What would Lehman Sisters do? The effect of board gender diversity
on M&A decision-making and performance
Lorian Micu S2036835 University of Groningen Faculty of Economics and Business
Abstract
Using different M&A transaction characteristics such as number and value of yearly transactions, premium paid, percentage of stock acquired, industry-related, cross-border and payments in stock and cash as dependent variables, we find that the presence of female directors, their number and percentage on boards of directors influence the firm to make decisions which can be attributed to the risk-aversion behavioural trait of women suggested in previous literature. However, there are also unexpected findings that suggest female directors are more likely to steer the acquirer firm to pay in cash, an option substantially riskier than the payment in stock and to acquire a larger stake in the target. Moreover, contradictory to previous findings, we find evidence indicating a negative relationship between gender diversity and the market reaction to M&A announcements.
Keywords: mergers, corporate governance, director gender, acquisitions, firm acquisitiveness,
industry related, cross-border, bid premium, stake size acquired, payment method, cumulated abnormal returns.
JEL Classification: G30, G34, J16, M14
Study Programme: MSc International Financial Management Other program: MSc Finance
Acknowledgements
Firstly, I thank my supervisors Prof. Dr. Melsa Ararat and Dr. Halit Gonenc, who have shown great support and encouragement while writing my thesis. Your challenging questions and suggestions were of tremendous help during the process of writing this thesis.
Secondly, I dedicate this humble work to the women who have enabled me to complete it. My grandmother who has put the basis of my education, my aunt who has made it possible for me to move to the Netherlands and my mother, who’s memory fuels my will to keep improving.
1
1. Introduction
Mergers and acquisitions (M&A) are usually the company’s largest form of discretionary spending and it often has a greater effect on shareholder value than any other event (DePamphilis, 2011; Gaughan, 2015). Committing the company to a certain strategy for a long period of time, this growth strategy becomes increasingly more difficult to abandon as implementation begins (Mellen & Evans, 2010; Schmidt, 2015). To understand the magnitude of M&A transactions, between 2002 and 2010 the worldwide value of announced M&As were, on a yearly basis, between 2.9-4.9 trillion USD, representing about 3% of the worldwide GDP or more than the GDP of Germany (Institute for Mergers Acquisitions and Alliances, 2015).
An investment decision on its own, going into a merger or acquisition, has the purpose of increasing shareholder wealth, however, about 60% of M&A deals are value-destroying (Bruner, 2002). According to Pereiro (2016) one of the reasons for the failure of the majority of M&A transactions is misvaluation by managers, who believe substantial synergies exist in a deal, bid at excessive prices and pay for targets more than they are actually worth. The decision to engage in an M&A transaction is proposed by the management however it requires the approval of the board of directors (Hegendorff, Collins, & Keasey, 2010; Gaughan, 2015).
The board of directors is one of the most important internal governance mechanisms that monitor and advise management in order to protect shareholder interest (Francis, Hasan & Wu, 2012). The board is the ultimate legal authority with respect to decision making in a firm (Adams & Ferreira, 2007) and has the power to hire, fire and compensate top-level managers and to ratify and monitor important decisions such as M&A transactions (Fama & Jensen, 1983). By representing shareholders’ interests, the board of directors can mitigate the likelihood that managers endowed with free cash flow will invest in negative present value projects (Jensen, 1986; Haleblian, et al, 2009), such as value destroying acquisitions, or that the CEO will engage in empire building (Williamson, 1964) through the same potentially value destroying activity.
2 Our research is driven by the continous debate and the recent legal support (corporate gender quotas across an increasing number of countries) shown for The Lehman Sisters hypothesis and leads us to the following research question: Is the presence of women on boards of directors influencing decision making in M&A transactions?In other words, is there empirical proof to infer that the legal and corporate efforts to increase gender diversity can lead to lower risk-taking, improved corporate decision making and shareholder value creation?
In 2014, according to Terjesen, Aguilera & Lorenz (2015), 10 countries have already implemented quotas for female representation on publicly traded or state owned enterprises (Norway, Spain, Finland, Quebec, Israel, Iceland, Kenya, France, Italy, Germany) and 15 other countries have included gender diversity requirements in their corporate governance codes.
The influence of gender diversity on corporate decision making discovered by previous studies attests to the importance of further exploring this subject. The theoretical arguments supporting gender diversity are, in essence, that female directors, through their cognitive frames, which are different from men’s, impact firm outcomes differently (Hambrick, 2007). Furthermore, because of their different experiences and knowledge, women can expand the human capital present on board (Carter, D'Souza, Simkins, & Simpson, 2010; Hillman, Cannella, & Harris, 2002), can increase team collaboration through their inclination for benevolence, tolerance and interdependence (Adams & Funk, 2012) and can provide improved monitoring of the CEO (Adams & Ferreira, 2009). Moreover, according to Dhir (2015), female directors present more enhanced diligence as they are more likely than their male counterparts to probe deeply into the issue at hand via an assertive presentation of inquiries and more challenging and counterintuitive questions. Dhir (2015) also argues that males are more afraid to show that they do not know everything and most frequently having the same background, expertise and education spend time on items they have a lot of knowledge about, while failing to recognize potential risks.
Considering the findings on gender diversity and firm performance, the debate is far from settled. While some academics report a negative relationship between gender diversity and stock market performance (Haslam, Ryan, Kulich, Trojanowski, & Atkins, 2010), others find that markets react positively to the appointment of female directors (Campbell & Vera, 2009). With respect to accounting returns, Nguyen and Faff (2012) discover a positive relationship, while Carter et al (2010) find no relationship and Ahern and Dittmar (2012) find a negative relationship.
3 analyzed in the financial literature. More specifically, we analyze the effect of the presence, number and proportion of female directors on boards, on the acquisitiveness of the firm, the transaction size, the premium paid and shareholder value creation around the announcement of M&A transactions. Furthermore, we also look at the probability of gender diverse boards engaging in cross-border and intra-industry transactions and at their preferences for the method of payment.
While Huang et al. (2013) focus only on executive directors of US listed firms to examine the influence of gender on corporate financial and investment decisions such as acquisitions and debt issuance, Dowling et al.(2013) and Levi et al. (2014) focus only on the influence of gender diversity on bid initiation and on the premium paid using a sample of listed companies in the UK and US, respectively. Our study is, to the best of our knowledge, the first to analyze, a wide spectrum of M&A decisions using an international sample of firms from 23 countries over a period of 9 years, thus overcoming methodological flaws of previous studies concerning sample sizes that are both geographically limited and covering short periods of time (Aguilar, 2013).
In addition to previous studies on gender diversity in the context of M&A, we find evidence suggesting that the spectrum of decisions in M&A transactions by gender diverse boards cannot be attributed to lower overconfidence and risk aversion alone (Huang et al. 2013; Levi, et al 2014), and that to understand the decision making of gender diverse boards, more in-depth research regarding the motivation behind certain decisions is needed.
Using an unbalanced panel dataset of more than 9000 firm-year observations spanning the period from 2002 to 2010 and employing different methodologies to account for dichotomous count data in dependent variables and potential endogeneity arising from selectivity bias, we find that gender diverse boards are negatively associated with the number of yearly acquisitions, cross-border acquisitions and the probability of paying in stock, and positively associated with the share of stock acquired and with the probability to pay in cash. Altough we find a negative relationship between female directors and shareholder value creation using OLS, the results are marginally significant only for the 5-days event window when using a Heckman two-step model to account for selectivity bias.
4
2. Literature review
The following section discusses theories on gender diversity and findings of studies focusing on this topic and its connection to corporate decision making and financial performance.
According to Adams, Hermalin and Weisbach (2010) boards have the responsibility to nominate, appoint and monitor executive directors and to set and guide firm strategy. In their quest for understanding the relationship between board characteristics and firm performance, academics have looked at corporate boards through the lenses of Resource Based View and Agency Theory.
