypothesis 5 states that CEO duality strengthens the relationship between CEO sources of overconfidence and risk-taking in M&As. Model 6 in Table VI shows that the effect of the interaction between self-importance and duality is positive (β=0.142) and insignificant (p=0.636). Model 7 in Table VI shows the interaction effect between recent organisational performance and duality. The effect of the interaction is positive (β=1.782) but insignificant (p=0.459). Lastly, Model 8 in Table VI shows the interaction between frequency of M&As and duality. This effect is also positive (β=0.048) and also insignificant (p=0.899). Therefore, hypothesis 5 is also unsupported.

(43)

42

Table V: Overview of the hypotheses

Hypothesis Result

H1: “Hypothesis 1: CEOs self-importance encourages the CEO to complete cross-border M&As rather than completing domestic M&As.”

Supported

H2: “Hypothesis 2: CEOs recent organisational success encourages the CEO to complete cross-border M&As rather than completing domestic M&As.”

Not supported

H3: “Hypothesis 3: CEOs who completed more than 5 M&A in 5 years are encouraged to complete cross-border M&As rather than completing domestic M&As.”

Not supported

H4: “Hypothesis 4: Board gender diversity negatively influences the relationship between CEO overconfidence and risk-taking in M&As.”

Not supported

H5: “Hypothesis 5: CEO duality strengthens the relationship between CEO sources of overconfidence and risk-taking in M&As.”

(44)

43

Table VI: Logistic regression

(1) (2) (3) (4) VARIABLES Risk-taking in M&As Dom/CB Risk-taking in M&As Dom/CB Risk-taking in M&As Dom/CB Risk-taking in M&As Dom/CB CEO Age 0.0157 0.0132 0.0146 0.0156 (0.0121) (0.0124) (0.0122) (0.0122) CEO Gender 1.784*** 1.884*** 1.793*** 1.783*** (0.639) (0.646) (0.641) (0.639) CEO Tenure -0.0037 -0.0032 -0.0031 -0.0037 (0.0118) (0.0119) (0.0119) (0.0118) Board size 0.0562 0.0635 0.0587 0.0559 (0.0416) (0.0419) (0.0417) (0.0417)

Firm size (thousands) 0.0024*** 0.0028*** 0.0024*** 0.0024***

(0.0007) (0.0008) (0.0007) (0.0008)

Firm age 0.0054* 0.0048* 0.0048* 0.0054*

(0.0028) (0.0028) (0.0028) (0.0028)

Deal size (millions) -0.0005*** -0.0005*** -0.0005*** -0.0005***

(0.0001) (0.0001) (0.0001) (0.0001) Self-importance 0.261** (0.123) Recent performance 1.344 (1.114) Frequency of M&As -0.0254 (0.186) Constant -4.041*** -4.471*** -4.079*** -4.027*** (0.975) (1.009) (0.979) (0.980) Observations 878 878 878 878 Pseudo R-squared 0.0611 0.0659 0.0626 0.0611 Log Likelihood -501.6 -499.1 -500.8 -501.6

(45)

44

Table VI: Logistic Regression (continued)

(5) (6) (7) (8) VARIABLES Risk-taking in M&As Dom/CB Risk-taking in M&As Dom/CB Risk-taking in M&As Dom/CB Risk-taking in M&As Dom/CB CEO Age 0.0115 0.0143 0.0135 0.0149 (0.0125) (0.0127) (0.0122) (0.0122) CEO Gender 1.997*** 1.896*** 1.934*** 1.900*** (0.642) (0.644) (0.641) (0.637) CEO Tenure 0.0006 -0.0024 0.00001 -0.0006 (0.0126) (0.0128) (0.0127) (0.0126) Board size 0.0494 0.0482 0.0398 0.0406 (0.0428) (0.0432) (0.0426) (0.0427)

Firm size (thousands) 0.0024*** 0.0023*** 0.0022*** 0.0022**

(0.0008) (0.0008) (0.0008) (0.0008)

Firm age 0.0038 0.0047* 0.0041 0.0048*

(0.0028) (0.00289) (0.0028) (0.0028) Deal size (millions) -0.0005*** -0.0005*** -0.0005*** -0.0005***

