• No results found

Women on the Board: The Effect of Female Directors on the Returns of U.S. Acquirers

N/A
N/A
Protected

Academic year: 2021

Share "Women on the Board: The Effect of Female Directors on the Returns of U.S. Acquirers"

Copied!
37
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

Women on the Board: The Effect of Female Directors

on the Returns of U.S. Acquirers

MSc Finance Thesis – semester II 2019-2020

University of Groningen, Faculty of Economics and Business Word count (incl. tables and references): 8751

Supervisor: Prof. Dr. Niels Hermes

Henning Kruse S3723771 h.kruse@student.rug.nl

Date: 4.6.2020

Abstract

A large literature documents the differences between women and men in corporate decision making, but little is known about the effect of women on acquisitions. This research investigates the relationship between the fraction of female directors on a board and the value created in acquisitions. Based on the existing literature and the fact that women tend to be less overconfident than men, a positive relation is expected. Acquirers generate significantly positive returns over the specified event window. Analysing the relation between the fraction of female directors and the generated return does not find a significant relation. Next to that this research also shows that there are no significant differences between gender-homogeneous and gender-heterogeneous boards. Lastly, boards with a fraction of female directors above a 25% threshold generate significantly lower returns than companies with a fraction below the threshold. However, it is found that there is no relation between the value that is created in acquisitions and the fact that a board has more or less than 25% female directors.

(2)

Introduction

The question whether women should play an increasing role in corporate decision making is around for quite a while and is still gaining more and more attention. During the recent years multiple countries decided to legally require a specified percentage of women in the management and especially the board of companies. The “2019 Report on equality between women and men in the EU”1 by the European commission reports that the fraction of female board members increased from 22.7% to 26.7% between 2016 and 2019. Furthermore, it is stated that this trend is continuing meaning that the fraction of female board members will further increase in the future. This trend is supported by possibly new legislation requiring companies to have a specified minimum of female board members.

There is a considerable body of literature investigating the effect of gender differences on the board on different aspects of the company. The majority of the literature concerning this topic is considered with either social performance or financial performance. Despite the fact that there is already a considerable amount of literature concerning this topic, almost no one investigated the effect of female directors on value creation in M&A deals. Therefore, this research will focus on aspects of the financial performance, especially concerning the performance in M&A deals. The reason for that is that M&A is one of the main growth strategies for a lot of companies and it can have a huge influence on shareholder value. The starting point of this research is the widespread opinion that male executives are more overconfident than their female counterparts, possibly leading to decisions that are bad for the value that is created for shareholders. Multiple articles suggest a negative relationship between the fraction of female board members and the risk a company is taking (Palvia, Vähämaa, and Vähämaa, 2015; Huang and Kisgen, 2013). Besides that, multiple women on the board are generally associated with increased financial performance leading to an increase in shareholder value (Vafaei, Ahmed, and Mather, 2015; Horak and Cui, 2017; Liu, Wei, and Xie, 2014; etc.). The relationship between female board members and M&A deal characteristics is not well researched and gives quite some room for new research trying to explain this relationship. Levi, Li, and Zhang (2014) find that more female directors lead to less bid initiations and lower bid premiums. To my knowledge there is no paper that tries to investigate the relationship between female directors and the value creation in acquisitions.

The majority of research regarding M&A deal announcement returns finds no positive significant returns but rather that acquisitions destroy shareholder value for bidder firms. Based on the perception that women are less overconfident and the fact that they tend to take less risk

1

(3)

and increase financial performance, one would expect boards with a higher fraction of female board members to make better decisions regarding acquisitions. That is why this research will investigate whether a higher fraction of female board members is associated with higher acquisition announcement returns. To be able to do so, this research tries to answer the following research question:

Does a higher fraction of female directors on the board have a positive effect on the value creation in acquisitions?

This paper contributes to the strand of literature that is concerned with gender differences in the context of finance. It adds especially to the strand that is concerned with the relationship between female directors and acquisition announcement returns. The purpose of this study is to investigate whether a larger fraction of female directors is associated with higher acquisition announcement returns.