2.1. Board diversity and Resource Based View
According to Pfeffer and Salancik (1978), who take the resource based perspective, the firm needs tools which allow it to overcome challenges and uncertainties and to take advantage of the opportunities present in the external environment. A diverse board of directors can help the firm respond to changes in the external environment through the access it can provide to certain resources (Johnson, Daily, & Ellstrand, 1996), through its superior understanding of the environment and through its improved quality of advice (Carter, Simkins, & Simpson, 2003). Moreover, Hillman, Cannella and Paetzold (2000) argue that diverse boards have larger networks links to a more diverse set of external stakeholders which can create increased opportunities for generating firm value.
A way of creating more diverse boards is by increasing age, ethnicity, gender and expertise diversity on boards. Supporting this rationale, Stiles and Taylor (2001) argue that board diversity increases the probability of favourable outcomes for the firm in its pursuit of securing the resources it needs. However, according to Hillman, Shropshire and Cannella (2007) legitimacy, or the perception of acting according to acceptable norms, is a vital aspect in securing these resources, without which exchanging resources with the external environment can become problematic. Furthermore, van der Walt and Ingley (2003), argue that diverse boards can create legitimacy through inclusion and equal representation of various societal groups.
5 a wide range of interpretations, alternatives and consequences”. In addition, women can improve conversational turn-taking enabling the group members to be responsive to one another and to make the best use of the knowledge and skills of the members (Bear & Wooley, 2011), in turn, increasing the collective intelligence of the board (Dhir, 2015).
2.2. Board diversity and Agency Theory
According to Adams and Ferreira (2007), the two main functions of the board are advising and monitoring. The monitoring function regards the extent to which the board aims to control managerial opportunism and agency costs and to oversee firm actions through different decisions (Post & Byron, 2015).
Taking the agency perspective, we regard boards of directors as having a crucial role in diminishing agency costs of self-interested managerial behaviour or poor managerial performance (Jensen & Meckling, 1976; Fama & Jensen, 1983). From this viewpoint, the board is considered a mechanism which regulates managerial actions and which ensures the firm is creating shareholder value. The level of independence and monitoring of the board has been found to have influence on managerial self-serving, CEO turnover and M&A transactions (Lu & Wang, 2015). With regards to M&A transactions, women are less prone towards empire building (Levi, Li, & Zhang, 2014) and thus may be less prone to undertake negative net present value projects ensuring shareholders’ interests are respected.
Studies have looked at board gender diversity and its effect on the monitoring function. Adams and Ferreira (2009) argue that female directors can improve monitoring through independent thinking and by not following the “old-boys’ club” when it comes to decision making. Moreover, the presence of female directors on boards, can promote cognitive diversity and constructive conflict mitigating groupthink in the boardroom (Fondas & Sassalos, 2000; Dhir, 2015).
The presence of female directors on boards seem to also improve the overall board meeting attendance, thus also contributing to the monitoring ability of the board indirectly, by influencing their colleagues to show more participation (Adams & Ferreira, 2009). There are, however, views that argue that greater board diversity can lead to less cohesive groups, less cooperation and more conflicts (Ali, Ng, & Kulic, 2014).
6
2.3.Gender diversity and firm performance
The impact of women directors on firm performance is yet to be settled, as so far, gender seems to be improving performance in firms with weak governance and decreases shareholder value in firms with strong governance (Liang, Xu, & Jiraporn, 2013).
The difference that women on boards can make towards the financial performance of firms, stems mainly from the differences in psychological traits between men and women. According to (Eckel & Grossman, 2008), women seem to care less about competition and are more risk averse then men. This is one reason for which gender diverse boards can be more prone to ratify less risky projects and reject the riskier ones. As stated by Luis Aguilar (2013), commissioner at U.S. Securities and Exchange Commission, the presence of women on board has been linked to better organizational performance, higher rates of return and more effective risk management. He further argues these positive results can arise because of women’s more effective approach towards risk through decision making that takes the interests of multiple stakeholders into account.
Other supporting evidence of the gender diversity-firm performance relationship was also found in a study on S&P 1500 firms, by Dezsö and Ross (2012) which showed that the presence of female top managers influences firm performance positively, however, this relationship is established for companies which have an innovation-oriented strategy. Also, the positive relationship between board gender diversity and Tobin’s Q was found significant by Nguyen, Locke and Reddy (2015) for Vietnamese firms, by Carter, et al. (2003) for US firms and by Campbell and Minguez-Vera (2008) for Spanish firms.
There are, however, also studies which show that the presence of women on boards does not affect firm performance, as found by Francis, Hasan and Wu (2012) during the financial crisis of 2007 to 2009. Moreover, the stock market does not seem to recognize the addition of women on boards as was found by Farrell & Hersch (2005) when analysing the abnormal returns of appointment announcements of women on boards in contradiction to the findings of Campbell and Vera (2009).
2.4.Gender diversity and M&A
7 perspectives gender diverse boards are better able to ascertain the risk of acquisitions and reach better decisions (van der Walt et al. 2006). Furthermore, the M&A process could benefit from increased criticism on decision making as board diversity can stimulate the critical analysis of decisions that are less likely to be discussed when directors have the same experience and background (Dhir, 2015).
The potential of gender diversity affecting the M&A process positively is supported by Huang and Kisgen, (2013) who find a positive relationship between gender diversity and acquisition announcement returns. Moreover, the same authors find that the number of acquisitions a firm engages in is negatively affected by board gender diversity.
The findings of previous studies focusing on gender diversity and M&A transactions are motivated by two behavioural traits, risk taking attitude and overconfidence.
According to Sapienza, Zingales and Maestripieri (2009), testosterone, which has been found to be linked to risk-taking, is present at substantial lower levels in women compared to men. Moreover, the difference in risk-taking can be witnessed in the safer approach of women towards sex, drug use, gambling, driving employment choices (Levi, Li, & Zhang, 2014) and trading behaviour (Barber & Odean, 2001). Furthermore, Berger, Kick and Schaek (2014) suggest that director age, gender and educational qualifications influence the level of portfolio risk that banks take. These findings suggest that gender diversity can have a fundamental impact on the level of risk boards are willing to take when engaging in mergers and acquisitions.
There is also evidence pointing to the fact that corporate executives can differ from the general population regarding such behavioural traits as risk-taking. While Adams and Funk, (2012) find that female directors are more risk-seeking that male directors, Graham, Harvey and Puri (2013) find that US CEOs are more positive and risk-tolerant than the lay population. This evidence suggest that the view of women being less risk tolerant, might not hold when it concerns corporate executives.
Regarding overconfidence, Kruger (1999) describes it as an excessive belief in one’s abilities, belief that desired outcomes are under one’s control and believing one’s estimate is more accurate. In the words of Pereiro (2016), overconfidence can lead to excesively high bid prices, while the lack of it can lead to the downplay of bids and bypassing of potentially profitable M&A deals. This view is supported by the findings of Levi et al (2014) which show that between gender diversity and the premium paid there is a negative relationship, finding which they attribute to the lower overconfidence of women in estimating M&A synergies.
According to Huang and Kisgen (2013) female executives place wider bounds on earnings estimates and exercise stock options early supporting the view that women are less likely to be overconfident in comparison to men. These findings also suggest that women can potentially recognize more possible outcomes in the future, not just the positive ones and have less certainty about accurately predicting the final state of the world.
8 and synergies from acquiring other companies (Malmendier & Tate, 2008). If the differences in behavioural traits between genders are expected to lead to different quality of M&A-related decisions, then it can be expected that financial markets are aware of this difference and will react accordingly around the announcement of M&A transactions.
3. Methodology
3.1. Hypotheses development
In the following paragraphs we look at the relationship between gender diversity and M&A decisions by taking the perspective of the Resource Based View and of the Agency Theory. Our goal is to analyse if board gender diversity indeed has an effect on the various decisions surrounding M&A transactions and if there is empirical evidence to support the Lehman Sisters hypothesis. In addition to these theories, we also present previous findings to provide the rationale for our hypotheses with regards to the characteristics of M&A deals that gender diverse boards engage in. Our investigation in the acquisition behaviour of gender diverse boards begins with looking first at the relationship between board gender diversity and acquisitiveness, the acquisition premium paid and deal characteristics (cross-border, industry relatedness, size of stake acquired and payment method) and ends by analysing the market reaction to the acquisitions made by gender diverse boards.