(0.0001) (0.0001) (0.0001) (0.0001) Self-importance 0.260** -1.435 (0.122) (1.049) Recent performance 0.942 4.842 (1.098) (9.736) Frequency of M&As 0.0270 -0.0585 (0.190) (1.444)

BOD Gen. Div. -1.574** -4.251** -1.211 -1.511*

(0.745) (1.935) (0.940) (0.824)

Duality -0.140 -0.368 -0.226 -0.145

(0.170) (0.492) (0.222) (0.188)

Self-importance*BOD Gen. Div. 1.835

(1.228)

Self-importance *Duality 0.142

(0.300)

Recent performance*BOD Gen. Div. -4.685

(10.81)

Recent performance * Duality 1.782

(2.406)

Frequency*BOD Gen. Div. 0.0291

(1.610) Frequency*Duality 0.0484 (0.383) Constant -2.995** -0.494 -2.852** -2.604** (1.213) (1.931) (1.290) (1.212) Observations 878 878 878 878 Pseudo R-squared 0.0719 0.0742 0.0680 0.0660 Log Likelihood -495.9 -494.6 -498 -499

(46)

45

4.4 Robustness tests

(47)

46

Table VII: Robustness test sample unrelated M&As

(1) (2) (3) (4) VARIABLES Risk-taking in M&As Dom/CB Risk-taking in M&As Dom/CB Risk-taking in M&As Dom/CB Risk-taking in M&As Dom/CB CEO Age 0.0053 0.0018 0.0042 0.0038 (0.0183) (0.0187) (0.0183) (0.0184) CEO Gender 0.7470 0.7500 0.6920 0.8160 (0.809) (0.809) (0.813) (0.811) CEO Tenure -0.0003 0.0017 0.0006 -0.0003 (0.0181) (0.0182) (0.0181) (0.0180) Board size 0.0281 0.0430 0.0325 0.0222 (0.0612) (0.0621) (0.0615) (0.0616)

Firm size (thousands) 0.0020** 0.0023** 0.0020** 0.0024**

(0.0009) (0.0009) (0.0009) (0.0009)

Firm age 0.0025 0.0018 0.0019 0.0025

(0.0038) (0.0038) (0.0038) (0.0038)

(0.0001) (0.0001) (0.0001) (0.0001) Self-importance 0.257 (0.182) Recent performance 1.272 (1.607) Frequency of M&As -0.300 (0.267) Constant -1.983 -2.333* -1.979 -1.837 (1.323) (1.355) (1.330) (1.334) Observations 410 410 410 410 Pseudo R-squared 0.0494 0.0540 0.0508 0.0519 Log Likelihood -247.6 -246.4 -247.3 -247

(48)

47

Table VII: Robustness test sample unrelated M&As (continued)

(5) (6) (7) (8) VARIABLES Risk-taking in M&As Dom/CB Risk-taking in M&As Dom/CB Risk-taking in M&As Dom/CB Risk-taking in M&As Dom/CB CEO Age -0.0072 -0.0066 -0.0006 -0.0024 (0.0191) (0.0198) (0.0189) (0.0186) CEO Gender 1.255 1.210 1.222 1.309 (0.837) (0.833) (0.847) (0.841) CEO Tenure -0.0004 -0.0020 -0.0056 -0.0038 (0.0195) (0.0200) (0.0198) (0.0196) Board size 0.0177 0.0168 0.0035 0.0014 (0.0639) (0.0645) (0.0646) (0.0633)

Firm size (thousands) 0.0019* 0.0016 0.0013 0.0016

(0.0010) (0.0010) (0.0009) (0.0010)

Firm age -0.0004 3.1200 -1.0000 0.0006

(0.0040) (0.0040) (0.0040) (0.0039)

(0.0002) (0.0002) (0.0002) (0.0002) Self-importance 0.246 -1.442 (0.182) (1.709) Recent performance 0.778 -8.073 (1.687) (15.78) Frequencyof M&As -0.290 -0.998 (0.276) (2.077)

BOD Gen. Div. -3.269*** -5.850* -3.746** -3.447***

(1.130) (3.290) (1.522) (1.336)

Duality 0.0763 -0.799 0.0394 0.0364

(0.248) (0.761) (0.361) (0.272)