The results show that on average acquirers generate significantly positive returns. Controlling for different firm and deal characteristics results in no significant relation between the fraction of female directors and the value that is created in acquisitions. When investigating different subsamples, it is found that boards with no female directors do not generate significantly higher returns than boards with at least one female director. Following the findings of the literature, it is analysed whether boards with a fraction of female directors above a certain threshold generate different returns. The threshold is defined by testing different fractions and whether there are significant differences between the subsamples. It is found that boards with a fraction of female directors below a threshold of 25% generate significantly higher returns. Analysing whether these differences are due to the fraction of female directors, gives insignificant results. This means that there is no relation between the fraction of female directors and the generated value in an acquisition.

(4)

Literature review

This research is related to prior studies that investigate the effect of gender differences in business and especially finance. The fact that there are differences in the behaviour of women and men cannot be denied. Due to less overconfidence, women are seen to be more cautious and tend to take less risks. People that are overconfident tend to overestimate their precision of knowledge and the effect they have on the outcome of decisions they make (Barber and Odean, 2001). Furthermore, it leads to decisions that are more likely to have a negative net present value (Huang and Kisgen, 2013) possibly leading to acquisitions that destroy value.

In contrast to the findings of gender differences in general behaviour, studies that investigate gender differences in a financial setting do not find consistent results of significant statistically or economically differences.

Due to less overconfidence, women tend to take less risk compared to men in their mutual fund investments. However, these results are not as strong when controlling for financial investment knowledge (experience and education), suggesting that previous studies are upward biased due to less specific investment knowledge controls (Dwyer, Gilkeson, and List, 2002). This means that there are almost no differences in risk taking between male and female fund managers. Atkinson, Baird, and Frye (2003) also find that female and male fund managers are similar with regards to risk taking, performance, and other fund characteristics. Based on this no differences between men and women are assumed when it comes to risk taking.

(5)
(6)

Table1. Overview of the most important literature concerning the effect of female directors on firm performance

Authors Sample

period

Market Sample

size

Performance measure Results

Vafaei, et al. (2015) 2005-2011 Australia 500 ROA, ROE, Tobin's Q +

Horak and Cui (2017) 2008-2011 China 66 ROE, Asset growth, sales

growth

+ (except ROE) Reguera-Alvarado, de Fuentes, and Laffarga (2017) 2005-2009 Spain 125 Tobin's Q +

Liu, et al. (2014) 1999-2011 China 16,964 ROA, ROS +

Galbreath (2018) 2004-2005 Australia 296 ROE + (indirect through CSR)

Li and Chen (2018) 2007-2012 China 2,936 Tobin's Q +

Bennouri, Chtioui, Nagati, and Nekhili (2018) 2001-2010 France 394 ROA, ROE, Tobin's Q Tobin's Q : - ROA & ROE: + Chapple and Humphrey (2014) 2004-2011 Australia 300 One factor model & four

factor model /

Owen and Temesvary (2018) 1999-2015 US 90 ROA, Stock price growth,

Sharpe ratio non-linear

(7)

Levi, et al. (2014) investigate whether gender enhances CEO empire building and the effect on bid premiums. They identify that each additional female director corresponds to a drop of 7.6% in acquisition bids. Next to that, the premium offered in an acquisition drops by 15.4% for each additional female director on the bidders’ board, supporting the notion that female board members help to create shareholder value. Chen, et al. (2016) and Dowling and Aribi (2013) support the finding that the presence of more female directors reduces the amount of acquisition bids of a company. Companies with less experience in acquisitions generate lower returns than companies with more experience (Canina, Kim and Ma, 2019; Boateng, Du, Bi, and Lodorfos, 2010). Having more women on the board leads to less acquisitions which could lead to lower announcement returns. However, Huang and Kisgen (2013) investigated whether there are differences in announcement returns, depending on the fact whether this announcement was made by a male or female CEO. Acquisitions made by companies with a male CEO seem to have approximately 2% lower announcement returns than those made by female CEOs. This is seen as evidence for male overconfidence in significant corporate decision making compared to their female counterpart. Table 2 presents a summary of the studies concerning the effect of female directors on different characteristics in M&A deals.