3.1.1. Gender diversity, firm acquisitiveness and M&A transaction characteristics
Previous research by Jensen & Zajac (2004) reports a negative relationship between the number of finance non-executive directors on board and acquisition activity. This suggests that directors’ characteristics can influence the acquisitiveness of the firm. Moreover, country-level evidence supports the idea that director characteristics can influence the firm’s acquisitiveness (Dowling & Aribi, 2013; Levi, Li, & Zhang, 2014).
9 transaction. Moreover, because of increased knowledge and more diverse perspectives, gender diverse teams would have the collective intelligence necessary to consider a larger number of targets from different geographical areas and different industries (Anderson, Reeb, Upadhyay, & Zhao, 2011; Adams, de Haan, Terjesen, & van Ees, 2015; Dhir, 2015).
By taking the perspective of the Agency Theory when considering the acquisition behavior of gender diverse boards, we can infer that gender diversity can influence firm acquisitiveness negatively. Academic findings suggest women directors are more likely to improve board monitoring due their assertive and inquiring nature and to not being part of the “boys’ club” bringing decisions concerning M&A transactions under scrutinity which can result in less yearly acquisition (Dhir, 2015). Moreover, while male board members may, theoretically, misvalue potential targets due to overconfidence resulting in higher expected synergies and higher bid premiums, female directors may influence the decision making in the opposite direction by downplaying the potential synergies a target may have to offer. In addition, concerning also the findings which suggest women are less risk tolerant, gender diversity would influence negatively the board’s engagement in M&A deals concerning targets from different countries or different industries from the acquirer’s, because of the perceived high risk and down-play of expected synergies.
We acknowledge the fact that gender diverse boards can influence the acquisitiveness of the firm positively through access to larger networks and resources and increased collective intelligence. However, based on previous studies, at a personal level, we also acknowledge that the psychological traits of lower overconfidence and risk-taking can influence the decision making of female directors negatively. In conclusion, we expect gender diversity to be negatively associated with firm acquisitiveness, premium paid, industry diversification and cross-border acquisitiveness.
Hence we state our first four hypotheses:
H1. The proportion of female directors on board is negatively associated with firm acquisitiveness H2. The proportion of female directors on board is negatively associated with the premium paid by
the acquirer
H3. There is a negative relationship between the gender diversity present on boards and the
likelihood to engage in M&A transactions outside of their two-digit industry SIC code
H4. There is a negative relationship between the gender diversity present on boards and the
10 3.1.2. Gender diversity, percentage of stock acquired and payment method
We continue our analysis of the influence of women directors on decision making around M&A transactions by analysing the relationship between gender diversity and the percentage of stock acquired and also between gender diversity and the payment method with either cash or stock. We proceed by first mentioning the importance of these decisions after which we present our hypotheses.
The percentage of stock acquired in the target and the method of payment, both have significant implications for the sharing of control and of the risk and reward, between acquirer and target.
Firstly, regarding the sharing of control, the larger the percentage of stock acquired in the target, the more decisional control the acquirer has over the target. Since directors on board are being chosen by shareholders through the voting power of the stock they hold, the acquirer receives the right to have its interests represented on the board through a number of directors, according to the size of the stock acquired and the conditions negotiated upon. Concerning the method of payment, paying in cash leads to a very clear attribution of ownership and control which cannot be said about the payment in stock where acquirer and target both hold a stake in each other (DePamphilis, 2011).
Secondly, looking at the sharing of risk and reward, according to the percentage of acquired stock, the acquirer is exposed to the potential risk and reward of the cash flows expected from the target’s operations. We can see why, in cases of high uncertainty about the cash flows of the target, the acquirer would rather hold a smaller stake in the target. Regarding the method of payment, while stock payments result in both acquirer and target sharing the risk and reward of the expected synergies, the cash payment leaves the potential risk and reward entirely with the acquirer.
11
Stating our fifth and sixth hypotheses:
H5. The proportion of female directors on board is associated with the percentage of stock acquired
in the M&A transaction
H6. The proportion of female directors on board has an influence on the type of payment method used
in M&A transactions
3.1.3. Gender diversity and cumulated abnormal returns around M&A announcement dates
We acknowledge the fact that the sign relationship between board gender diversity and firm performance, or firm value, is not simple to infer using only the theory at hand, reason for which, in addition to using theories, we also discuss current results to build our seventh, and last, hypothesis.
Considering the influence women directors can have on the market reaction, we can infer, from the Resource Based View, that through their augmented monitoring capabilities, collaborative nature and different networks female directors can add value to the firm through improved decision-making and this can be recognized by the market through a positive reaction to the announcement of M&A deals. The same can be said if female directors are indeed less overconfident and downplay the bid price for the target, case in which the acquirer’s shareholders will pay less for an acquisition. However, the market could also react negatively because, considering the positive risk-reward relationship and female directors’ risk-averse attitude, the deals that gender diverse boards engage in are, in fact, not risk-taking enough to create positive expectations regarding future earnings from the M&A deals in question.
There are however findings that support the idea that increased gender diversity on boards can add value and improve performance. A positive relationship between gender diversity and firm value is found by Carter et al. (2003) for a sample of Fortune 1000 firms and by Campbell and Minguez-Vera (2008). Moreover Erhardt, Werbel and Shrader (2003) find that gender diversity influences return on assets (ROA) and return on investment (ROI) positively. Furthermore, Kang, Ding and Charoenwong (2010) and Campbell and Minguez-Vera (2010) find a positive market reaction to the announcement of women directors, in contrast to the findings of Farrell and Hersch (2005).
12
H7. The proportion of female directors on board is positively associated with the cumulated abnormal
returns surrounding the M&A transaction announcement
3.2. Econometric models
In this section we describe the econometrical methods used to test the above mentioned hypotheses. We evaluate the first hypothesis using a negative binomial regression and a zero-inflated Poisson regression, the second and fifth hypothesis using an OLS regression. The third, fourth and sixth hypothesis, we test by using a probit regression and the seventh hypothesis, by employing an event study methodology together with a Heckman two-step model. Furthermore, as additional analysis, we address the potential endogeneity arising in this models by employing an instrumental variable methodology.
3.2.1. Negative binomial regression and zero-inflated Poisson regression for count data
Negative binomial regression is used with count data to account for overdispersion, or the case when the observed variance is larger than the mean. Not accounting for overdispersion leads to underestimation of standard error coefficients which leads to too narrow confidence intervals and too small p-values (Hausman, Hall, & Griliches, 1984, Chatterjee & Simonoff, 2013). Moreover, not accounting for the fact that many firm-year observations equal 0, leads to the homoscedasticity assumption of the linear regression model to be violated (Nadolska & Barkema, 2007). We employ negative binomial regression to test the first hypothesis because we use the yearly number of acquisitions which can range from zero to many, as dependent variable. In our case, as usual when dealing with count data, the variance of our dependent variable exceeds its mean as we observe by doing a conditional mean analysis of the number and value of acquisitions and gender diversity. This, however, is not provided for reasons of brevity.