(2.038)

Self-importance*Duality 0.558

(0.467)

Recent performance*BOD Gen. Div. 9.849

(17.47)

Recent performance*Duality 1.316

(4.035)

Frequency*BOD Gen. Div. 0.749

(2.365) Frequency*Duality 0.174 (0.534) Constant 0.831 3.516 1.412 1.266 (1.710) (3.144) (1.871) (1.756) Observations 410 410 410 410 Pseudo R-squared 0.0761 0.0789 0.0705 0.0715 Log Likelihood -240.7 -240 -242.1 -241.9

(49)

48

Table VIII: Robustness test sample related M&As

(1) (2) (3) (4) VARIABLES Risk-taking in M&As Dom/CB Risk-taking in M&As Dom/CB Risk-taking in M&As Dom/CB Risk-taking in M&As Dom/CB CEO Age 0.0213 0.0193 0.0202 0.0220 (0.0167) (0.0171) (0.0167) (0.0167) CEO Gender 3.040*** 3.223*** 3.087*** 3.104*** (1.142) (1.164) (1.146) (1.155) CEO Tenure -0.0023 -0.0031 -0.0022 -0.0029 (0.0160) (0.0161) (0.0160) (0.0161) Board size 0.0725 0.0730 0.0732 0.0739 (0.0592) (0.0594) (0.0590) (0.0592)

Firm size (thousands) 0.0036* 0.0040** 0.0035* 0.0033*

(0.0019) (0.0019) (0.0019) (0.0019)

Firm age 0.0088** 0.0084** 0.0084** 0.0088**

(0.0041) (0.0041) (0.0041) (0.0041)

(0.0001) (0.0001) (0.0001) (0.0001) Self-importance 0.283 (0.176) Recent performance 1.490 (1.586) Frequency of M&As 0.235 (0.263) Constant -5.991*** -6.494*** -6.055*** -6.150*** (1.537) (1.592) (1.541) (1.554) Observations 468 468 468 468 Pseudo R-squared 0.0770 0.0823 0.0788 0.0784 Log Likelihood -250.7 -249.2 -250.2 -250.3

(50)

49

Table VIII: Robustness test sample related M&As (continued)

(6) (7) (8) (9) VARIABLES Risk-taking in M&As Dom/CB Risk-taking in M&As Dom/CB Risk-taking in M&As Dom/CB Risk-taking in M&As Dom/CB CEO Age 0.0201 0.0223 0.0210 0.0229 (0.0174) (0.0173) (0.0170) (0.0170) CEO Gender 3.417*** 3.064*** 3.347*** 3.112*** (1.186) (1.159) (1.174) (1.161) CEO Tenure 0.0071 0.0038 0.0078 0.0066 (0.0173) (0.0173) (0.0172) (0.0171) Board size 0.0810 0.0783 0.0681 0.0756 (0.0615) (0.0616) (0.0616) (0.0620)

Firm size (thousands) 0.0040** 0.0042** 0.0041** 0.0036*

(0.0019) (0.0020) (0.0019) (0.0019)

Firm age 0.0086** 0.0098** 0.0083* 0.0093**

(0.0042) (0.0043) (0.0043) (0.0042)

(0.0001) (0.0001) (0.0001) (0.0001) self-importance 0.337* -1.310 (0.186) (1.543) Recent performance 1.361 24.49 (1.581) (16.73) Frequency of M&As 0.476* 1.522 (0.280) (2.215)

BOD Gen. Div. -0.0423 -2.703 1.804 0.183

(1.013) (2.770) (1.570) (1.078)

Duality -0.507** -0.196 -0.658** -0.437

(0.249) (0.690) (0.313) (0.273)

(1.770)

Self-importance *Duality -0.126

(0.419)

Recent performance*BOD Gen. Div. -27.87

(19.24)

Recent performance * Duality 4.355

(3.374)

Frequency*BOD Gen. Div. -1.142

(2.350)

Frequency*Duality -0.275

BOD Gen. Div. (0.643)

Constant -6.901*** -4.197 -7.733*** -6.310***

(1.945) (2.911) (2.192) (1.930)

Observations 468 468 468 468

Pseudo R-squared 0.0945 0.0902 0.0901 0.0860

Log Likelihood -245.9 -247.1 -247.1 -248.2

(51)

50

5. Discussion

This section discusses the results obtained by performing the logistical regression analyses. These findings are summarised, and then the theoretical and practical implications are considered. Followed by an analysis of the limitations of the study and a look towards future research. This section finishes with the authors concluding thoughts on the study.