Table 2. Overview of the most important literature concerning the effect of female directors on M&A

Authors Sample

period Market Sample size Topic Measure Results

Levi, et al.

(2014) 1997-2009 US 1500 Bid initiation and bid premium 1. log (bid initiation) 2. bid premium 1. - 2. - Chen, et

al. (2016) 1998-2010 US 13,248 (H2: 2,825) Corporate acquisition intensity and size 1. number of acquisitions 2. total value of all transaction scaled by anual sales of bidder 1. - 2.- Dowling and Aribi (2013) 2000-2011 GB 364 Female directors and acquisitiveness number of acquisitions per year -

(8)

not tend to overestimate their precision of knowledge and the effect they have on the outcome of decisions they make. Next to that women are likely to see future outcomes less favourable than men do (Barber and Odean, 2001; Malmendier and Tate 2008). Both of these types show that women are less overconfident than men. Based on this one would expect companies with more female directors to make better decisions regarding acquisitions.

This research will investigate whether the presence of women on the bidders’ board is associated with higher acquisition announcement returns. This assumption is made based on the fact that decision made with less overconfidence are likely to yield higher returns. Making that assumption, the following hypothesis is developed:

H1. The fraction of female directors on a board is positively associated with announcement returns in acquisitions.

(9)

Methodology

Sample selection and criteria

The deal data that is used during this research is taken from Zephyr database. Data with regards to board composition is collected from the BoardEx database, whereas the company financials are taken from COMPUSTAT North America database and CRSP database. For the purpose of this research the sample is limited to acquirers and targets that are located within the United States. Literature suggests that there are differences in value creation between acquirers from different countries, making the outcome for one country not generalizable for other countries. Another reason for a single country sample is the fact that different countries have different requirements with regards to the minimum fraction of female directors on the board. By using U.S. only based companies, the sample should not be biased due to different regulations. Given the fact that most of the available data and literature is based on U.S. data, a sample with U.S. based acquirers and targets seems to be the most appropriate approach for filling the corresponding gap in the literature. Furthermore, only deals that were completed at the start of this research were included. In order to avoid making a choice of who is the acquiring part in a merger, only acquisitions where more than 50% (controlling interest) is bought are taken into consideration. In the next step, deals without a known deal value are eliminated. To ensure a sample that is as large as possible and therefore representative, only the acquirer has to be listed on a stock exchange. Targets can either be listed or private. Table 3 shows how many events remain in the sample after adding each single selection criteria. The sample started with 216,601 events and after adding all the criteria the sample was left with 824 acquisitions over the 2005-2019 period. Table 4 shows the frequency of events throughout the years.

Table 3. Selection criteria

Selection criteria Search result

1. Country: United States of America (bidder and target) 216,601

2. Current deal status: Completed - confirmed 184,977

3. Deal type: Acquisition

82,772

4. Sub-deal type: Acquisition with more than 50% 9,976 5. Deal value (USD): all deals with known value 5,961 6. Listed acquiror and listed and unlisted target

1,109

7. Time period: on and after 01/01/2005 and up to and including 31/12/2019 824

(10)

Table 4. Announcements per year

Year Amount Percentage

2019 38 4.61% 2018 55 6.67% 2017 54 6.55% 2016 65 7.89% 2015 70 8.50% 2014 88 10.68% 2013 58 7.04% 2012 69 8.37% 2011 62 7.52% 2010 55 6.67% 2009 20 2.43% 2008 40 4.85% 2007 57 6.92% 2006 39 4.73% 2005 54 6.55% Total 824 100.00%

Event study methodology

For the purpose of this research the Event Study Methodology, which was first described by McKinlay (1997), is used. An event study investigates the effect of an unexpected corporate event on the stock performance of the corresponding firm by calculating abnormal returns (AR) around the event day. The event window starts 30 day before and ends 30 day after the announcement day. A longer event window is chosen to make sure to capture all possible effects such as leakage effects or post announcement drifts (MacKinlay, 1997). Abnormal returns are calculated by taking the difference between the expected return and the actual return. The expected return is calculated by using a single index model, where ai and bi are determined using a 300-tradingday estimation window. !!" is defined as the return on the corresponding market index. Therefore, the expected return is calculated as follows:

!#" = ## + %#!!"+ &#" (1)

(11)

Abnormal returns are calculated by subtracting the expected return from the actual return (!#%). The calculation is as follows:

/!#% = !#%− #1#− %2#!!% (2)

with τ representing the event time.