13 H1: 𝐹𝑖𝑟𝑚 𝑎𝑐𝑞𝑢𝑖𝑠𝑖𝑡𝑖𝑣𝑒𝑛𝑒𝑠𝑠 = 𝛼1+ 𝛽1𝐺𝑒𝑛𝑑𝑒𝑟 𝐷𝑖𝑣𝑒𝑟𝑠𝑖𝑡𝑦 + 𝛽2𝐵𝑜𝑎𝑟𝑑 𝑠𝑖𝑧𝑒 + 𝛽3𝐶𝐸𝑂 𝐷𝑢𝑎𝑙𝑖𝑡𝑦 + 𝛽4𝐼𝑛𝑑𝑒𝑝 𝐷𝑖𝑟𝑒𝑐𝑡𝑜𝑟𝑠 + 𝛽5𝐹𝐶𝐹 + 𝛽6𝑇𝑜𝑏𝑖𝑛𝑠 𝑄 + 𝛽7𝐿𝑒𝑣𝑒𝑟𝑎𝑔𝑒 + 𝛽8𝑅𝑂𝐴 + 𝛽9𝑀𝑎𝑟𝑘𝑒𝑡 𝐵𝑜𝑜𝑘 + 𝛽10𝑀𝑎𝑟𝑘𝑒𝑡 𝐶𝑎𝑝 + 𝜖𝑖𝑡 3.2.3. OLS regression
We employ a standard OLS regression using the deals dataset to test the second and fifth hypothesis. For the second hypothesis we use as dependent variable the premium paid (H2) in acquisitions expressed as a ratio of, the offering price paid by the acquirer, to the target book value per share. For the fifth hypothesis we use the percentage of stock acquired (H5) in the M&A transaction as the dependent variable. In both models the proportion of female directors is used as independent variable together with a number of control variables as found in the M&A literature (Faleye et al., 2011).
14 3.3.4. Probit regression model
15 3.3.5. Event study methodology
At the basis of the event study methodology is the primary assumption of efficient markets (Brown & Warner, 1985). In an efficient market, the returns around the announcement-period will reflect the wealth effect of acquisitions by gender diverse boards (Faleye, 2007). If acquisitions made by gender diverse boards are value-destroying, the firm’s stock price will experience a negative change and vice versa.
The dates at which M&A deals are disclosed are used as the date of announcement. The abnormal returns are calculated as the differences between the stock returns and the local market index from each of the 23 countries in our sample:
𝐴𝑅𝑖𝑡 = 𝑅𝑖𝑡− 𝑅𝑚𝑡
Furthermore, cumulative abnormal returns (CARs) are calculated by adding the abnormal returns during the 5 days [-2, +2] and 31 days [-15, +15] event windows.
H7: 𝐶𝐴𝑅 = 𝛼1+ 𝛽1𝐺𝑒𝑛𝑑𝑒𝑟 𝐷𝑖𝑣𝑒𝑟𝑠𝑖𝑡𝑦 + 𝛽2𝐵𝑜𝑎𝑟𝑑 𝑆𝑖𝑧𝑒 + 𝛽3𝐶𝐸𝑂 𝐷𝑢𝑎𝑙𝑖𝑡𝑦 + 𝛽4𝐼𝑛𝑑𝑒𝑝𝑒𝑛𝑑𝑒𝑛𝑡 𝐷𝑖𝑟𝑒𝑐𝑡𝑜𝑟𝑠 + 𝛽5𝐹𝐶𝐹 + 𝛽6𝑇𝑜𝑏𝑖𝑛𝑠 𝑄 + 𝛽7𝐿𝑒𝑣𝑒𝑟𝑎𝑔𝑒 + 𝛽8𝑅𝑂𝐴 + 𝛽9𝑀𝑎𝑟𝑘𝑒𝑡 𝑇𝑜 𝐵𝑜𝑜𝑘 + 𝛽10𝑀𝑎𝑟𝑘𝑒𝑡 𝐶𝑎𝑝 + 𝛽11𝐷𝑖𝑣𝑖𝑑𝑒𝑛𝑑 𝑌𝑖𝑒𝑙𝑑 + 𝛽12𝐷𝑒𝑎𝑙 𝑆𝑖𝑧𝑒 + 𝛽13𝐴𝑙𝑙 𝐶𝑎𝑠ℎ 𝐷𝑒𝑎𝑙 + 𝛽14𝐴𝑙𝑙 𝑆𝑡𝑜𝑐𝑘 𝐷𝑒𝑎𝑙 + 𝛽15𝑃𝑒𝑟𝑐𝑒𝑛𝑡𝑎𝑔𝑒 𝐴𝑐𝑞𝑢𝑖𝑟𝑒𝑑 + 𝛽16𝑀𝑜𝑟𝑒 𝐶𝑎𝑠ℎ 𝑇ℎ𝑎𝑛 𝑆𝑡𝑜𝑐𝑘 + 𝛽17𝐻𝑜𝑠𝑡𝑖𝑙𝑒 + 𝛽18𝑁𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝐵𝑖𝑑𝑑𝑒𝑟𝑠 + 𝛽19𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦 𝑅𝑒𝑙𝑎𝑡𝑒𝑑𝑛𝑒𝑠𝑠 + 𝛽20𝐶𝑟𝑜𝑠𝑠 𝐵𝑜𝑟𝑑𝑒𝑟 + 𝛽21𝑃𝑟𝑖𝑐𝑒 𝑅𝑢𝑛 𝑈𝑝 + 𝜖𝑖𝑡
3.3.6. Heckman two step model
Selectivity bias arises when the dependent variable is observed for a non-random sample and we employ this model in order to account for sample selectivity (Heckman, 1979) which can depend on some unobservable factors, such as private information, that leads the firm to engage in M&A transactions, making the decision to attempt acquisitions non-random (Li & Prabhala, 2007; Faleye et al. 2011).
16 between board gender diversity and CARs while accounting for selectivity bias in our sample (Gujarati, 2003; Kennedy, 2008;Faleye, et al. 2011). We base the first stage of selectivity in M&A transactions on variables used in the prior work, such as past performance, leverage, firm size, investment opportunities, free cash flow and CEO duality (Faleye, et al. 2011).
3.4. Data
3.4.1. Bid initiation sample
Our sample consists of 1015 non-financial public firms from US, Canada and 21 European
Countries and spans between 1st January 2002 and 31st December 2010. While we obtain board data
from Asset4 database, we gather the M&A transaction data from Thomson One Banker SDC, stock information from Datastream and accounting data from Worldscope. The Asset4 database provides Environmental, Social and Governance data at firm level. Since Asset4 begins collecting data in 2002, our sample begins in the same year. According to the management of Asset4, the number of firms covered increased yearly and Asset4 did not retroactively complete the missing previous years of newly
added firms. This fact implies that we cannot attribute the lack of data for many of our firms for years
2002 to 2004 to the lack of transparency and lower governance standards, but to the Asset4 collection practices. Although our final datasets are compiled from the same sources, they are used to test different hypotheses and after excluding financial acquirers (SIC code 60) our bid initiation dataset contains an unbalanced panel of 9136 firm-year observations and our M&A transactions dataset a total of 4059 observations. Moreover, both datasets are Winsorized at 1% and at the 99% level.
Moreover, for testing the first hypothesis we the bid initiation dataset, while to test the other hypotheses we use the M&A transactions dataset which contains the acquisitions done by the firms from the first dataset.
3.4.2. M&A transactions sample
17 Morck, Shleifer and Vishny (1990) suggest that the board of directors gets involved only on transactions that have a value above 5% of their firm’s market capitalization. However, Dowling and Aribi (2013) argue that this claim has not yet been empirically validated, reason for which our analysis does not exclude deals under the theoretical 5% threshold. A description of our sample is contained in the following three tables.