The main aim of this study was to determine whether risk-taking in M&As could be explained by CEO overconfidence. As M&As commonly lead to value destruction for the acquiring shareholders and ruined careers, it was interesting to investigate whether CEO overconfidence was a possible cause. This study used three sources as proxies for CEO overconfidence. It was hypothesised that the three sources (self-importance, recent organisational performance, frequency of M&As) were positively related to risk-taking in M&As. In addition, it was predicted that the relationship would be weakened by having a gender diverse board. On the other hand, the predictions were that the relationship would be strengthened by duality.

(52)

51

The results of this study did not entirely establish the proposed relationship. Therefore, a set of robustness tests were performed to test the initial analysis and determine whether general conclusions remained the same with the changed assumptions. The results of the robustness tests are similar to the initial results. However, CEO self-importance is statistically insignificant on all the models when the regression was performed with the unrelated M&As sample, see Table VII. Additionally, when the regression was performed with the related sample, CEO self-importance was insignificant when regressed solely (model 2), but significant when regressed with the other sources (model 6), see table VIII. Therefore, the results of this study need to be interpreted with caution.

5.1 Theoretical implications

The results of this study only confirmed one of the five hypotheses, which means that the relationship between CEO overconfidence and risk-taking in M&As is not entirely confirmed. This research contributes to the existing literature in such a way that it enlarges the current amount of information that is available. It demonstrates that CEO self-importance is influential on risk-taking in M&As. Nonetheless, the null-findings will still provide future researchers with additional attestation for their work. This study will now discuss the findings of this study further. It will offer an explanation and interpretation of the null-findings and examine how the confirmed hypothesis will contribute to the existing literature.

(53)

52

measure CEO overconfidence. A second explanation for the divergent results could be that contextual factors are influencing the proposed relationship. For instance, previous experience could influence the relationship (Hambrick, 2007; Trichterborn and Zu Knyphausen‐Aufseß, 2016; Cuypers et al., 2017). It must be noted that the CEO’s experiences were not evaluated.

Hypothesis 2 predicted that ‘recent organisational performance’ would be positively related to risk-taking in M&As. However, model 3 in Table VI of the logistic regression does not show any significant correlation between the recent organisational performance and risk-taking in M&As. Recent organisational performance can be measured in several ways, and this study has chosen to measure it by ROA%. One possible explanation could be that measuring by ROA% was not the best method for this study, and perhaps other measures would have been more appropriate, for example, stock prices. Another explanation is provided by Tourish (2019) he argues that attributions related to performance influence the expectations of their own abilities. If the received attributions by the CEO were perceived negative recently, it is possible his/her expectations of the ability in performing M&As had been influenced. Adding to that, Lane and Kyprianou (2004) argue that their own judgements about recent performance influence future efforts. Tourish (2019) and Lane and Kyprianou (2004) indicate that CEO behaviour is dependent on the situation. Therefore, it is possible that the CEO’s perception of the situation has influenced the proposed relationship, which possibly explains the divergence in the results.

(54)

53

overconfidence and M&A activity when examining Chinese publicly traded companies. Therefore, it could be possible that frequency of M&As is not an indicator for overconfidence. Another explanation can be found in the study of Friedman et al., (2016). It is possible that the CEOs in this dataset recently had negative experiences performing an M&A. Friedman et al., (2016) argue that it is likely that the CEO would become more defensive after the negative attention and loss of face. Since M&As often fail and lead to shareholder value destruction (Yim, 2013; McManus, 2018), it is a possible explanation that these previous negative experiences influenced the results as these were not taken into account.