18 Table 1 Variable definitions/calculations and sources of origin
Variable Definition Source
Board characteristics
Ln Board size Number of directors on board Asset4 CEO duality 1 if the CEO is also the chairman of the board, 0 otherwise Asset4 Independent directors The number of independent directors divided by the total number of directors on
board
Asset4
Gender diversity Number of women on board divided by total number of directors on board Asset4 Number of women The number of women on board of directors
Presence of women Coded 1 if there is at least 1 woman on board, 0 otherwise
Firm characteristics
Net incomet-1 Net income in US dollars Worldscope
ROAt-1 Net income divided by total assets in US dollars Worldscope
Tobin’s Qt-1 Ratio of (market cap + total liabilities) to (common stock + total liabilities) Worldscope
Leveraget-1 Ratio of total debt to total assets Worldscope
Free cash flowt-1 The ratio of free cash flow to total assets Worldscope
Ln Market Capitalizationt-1 The total market value of the company’s outstanding shares during the previous
year
Worldscope
Ln MarkettoBookt-1 The market value of the company’s stock to its book value Worldscope
Dividend yield Annual dividends per share divided by the share’s price Worldscope Worldscope
Deal characteristics SDC
Individual acquisitions The number of acquisitions per year of each firm in our sample SDC Stake acquired The percentage of shares acquired in the target firm SDC Deal size The ratio of deal value to total assets SDC Industry relatedness Coded 1 if the first two digits of the SIC code match between target and acquirer,
0 otherwise
SDC
All cash deal Coded 1 if the deal value is paid 100% in cash, 0 otherwise SDC All stock deal Coded 1 if the deal value is paid 100% in stock, 0 otherwise SDC More cash than stock Coded 1 if the % of cash paid in transaction is larger than % of stock, otherwise 0 Hostile Coded 1 if the deal is hostile, 0 otherwise SDC Number of bidders The number of firms bidding for the target SDC Deal premium Offering price as percentage of the target book value per share SDC Price run-up CARs for the period (-30, -1) days before the announcement SDC Deal performance
CAR (-2, 2) The sum of abnormal returns for the period (-2,2) around the announcement date Datastream CAR (-15, 15) The sum of abnormal returns for the period (-15,15) around the announcement
date
19
Table 2 Country and industry distribution of the firms in our dataset and their acquisitions
Acquisition activity sample distribution by country between 2002-2010 retrieved from Asset4 and M&A transactions for the same sample retrieved from Thomson One Banker SDC.
Panel A: Country distribution of firms and M&As
Country Firms
in Asset4 Universe
From which- acquiring firms
M&A transactions undertaken by acquiring firms Austria 12 8 19 Belgium 7 6 33 Canada 95 68 270 Czech Republic 1 1 15 Denmark 10 10 20 Finland 17 16 75 France 47 41 202 Germany 30 27 112 Greece 3 2 5 Hungary 1 1 2 Ireland-Rep 9 9 42 Italy 14 12 75 Luxembourg 1 1 12 Netherlands 16 14 54 Norway 8 3 17 Poland 4 5 23 Portugal 3 3 10 Russian Fed 16 13 93 Spain 24 19 101 Sweden 24 20 101 Switzerland 22 22 96 United Kingdom 158 132 633 United States 493 436 2048 All countries 1015 869 4059
Panel B: Industry distribution of Firm and M&A
Industry Number of firms from
Asset4 dataset
Number of transactions by firms from Asset4 dataset
Mining 79 343
Construction 27 96
Manufacturing 515 2590
Transportation & Communications 96 410
Wholesale trade 30 145
Retail trade 65 190
Services 52 283
Non classifiable 151 2
All industries 1015 4059
20 and Canada (6.7%). Regarding industries, the deal distribution is shared among manufacturing (63.8%), transportation and communications (10.1%) and mining (8.5%).
Table 3 Acquisition activity sample descriptive statistics
Panel A: Annual distribution of proportion of women on board
Year Firms reporting
board gender diversity
Gender diversity Firms reporting M&A transactions
Average value spent on acquisitions/firm (mil. USD) Average number of acquisitions/firm 2002 415 8.55% 1015 131.85 0.44 2003 420 9.44% 1015 127.99 0.42 2004 613 9.24% 1015 162.67 0.44 2005 708 10.05% 1015 249.15 0.51 2006 711 10.74% 1015 280.43 0.54 2007 748 11.23% 1015 323.96 0.53 2008 876 11.20% 1015 210.06 0.40 2009 935 11.60% 1015 167.16 0.31 2010 920 12.59% 1015 125.20 0.30 All years 6346 10.81% 9135 197.61 0.43
Panel B: Summary statistics for all firm years
Variable Sample Mean Median 25th Percentile 75th Percentile St. deviation
Board characteristics
Board size 6,344 10.44 10 8 12 3.07
CEO duality dummy 6,363 0.37 0 0 1 0.48
Indep. directors 5,718 0.67 0.73 0.54 0.86 0.24 Gender diversity 6,346 0.11 0.1 0 0.17 0.1 Company characteristics Market Cap 7,776 10,337 2,976 1,131.80 9,015.56 21,011.58 Tobin’ s Q 7,516 1.86 1.53 1.18 2.14 1.14 ROA 7,775 6.63 6.53 3.35 10.27 8.02 Leverage 7,775 23.47 22.74 11.28 33.85 16.33 Net Income 7,775 599.5 163.31 43.28 511.75 1,393.54 FCF 7,775 524.62 86.56 0 410.09 1,214.95 Market/Book ratio 7,776 2.95 2.26 1 3.9 3.33
21 At the firm level, we notice that with the exception of 2008, gender diversity on boards has continued to increase from 8.6% in 2002 to 12.6% in 2010. Furthermore the number and value of acquisitions presents a rising trend between 2002-2007 after which, the trend shows a decline between 2008-2010 below the 2002 levels suggesting the severity of the financial crisis between 2007-2009.
Table 3 Acquisition activity sample descriptive statistics (continued)
Panel C: Summary statistics M&A transactions sample
Variable Sample Mean Median 25th Percentile 75th Percentile St. deviation
Board characteristics
Board size 3084 10.7 3.1 3 9 10
CEO duality dummy 3089 0.39 0 0 1 0.49
Indep. directors 2735 0.67 0.73 0.5 0.86 0.24 Gender diversity 3085 0.1 0.1 0 0.17 0.09 Company characteristics FCF 3992 0.04 0.04 0 0.07 0.08 Tobin’s Q 3994 2.92 1.98 1.29 3.02 3.27 Leverage 4042 0.24 0.23 0.13 0.33 0.14 ROA 4031 0.08 0.07 0.04 0.11 0.07 MarketBook 4058 3.18 2.58 1.4 4.2 2.88 MarketCap 4009 18025.15 4554.2 1813.39 15464.35 33708.14 Dividend Yield 4031 0.02 0.02 0 0.03 0.02 Deal characteristics Deal size 4058 0.08 0.02 0.01 0.07 0.26
All cash deal 4058 0.43 0 0 1 0.5
All stock deal 4058 0.43 0 0 1 0.5
Percentage acquired 4055 0.9 1 1 1 0.23
More cash than stock 4058 0.08 0 0 0 0.28
Hostile 4058 0 0 0 0 0.06 Number of bidders 4058 1.02 1 1 1 0.17 Industry relatedness 4058 0.32 0 0 1 0.47 Cross-border 4058 0.45 0 0 1 0.5 Price run-up 3716 0.01 0.01 -0.04 0.07 0.09 CAR (-2, 2) 3716 0.01 0.00 -0.02 0.03 0.06 CAR (-15, 15) 3716 0.02 0.02 -0.04 0.07 0.11
22
4. Results
In this section we report and discuss the results of the econometric models employed to test our hypotheses.
4.1. Gender diversity and firm acquisitiveness, industry relatedness and cross-border
transactions
To test our first hypothesis, we first count the number of yearly acquisitions each firm engages in as found in our M&A transactions sample. We then calculate the number and value of deals each firm in our sample undertakes yearly and use these two variables as dependent variables in two distinct econometric models. We use both models with robust standard errors to control for potential violation of assumptions as recommended by Cameron & Trivedi (2009).
Table 4 presents the results of the negative binomial regression ran to test the association between gender diverse boards and bid initiation or firm acquisitiveness. We find a negative and statistically significant (p-value<0.01) relationship between gender diversity, the number of women on board, the presence of women on board and the number of yearly acquisitions firms engage in. Moreover, the number of women on board is statistically significant (p-value<0.05) and negatively related to the value spent on yearly acquisitions. Regarding economic significance, each 10% increase in the percentage of women on board leads the number of bids to decrease by (=1-exp (-1.144 x 0.1)) 10.8%.