(55)

54

Hypothesis 5 was based upon the statements of the agency theory, which suggests that duality leads to lower levels of vigilance of the board. Hypothesis 5 stated that duality would strengthen the relationship between the sources of CEO overconfidence and risk-taking. However, models 6-8 in Table VI in this study do not show a significant positive correlation, and thus, it rejects the hypothesis. A possible explanation for this divergent result can be found in the research of Peng et al., (2007) who finds empirical evidence that duality enhances performance, this is consistent with the Stewardship theory. Therefore, it could be possible that CEO duality does not affect the relationship as M&As are often associated with bad firm performance. Another possible explanation is found in the results of Yan Lam and Kam Lee (2008), their work explains that the advantageous and disadvantageous are very dependent upon the situation, e.g. family-owned or non-family owned firms. This study contained publicly listed firms from the US, and these are predominantly non-family owned firms which is beneficial for CEO duality according to Yan Lam and Kam Lee (2008). Thus, this indicates that it depends on the situation. Despite the fact that CEO duality enhances power, it does not mean that every CEO would take advantage of it and pursue more radical actions. This study did not take the specific situations of the M&As into account, which would possibly explain the insignificant results.

5.2 Practical implications

(56)

55

is likely that the CEO's salary compared to the second-ranked officer is an acceptable proxy to measure CEO overconfidence.

Furthermore, this study aimed to contribute to the literature and knowledge on M&As by explaining whether risk-taking could be explained by CEO overconfidence. At the time of writing, we can only inform shareholders, organisations and CEOs, that CEOs with high levels of self-importance are more likely to perform cross-border M&As. Whether CEO overconfidence is an actual explanation for risk-taking in M&As remains unclear. The limitations and future directions provide further insight into how this question could be answered.

5.3 Limitations and future directions

When interpreting the results, the limitations of this study should be considered. However, it must also be noted that limitations do also provide new opportunities and insights for future research. This dataset consisted only US publicly listed firms when examining the acquiring organisation. The US contains high levels of M&A activity. Therefore, it would be interesting to investigate other national settings as it could be possible that a differing set of results are found when another country is chosen.

(57)

56

now, very little had been written regarding CEO overconfidence and risk-taking in M&As. For this reason, the current literature has not established a perfect method to measure risk. For this reason, it could be possible that measuring it by domestic and cross-border is not the best way to do it. Furthermore, as mentioned previously, there could be other external factors influencing the relationship that could not be evaluated.

(58)

57

research, it might be interesting to take more approaches to measure recent organisational performance and investigate whether there is a difference. The last measure for CEO overconfidence was frequency of M&As. Frequency of M&As also has its limitations as it is based upon previous research (Doukas and Petzemas, 2007; Malmendier and Tate, 2004). Frequency is an indicator for overconfidence, but there are influential external factors, for example, when a company has an aggressive internationalisation strategy. Therefore, these proxies have limitations. Further research should carefully interpret these results and determine if these proxies match are usable for their research.

In addition, this research has limitations regarding the sample. The distribution of the data is skewed, meaning that the total amount of domestic M&As were overrepresented. In total, there were 617 domestic M&As and 261 cross-border, and thus, it is possible that the outcomes would have been different if the data were evenly distributed. Also, as suggested previous, another country other than the USA as the home country could possibly influence results. In future research, it might be interesting to investigate different home countries to determine whether results differ, e.g. eastern countries.

5.4 Conclusion

(59)

58

(60)

59

References

Adams, R. and Ferreira, D. (2009). Women in the boardroom and their impact on governance and performance. Journal of Financial Economics, 94(2), pp.291-309.

Ahern, K.R., Daminelli, D. and Fracassi, C. (2015). Lost in translation? The effect of cultural values on mergers around the world. Journal of Financial Economics, 117(1), pp.165-189.

Aktas, N., de Bodt, E., Bollaert, H. and Roll, R. (2016) “CEO Narcissism and the Takeover Process: From Private Initiation to Deal Completion,” Journal of Financial and Quantitative Analysis. Cambridge University Press, 51(1), pp. 113–137.

Anand, J., Capron, L. and Mitchell, W. (2005). Using acquisitions to access multinational diversity: thinking beyond the domestic versus cross-border M&A comparison. Industrial and Corporate Change, 14(2), pp.191-224.

Arena, C., Michelon, G. and Trojanowski, G. (2018). Big egos can be green: A study of CEO hubris and environmental innovation. British Journal of Management, 29(2), pp.316-336.

Autio, E., Sapienza, H.J. and Almeida, J.G. (2000). Effects of age at entry, knowledge intensity, and imitability on international growth. Academy of management journal, 43(5), pp.909-924.

(61)

60

Berkovitch, E. and Narayanan, M.P. (1993). Motives for takeovers: An empirical investigation. Journal of Financial and Quantitative analysis, 28(3), pp.347-362.

Boeh, K. K. (2011). Contracting costs and information asymmetry reduction in cross‐border M&A. Journal of Management Studies, 48(3), pp.568-590.

Boeker, W. (1992). Power and managerial dismissal: Scapegoating at the top. Administrative Science Quarterly, pp.400-421.

Boyd, B.K. (1995). CEO duality and firm performance: A contingency model. Strategic Management Journal, 16(4), pp.301-312.

Brahma, S., Boateng, A. and Ahmad, S. (2018). Motives of mergers and acquisitions in the European public utilities: An empirical investigation of the wealth-anomaly. International Journal of Public Sector Management, 31(5), pp.599-616.

Bromiley, P. and Rau, D. (2016). Social, behavioral, and cognitive influences on upper echelons during strategy process: A literature review. Journal of Management, 42(1), pp.174-202.

(62)

61

Bruneel, J., Yli‐Renko, H. and Clarysse, B. (2010). Learning from experience and learning from others: how congenital and interorganizational learning substitute for experiential learning in young firm internationalization. Strategic entrepreneurship journal, 4(2), pp.164-182.

Business Insider. (2010). How Does Steve Jobs Manage To Get By On Just A Dollar A Year?. [online] Available at: https://www.businessinsider.com/how-does-steve-jobs-manage-to-get-by-one-just-a-dollar-a-year-2010-1?international=true&r=US&IR=T

Business Insider. (2019). Jeff Bezos got so rich in 2018 that he now makes more per minute than most people do in a year. [online] Available at: https://www.businessinsider.com/how-much-jeff-bezos-makes-per-minute-2018-12?international=true&r=US&IR=T

Campbell, W.K., Goodie, A.S. and Foster, J.D. (2004). Narcissism, confidence, and risk attitude. Journal of behavioral decision making, 17(4), pp.297-311.

Carter, D.A., Simkins, B.J. and Simpson, W.G. (2003). Corporate governance, board diversity, and firm value. Financial review, 38(1), pp.33-53.

(63)

62

Chatterjee, A. and Hambrick, D.C. (2007). It's all about me: Narcissistic chief executive officers and their effects on company strategy and performance. Administrative science quarterly, 52(3), pp.351-386.

Chin, M.K., Hambrick, D.C. and Treviño, L.K. (2013). Political ideologies of CEOs: The influence of executives’ values on corporate social responsibility. Administrative Science Quarterly, 58(2), pp.197-232.

Claxton, G., Owen, D. and Sadler-Smith, E. (2015) ‘Hubris in Leadership: A Peril of Unbridled Intuition’, Leadership 11, pp.57–78

Conyon, M. and He, L. (2017). Firm performance and boardroom gender diversity: A quantile regression approach. Journal of Business Research, 79, pp.198-211.

Croci, E. and Petmezas, D. (2015). Do risk-taking incentives induce CEOs to invest? Evidence from acquisitions. Journal of Corporate Finance, 32, pp.1-23.

Crossland, C. and Hambrick, D.C. (2007). How national systems differ in their constraints on corporate executives: A study of CEO effects in three countries. Strategic Management Journal, 28(8), pp.767-789.

(64)

63

Deloitte United States. (2019). Cross-border M&A risks and rewards | Deloitte US. [online] Available at: https://www2.deloitte.com/us/en/pages/mergers-and-acquisitions/articles/cross-border-m-and-a-risks-rewards.html

Doukas, J. A. and Petmezas, D. (2007). Acquisitions, overconfident managers and self‐ attribution bias. European Financial Management, 13(3), pp.531-577.

Echajari, L. and Thomas, C. (2015). Learning from complex and heterogeneous experiences: the role of knowledge codification. Journal of Knowledge Management, 19(5), pp.968-986.