As mentioned earlier, negative binomial regression is regarded as appropriate for count data, however, this model does not make a distinction between firms which are seeking to acquire but are not completing deals and those who are deliberately not seeking to acquire, reason for which we also use a
23 Table 4 Negative binomial regression of firm acquisitiveness on gender diversity. Gender diversity is defined as the % of female directors on board. We use two more proxies for gender diversity, the number of women on board and a dummy variable coded 1 if there is at least 1 women present on the board of directors and 0 otherwise. Board size is calculated as the logarithm of the number of directors on board and log transformation is also used for market capitalization. Moreover, robust standard errors are provided in the parentheses and significance levels of 1%, 5% and 10% are indicated by ***, ** and * respectively.
24 Table 5 Zero-inflated Poisson regression of firm acquisitiveness and gender diversity. Gender diversity is defined as the % of female directors on board. We use two more proxies for gender diversity, the number of women on board and a dummy variable coded 1 if there is at least 1 women present on the board of directors and 0 otherwise. Board size is calculated as the logarithm of the number of directors on board and log transformation is also used for market capitalization and market to book ratio. Moreover, robust standard errors are provided in the parentheses and significance levels of 1%, 5% and 10% are indicated by ***, ** and * respectively.
Zero-inflated Poisson Regression: gender diversity and firm acquisitiveness Variable Number of acquisitions per year Acquisitions value per year Gender diversity 1.701*** (0.648) - - 0.618* (0.332) - - Number of women - 0.186*** (0.063) - - 0.054* (0.030) - Presence of women - 0.264* (0.154) - - 0.068 (0.070) Board size -0.014 (0.104) 0.041 (0.108) -0.002 (0.106) 0.588*** (0.196) 0.588*** (0.196) 0.588*** (0.196) CEO Duality -0.018 (0.049) -0.017 (0.049) -0.017 (0.049) 0.222*** (0.084) 0.222*** (0.084) 0.222*** (0.084) Independent directors -0.185 (0.114) -0.179 (0.114) -0.191* (0.114) -0.119 (0.208) -0.119 (0.208) -0.119 (0.208) FCF 1.599*** (0.387) 1.599*** (0.386) 1.585*** (0.387) 0.379 (0.831) 0.379 (0.831) 0.379 (0.831) Tobin’s Q -0.069** (0.030) -0.069** (0.030) -0.071** (0.030) -0.008 (0.061) -0.008 (0.061) -0.008 (0.061) Leverage -0.007*** (0.002) -0.007*** (0.002) -0.007*** (0.002) 0.002 (0.003) 0.002 (0.003) 0.002 (0.003) ROA -0.001 (0.004) -0.001 (0.004) -0.001 (0.004) 0.016** (0.007) 0.016** (0.007) 0.016** (0.007) MarketBook 0.168*** (0.042) 0.169*** (0.042) 0.171*** (0.042) -0.126** (0.054) -0.126** (0.054) -0.126** (0.054) MarketCap 0.131*** (0.020) 0.131*** (0.020) 0.130*** (0.020) 0.309*** (0.032) 0.309*** (0.032) 0.309*** (0.032) Constant -1.167*** (0.242) -1.301*** (0.255) -1.173*** (0.248) 2.376*** (0.432) 2.376*** (0.432) 2.376*** (0.032) Vuong test p-value 0.000 0.000 0.000 0.000 0.000 0.000 Observations 4364 4364 4364 4364 4364 4364 Non-zero observations 1556 1556 1556 1556 1556 1556 Pseud R2 (Prob>Chi2) <0.01 <0.01 <0.01 <0.01 <0.01 <0.01
25 odds that the gender diverse boards will not engage in the process of seeking to acquire. Moreover, the inflate coefficient for gender diversity suggests that for each percentage unit increase in gender diversity, the odds of the company not actively seeking to acquire increase by (exp (1.701)=) 5.48 %. This supports the findings from the negative binomial regression model which shows a negative relationship between gender diversity and the number of acquisitions per year. In other words, the higher the percentage of women directors on board, the more likely it is that the firm will not seek to perform M&A transactions. Regarding the significant value of the Vuong test (p-value<0.00) which tests the inflated model against the regular Poisson model, we conclude that using the zero-inflated Poisson model is more appropriate. We move further with our analysis of the influence of gender diversity on decision making around M&A transactions by looking at the likelihood that the target will be from a related industry or from a foreign country.
The specific zero-inflated Poisson interpretation holds only for the three variables used as predictors of zero-inflated zeros (gender diversity, number of women on board and presence of women on board), for the other variables the interpretation being as by the standard OLS regression. In other words, the free cash flow the company is capable of generating, the company’s market to book ratio and market capitalization are all statistically significant and positively related to the yearly number of acquisitions (p-value<0.01), while the amount of leverage and Tobin’s Q are negatively related and statistically significant at 1% and 5% confidence level, respectively.
26 Table 6 Probit regression of industry relatedness/cross-border acquisition and gender diversity. Gender diversity is defined as the % of female directors on board. We use two more proxies for gender diversity, the number of women on board and a dummy variable coded 1 if there is at least 1 women present on the board of directors and 0 otherwise. Board size is calculated as the logarithm of the number of directors on board and log transformation is also used for market capitalization and market to book ratio. Moreover, robust standard errors are provided in the parentheses and significance levels of 1%, 5% and 10% are indicated by ***, ** and * respectively.
Probit regression: gender diversity and M&A transaction characteristics of cross-border and industry relatedness
Variable Industry relatedness Cross-border acquisition Gender diversity 0.277 (0.286) - - -0.655** (0.282) - - Number of women - 0.024 (0.028) - - -0.072*** (0.027) - Presence of women - - 0.063 (0.059) - - -0.147*** (0.057) Board size -0.218** (0.099) -0.237** (0.105) -0.238** (0.103) 0.468*** (0.097) 0.530*** (0.103) 0.513*** (0.101) CEO Duality -0.120** (0.056) -0.119** (0.056) -0.120** (0.056) -0.356*** (0.054) -0.355*** (0.054) -0.354*** (0.054) Independent directors 0.258** (0.119) 0.260** (0.119) 0.251** (0.120) -0.546*** (0.114) -0.537*** (0.114) -0.532*** (0.115) FCF -0.670* (0.405) -0.671* (0.405) -0.669* (0.405) -0.659* (0.393) -0.648* (0.393) -0.662* (0.393) Tobin’s Q -0.004 (0.009) -0.004 (0.009) -0.004 (0.009) 0.001 (0.009) 0.001 (0.009) 0.001 (0.009) Leverage -0.354* (0.197) -0.352* (0.197) -0.353* (0.197) -0.313 (0.190) -0.313 (0.190) -0.314* (0.191) ROA 0.406 (0.456) 0.407 (0.456) 0.408 (0.456) 0.250 (0.438) 0.253 (0.438) 0.250 (0.437) MarketBook 0.012 (0.009) 0.012 (0.009) 0.012 (0.009) 0.003 (0.009) 0.004 (0.009) 0.003 (0.009) MarketCap 0.028 (0.018) 0.028 (0.018) 0.028 (0.018) 0.035** (0.018) 0.034* (0.018) 0.035** (0.018) Constant -0.322 (0.277) -0.281 (0.284) -0.285 (0.281) -0.814*** (0.269) -0.948*** (0.277) -0.899*** (0.274) Observations 2660 2660 2660 2660 2660 2660 Pseudo R-squared 0.008 0.008 0.008 0.040 0.040 0.040
27 of entering a cross-border acquisition by 6.6%, while the addition of one more woman would decrease the same odds by 7.2%.
Our findings suggest that the percentage of female directors negatively influences the decision to acquire foreign targets. These findings are consistent with the view that female directors are more risk-averse and will influence the board to make decisions from a more risk-averse perspective.
4.2. Gender diversity, the percentage of stock acquired, premium paid and payment method
Considering that gender diverse boards are less likely to acquire and their relationship with the amount spent on yearly acquisitions is negative and statistically significant, we move our attention to the acquisition size as proxied by the percentage of stocks acquired. After testing this relationship, we find a positive and significant relationship for the percentage of women on board, their number and the
presence of women on board at 10%, 5% and 10%, respectively. This indicates that an increase in board
gender diversity by 10% will result in an increase of 0.81% acquired stock in the M&A transaction. This finding verifies our hypothesis and shows support for the view that the percentage of female directors on board influences the acquirer firm to take a larger stake in the target.