Erel, I., Liao, R.C. and Weisbach, M.S. (2012). Determinants of cross‐border mergers and acquisitions. The Journal of finance, 67(3), pp.1045-1082.

Evans, D. (1987). The Relationship Between Firm Growth, Size, and Age: Estimates for 100 Manufacturing Industries. The Journal of Industrial Economics, 35(4), pp.567-581.

Finkelstein, S., Cannella, S. F. B., Hambrick, D. C. and Cannella, A. A. (2009). Strategic leadership: Theory and research on executives, top management teams, and boards. Oxford University Press, USA.

(65)

64

Finkelstein, S. and D'aveni, R.A. (1994). CEO duality as a double-edged sword: How boards of directors balance entrenchment avoidance and unity of command. Academy of Management journal, 37(5), pp.1079-1108.

Goranova, M.L., Priem, R.L., Ndofor, H.A. and Trahms, C.A. (2017). Is there a “Dark Side” to Monitoring? Board and Shareholder Monitoring Effects on M&A Performance Extremeness. Strategic Management Journal, 38(11), pp.2285-2297.

Goodstein, J., Gautam, K., & Boeker, W. (1994). The effects of board size and diversity on strategic change. Strategic management journal, 15(3), pp.241-250.

Gul, F. A., Srinidhi, B., & Ng, A. C. (2011). Does board gender diversity improve the informativeness of stock prices?. Journal of Accounting and Economics, 51(3), pp.314-338.

Hambrick, D. (2007). Upper Echelons Theory: An Update. Academy of Management Review, 32(2), pp.334-343.

Hambrick, D. and Mason, P. (1984). Upper Echelons: The Organization as a Reflection of Its Top Managers. The Academy of Management Review, 9(2), pp.193.

Hayward, M. L. A., & Hambrick, D. C. (1997). Explaining the premiums paid for large acquisitions: Evidence of CEO hubris. Administrative Science Quarterly, 42(1), pp.103-127.

(66)

65

Hilary, G. and Menzly, L. (2006). Does Past Success Lead Analysts to Become Overconfident?. Management Science, 52(4), pp.489-500.

Hill, C.W. and Jones, T.M. (1992). Stakeholder‐agency theory. Journal of management studies, 29(2), pp.131-154.

Hiller, N. and Hambrick, D. (2005). Conceptualizing executive hubris: the role of (hyper-)core self-evaluations in strategic decision-making. Strategic Management Journal, 26(4), pp.297-319.

Hirschman, E.C. (1986). Humanistic inquiry in marketing research: philosophy, method, and criteria. Journal of marketing Research, 23(3), pp.237-249.

HISTORY. (2019). AOL-Time Warner formed. [online] Available at: https://www.history.com/this-day-in-history/aol-time-warner-formed

Hribar, P. and Yang, H. (2016.) CEO overconfidence and management forecasting. Contemporary Accounting Research, 33(1), pp.204-227.

(67)

66

Jianakoplos, N.A. and Bernasek, A. (1998). Are women more risk averse?. Economic inquiry, 36(4), pp.620-630.

Jiang, F., Stone, G. R., Sun, J. and Zhang, M. (2011). Managerial hubris, firm expansion and firm performance: Evidence from China. The Social Science Journal, 48(3), pp.489-499.

Kahneman, D. and Lovallo, D. (1993). Timid choices and bold forecasts: A cognitive perspective on risk taking. Management science, 39(1), pp.17-31.

Johan, A.P. and Handika, R.F. (2017). CEOS characteristics and the successful of turnaround strategy: Evidences from Indonesia. Academy of Strategic Management Journal, 16(1).

Kanadlı, S.B., Bankewitz, M. and Zhang, P. (2018). Job-related diversity: the comprehensiveness and speed of board decision-making processes—an upper echelons approach. Journal of Management and Governance, 22(2), pp.427-456.

King, D. R., Dalton, D. R., Daily, C. M., & Covin, J. G. (2004). Meta-analysis of post-acquisition performance: Indications of unidentified moderators. Strategic Management Journal, 25(2), pp.187–200.

Kelley, H.H. (1973). The processes of causal attribution. American psychologist, 28(2), p.107.

To what extent does CEO overconfidence effect M&As and how is this moderated by board gender diversity and CEO duality: Evidence from the United States of America