Significant at 1% level, the percentage of independent directors is also positively related to the
percentage of stock acquired in line with the suggestions of Machold et al. (2013) that women directors
are usually independent directors. It is important to note the positive relationship between gender diverse boards and the percentage of acquired stock cannot be attributed to gender diverse boards being more risk-averse, since risk-aversion would, most likely, lead to a negative relationship. This relationship can be rather attributed to the possibility that gender diverse boards engage in fewer acquisitions while purchasing larger stakes. In other words, because they are able to recognize more of the potential outcomes and decrease uncertainy, female directors influence firms to engage only in transactions about which they are confident and in these transactions they take greater risks.
Consistent with the theory of risk-aversion, we notice that the percentage of acquired stock is statistically significant and negatively related to the deal being a cross-border M&A transaction. The same relationship can be found between industry relatedness and the percentage of stock acquired.
28 Table 7 OLS regression of percentage of stock acquired/premium paid acquisition and gender diversity. Gender diversity is defined as the % of female directors on board. We use two more proxies for gender diversity, the number of women on board and a dummy variable coded 1 if there is at least 1 women present on the board of directors and 0 otherwise. Board size is calculated as the logarithm of the number of directors on board and log transformation is also used for market capitalization and market to book ratio. Moreover, robust standard errors are provided in the parentheses and significance levels of 1%, 5% and 10% are indicated by ***, ** and * respectively.
OLS regression: gender diversity, the percentage of stock acquired and premium paid
Variable Percentage of stock acquired Premium paid Gender diversity 0.081* (0.049) - - 0.018 (0.077) - - Number of women - 0.010** (0.005) - - 0.002 (0.007) - Presence of women - - 0.018* (0.010) - - 0.063 (0.042) Board size -0.131*** (0.019) -0.139*** (0.020) -0.136*** (0.020) 0.062 (0.038) 0.060 (0.036) 0.038 (0.027) CEO Duality 0.004 (0.009) 0.004 (0.009) 0.004 (0.009) -0.052 (0.039) -0.052 (0.039) -0.056 (0.041) Independent directors 0.074*** (0.022) 0.072*** (0.022) 0.072*** (0.022) -0.044 (0.041) -0.044 (0.041) -0.062 (0.051) FCF 0.089 (0.067) 0.088 (0.067) 0.090 (0.067) 0.093 (0.078) 0.092 (0.078) 0.088 (0.079) Tobin’s Q 0.001 (0.002) 0.001 (0.002) 0.001 (0.002) 0.008 (0.012) 0.008 (0.012) 0.008 (0.012) Leverage 0.078** (0.033) 0.078** (0.033) 0.078** (0.033) 0.067 (0.121) 0.066 (0.121) 0.074 (0.122) ROA -0.001 (0.075) -0.001 (0.075) 0.001 (0.075) -0.122 (0.266) -0.121 (0.266) -0.087 (0.263) MarketBook 0.001 (0.002) 0.001 (0.002) 0.001 (0.001) 0.016 (0.010) 0.016 (0.010) 0.015 (0.010) MarketCap -0.004 (0.003) -0.004 (0.003) -0.004 (0.003) 0.002 (0.015) 0.002 (0.014) 0.001 (0.014) Deal Size 0.049*** (0.017) 0.049*** (0.017) 0.049*** (0.017) -0.025 (0.024) -0.025 (0.024) -0.027 (0.025) All Cash Deal 0.001
(0.009) 0.001 (0.009) 0.001 (0.009) -0.116 (0.086) -0.116 (0.086) -0.119 (0.087) All Stock Deal 0.001
(0.022) 0.001 (0.022) 0.001 (0.022) -0.091 (0.061) -0.091 (0.061) -0.084 (0.056) Hostile -0.029 (0.048) -0.030 (0.049) -0.029 (0.049) -0.045 (0.049) -0.045 (0.049) -0.058 (0.057) Industry relatedness -0.031*** (0.009) -0.032*** (0.009) -0.032*** (0.009) -0.061 (0.040) -0.061 (0.040) -0.060 (0.040) Cross border acquisition -0.057***
(0.009) -0.057*** (0.009) -0.057*** (0.009) 0.051 (0.038) 0.052 (0.038) 0.056 (0.041) Constant 1.193*** (0.051) 1.212*** (0.053) 1.205*** (0.052) -0.036 (0.109) -0.032 (0.118) 0.001 (0.111) Observations 2658 2658 2691 507 507 507 R-squared 0.069 0.069 0.069 0.043 0.043 0.057
29 Table 8 Probit regression of all stock deal/all cash deal payment and gender diversity. Gender diversity is defined as the % of female directors on board. We use two more proxies for gender diversity, the number of women on board and a dummy variable coded 1 if there is at least 1 women present on the board of directors and 0 otherwise. Board size is calculated as the logarithm of the number of directors on board and log transformation is also used for market capitalization and market to book ratio. Moreover, robust standard errors are provided in the parentheses and significance levels of 1%, 5% and 10% are indicated by ***, ** and * respectively.
Probit regression: gender diversity and payment method Variable All stock deal All cash deal
Gender diversity -1.517** (0.714) - - 0.624 (0.277) - - Number of women - -0.104 (0.065) - - 0.058 (0.268) - Presence of women - - -0.285*** (0.112) - - 0.197*** (0.057) Board size 0.347* (0.201) 0.390* (0.211) 0.423** (0.200) -0.326*** (0.095) -0.373*** (0.100) -0.397*** (0.099) CEO Duality -0.291** (0.126) -0.295** (0.125) -0.305** (0.127) -0.048 (0.053) -0.047 (0.053) -0.053 (0.053) Independent directors 0.532* (0.279) 0.498* (0.278) 0.555** (0.274) 0.216* (0.113) 0.217* (0.113) 0.178 (0.114) FCF -0.479 (0.744) -0.503 (0.746) -0.489* (0.756) -0.062 (0.393) -0.065 (0.393) -0.067*** (0.392) Tobin’s Q -0.001 (0.021) 0.001 (0.021) -0.001 (0.021) 0.009 (0.009) 0.010 (0.009) 0.009** (0.009) Leverage -0.974** (0.405) -0.982** (0.404) -1.001** (0.405) -0.101 (0.188) -0.099 (0.188) -0.107 (0.189) ROA -1.086 (0.952) -1.076 (0.958) -1.119*** (0.960) 0.469 (0.433) 0.468 (0.433) 0.466*** (0.433) MarketBook 0.028 (0.018) 0.028 (0.018) 0.027 (0.018) -0.018** (0.009) -0.018** (0.009) -0.018* (0.009) MarketCap -0.061 (0.040) -0.063 (0.040) -0.060 (0.040) 0.002 (0.017) 0.003 (0.017) 0.002 (0.017) Deal size 0.687*** (0.213) 0.681*** (0.215) 0.688*** (0.214) -0.445*** (0.137) -0.446*** (0.138) -0.445*** (0.136) Hostile 0.883 (0.548) 0.848 (0.554) 0.889 (0.565) 0.105 (0.438) 0.110 (0.436) 0.092 (0.441) Cross-border -0.682*** (0.131) -0.668*** (0.130) -0.680*** (0.130) -0.061 (0.051) -0.060 (0.051) -0.057 (0.051) Constant -1.912*** (0.536) -2.006*** (0.558) -2.060*** (0.531) 0.506* (0.267) 0.609** (0.274) 0.636** (0.271) Observations 2660 2660 2660 2660 2660 2660 Pseudo R-squared 0.142 0.138 0.138 0.111 0.111 0.013
When looking at the payment method, we find a negative relationship between gender diversity and all stock deals statistically significant at 5% level and for the presence of women statistically significant at 1% level. The positive coefficients in the model where we used all cash deal as our dependent variable, although not statistically significant, suggest that gender diverse boards are more likely to pay in cash for acquisitions.
30
between target and acquirer, gender diverse boards are more likely to choose payments in cash taking upon their company the entire risk and reward presented with the M&A transaction.
This finding points to the idea mentioned above that gender diverse boards can be capable of decreasing uncertainty of the outcomes of the M&A transactions in which they engage, thus being aware
of the risk and rewards and choosing not to share it with the target.
We further notice, from the sign and statistical significance of the coefficient for board size, that bigger boards are more willing to pay with stock than with cash and that in the situations where the CEO is not the chairman of the board, the acquirer is more likely to pay in stock. Regarding the percentage of independent directors, it seems that its sign is positive for both methods of payment. This suggests that independent directors are proponents of both methods of payment, although the coefficient for the variable representing their percentage and the significance level, is higher for stock payments. We acknowledge however that this is not clear evidence for inferring the influence of the percentage of independent directors.
4.3. Gender diversity and cumulated abnormal returns (CARs)
To test our hypothesis that gender diversity will have a positive influence on the cumulated abnormal returns around the date of announcement we use 5 and 31 days CARs with an OLS regression. From Table 9 we observe a negative and statistically significant relationship between gender diversity, and number of women and the 5 days cumulated abnormal returns at 10% and 5% respectively. Regarding the 31 days CARs, we find that the increase by 1% in gender diversity on the board of directors, keeping all other things constant, will result in a 0.021% (approx. 2% annualized) decrease in the CARs 5 days and 0.039% (approx. 4.8% annualized) in the CARs 31 days, contrary to what we predicted. We also notice that the board size and the market capitalization of the firm have a negative influence on the CARs 5 days at 1% level, and so does the dividend yield the firm is paying (p-value<0.1). Moreover, the price run-up, which proxies the performance of the stock before the announcement is under all specifications, as expected positively related and statistically significant (p-value<0.01)
31 Table 9 OLS regression of CARs 5 days/CARs 31 days and gender diversity. Gender diversity is defined as the % of female directors on board. We use two more proxies for gender diversity, the number of women on board and a dummy variable coded 1 if there is at least 1 women present on the board of directors and 0 otherwise. Board size is calculated as the logarithm of the number of directors on board and log transformation is also used for market capitalization and market to book ratio. Moreover, robust standard errors are provided in the parentheses and significance levels of 1%, 5% and 10% are indicated by ***, ** and * respectively.
OLS regression: gender diversity and M&A announcement CARs
Variable CARs (-2,2) CARs (-15, 15)
Gender diversity -0.021* (0.011) - - -0.039* (0.021) - - - Number of women - -0.002** (0.001) - -0.003* (0.002) - Presence of women - - -0.002 (0.002) - - -0.008* (0.004) Board size -0.010*** (0.004) -0.009** (0.004) -0.010** (0.004) -0.011 (0.007) -0.009 (0.004) 0.009 (0.007) CEO Duality -0.002 (0.002) -0.002 (0.002) -0.002 (0.002) -0.004 (0.004) -0.005 (0.004) -0.004 (0.004) Independent directors -0.008* (0.005) -0.008* (0.005) -0.009* (0.005) -0.014* (0.008) -0.014* (0.008) -0.013 (0.008) FCF 0.016 (0.015) 0.016 (0.015) 0.015 (0.015) 0.023 (0.029) 0.023 (0.029) 0.023 (0.029) Tobin’s Q 0.001 (0.001) 0.001 (0.001) 0.001 (0.001) -0.001 (0.001) -0.001 (0.001) -0.001 (0.001) Leverage -0.009 (0.008) -0.009 (0.008) -0.010 (0.008) -0.014 (0.015) -0.014 (0.015) -0.014 (0.015) ROA 0.023 (0.019) 0.023 (0.019) 0.023 (0.019) 0.037 (0.039) 0.037 (0.039) 0.037 (0.039) MarketBook 0.001 (0.001) 0.001 (0.001) 0.001 (0.001) 0.001* (0.001) 0.001* (0.001) 0.001* (0.001) MarketCap -0.002*** (0.001) -0.002*** (0.001) -0.002*** (0.001) -0.005*** (0.001) -0.005*** (0.001) -0.005*** (0.001) Dividend Yield -0.084* (0.049) -0.083* (0.049) -0.086* (0.049) -0.413*** (0.094) -0.412*** (0.094) -0.421*** (0.095) Deal size -0.009 (0.007) -0.009 (0.007) -0.009 (0.006) -0.004 (0.008) -0.004 (0.008) -0.004 (0.008)
All cash deal 0.001
(0.002) 0.001 (0.002) 0.001 (0.002) -0.002 (0.004) -0.002 (0.004) -0.001 (0.004) All stock deal -0.007
(0.010) -0.007 (0.010) -0.007 (0.010) 0.020 (0.016) 0.020 (0.016) 0.019 (0.016) Percentage acquired -0.005 (0.005) -0.005 (0.005) -0.005 (0.005) -0.022*** (0.008) -0.022*** (0.008) -0.022*** (0.008) More cash than stock -0.013*
(0.007) -0.013** (0.007) -0.013* (0.007) -0.023* (0.012) -0.023* (0.012) -0.023* (0.012) Hostile -0.006 (0.019) -0.006 (0.019) -0.007 (0.018) 0.021 (0.034) 0.021 (0.034) 0.020 (0.034) Number of bidders 0.001 (0.008) 0.001 (0.008) 0.001 (0.008) -0.003 (0.012) -0.004 (0.012) -0.003 (0.012) Industry relatedness 0.002 (0.002) 0.002 (0.002) 0.002 (0.002) 0.002 (0.004) 0.002 (0.004) 0.002 (0.004) Cross-border 0.002 (0.002) 0.002 (0.002) 0.002 (0.002) 0.002 (0.004) 0.002 (0.004) 0.002 (0.004) Price run-up 0.053*** (0.014) 0.053*** (0.014) 0.053*** (0.014) 0.483*** (0.030) 0.483*** (0.030) 0.483*** (0.030) Constant 0.062*** (0.014) 0.058*** (0.014) 0.061*** (0.014) 0.123*** (0.026) 0.117*** (0.026) 0.118*** (0.026) Observations 2479 2479 2479 2479 2479 2479 R-squared 0.041 0.041 0.040 0.234 0.234 0.234
For robustness purposes, we run an independent t-test with equal variances to compare the
means of two groups: firms situated under the 25th percentile on the proportion of women on board, and
firms above the 75th percentile for each measure of performance, CAR2 and CAR15. In order to perform
32
the 75th percentile and 0 if it is under the 25th percentile, ultimately this subsample containing only the
two ranges (1st-25th, 75th-100th). Then we use the independent t-test by grouping according to this
dummy variable. The results support our current findings, with p-values smaller than 0.01 for both models, CAR2 and CAR15, leading us to reject the null hypothesis that the means of these two subsamples are equal. Moreover, the t-test findings indicate that, for both measures of performance, the mean of the CARs for the M&A announcement of firms with the proportion of women on board under
the 25th percentile is higher than the mean of the CARs for the M&A announcement of firms with the
proportion of women on board above the 75th percentile.
Table 10 Heckman twostep model of CARs 5 days/CARs 31 days and gender diversity. Gender diversity is defined as the % of female directors on board. We use two more proxies for gender diversity, the number of women on board and a dummy variable coded 1 if there is at least 1 women present on the board of directors and 0 otherwise. Board size is calculated as the logarithm of the number of directors on board and log transformation is also used for market capitalization and market to book ratio. Furthermore, we report the first stage variables at the top of the table, variables regarded as influencing the decision to engage in M&A transactions according to Faleye et al. (2011). Moreover, robust standard errors are provided in the parentheses and significance levels of 1%, 5% and 10% are indicated by ***, ** and * respectively.
Heckman twostep model: gender diversity and M&A announcement CARs
Variable CAR (-2,2) CAR (-15, 15)