The cumulative abnormal return is defined as the sum of the abnormal returns between two predefined days in the even window.

3/!#(4&, 4$) = ∑%%'%! "/!#% (3)

The average abnormal return is calculated by dividing the sum of all abnormal returns by the number of events. /! 7777% = & (∑ /!#% ( #'& (4)

Adding all the average abnormal returns between two predefined days in the event window results in the cumulative average abnormal return.

3/!

777777(4&, 4$) = ∑%%'%! "/!7777% (5)

Figure 1 provides an overview of the development of the average abnormal return (AAR) and the cumulative average abnormal return (CAAR) during the event window. As can be seen, the announcement to acquire another company seems to have a positive effect on the stock performance of the bidding firm. Especially the announcement day and the day after do stick out with both being statistically significant at the 1% level. However, based on the magnitude of both abnormal returns, one could argue that they are not economically significant, which would be in line with the majority of the literature saying that bidder to not earn positive abnormal returns in acquisition announcements.

(12)

Figure 1. AAR & CAAR over the event period

Variables, Descriptive statistics and correlation

The dependent variable that will be used to analyse the relationship between value creation and the fraction of female directors will be a cumulative abnormal return. To be able to decide which cumulative abnormal return to use for the regression, first the abnormal returns have to be tested for significance. Therefore, one parametric and two non-parametric tests are

employed. The first test that was conducted, is the standardized cross-sectional t test for abnormal returns. While common event study methods reject true nulls too frequently, this test avoids this problem without reducing the test’s power significantly (Boehmer, Masumeci, Poulsen, 1991). The standardized cross-sectional is less powerful than the non-parametric tests but can provide a useful robustness check (Campbell, Cowan and Salotti, 2010). It is tested if 8): AAR = 0. First the abnormal returns are standardized as follows:

9/!#(:&, :$) = /!#(:&, :$)/<*+#(-",-!) (6)

Where <*+#(-",-!)is the standard deviation across firms during the event window.

<*+#(-",-!) = ( & 0#1$ ∑ /!#2 $ 13& 2'133& )&/$ [?#@1 + 50# #+ (∑)!%*)"+$%15#+7$_'(%)! ∑+,"%*+,,"8+$%1 +7$_'(%:! B]&/$ (7)

(13)

mean daily national market-index return in the estimation window. The standardized cross-sectional statistic is calculated as follows:

E" = ∑ >*+#(-",-!) -#*" √(=./0 (8) where <>*+ = [(1&& ∑( (9/!#

#'& (:&, :$) −(&∑(#'&(9/!#(:&, :$) (9)

Both Cowan’s generalized sign test and Corrado’s rank test are used to ensure balanced power and correct specification (Campbell, Cowan and Salotti, 2010). Overall, the rank test seems to perform better in an ideal setting. The generalized sign test performs better than other non-parametric tests in event windows that are longer than two days (Cowan, 1992). Furthermore, it is not as sensitive to return variance increases on the event date and thin trading. However, both tests show correct specification under most circumstances (Campbell, Cowan and Salotti, 2010; Corrado, 2011; Cowan, 1992). Cowan’s generalized sign Test tests whether the fraction of positive CAR in the event window is in line with the fraction of positive CAR in the estimation window. The positive fraction of CAR is calculated as follows:

F̂ = (& ∑ 0& #

(

#'& ∑-"'-" 19#," (10)

Where 9#," is 1 if the sign is positive and 0 otherwise. D# is the number of non-missing returns in the estimation window for event i. The Generalized Sign Test statistic that tests 8): CAAR

= 0 is

H@=#@A= (B1(CD)

E(CD(&1CD) (11)

where w is the number of positive CAR during the event window.

Corrado’s (1989) transforms the measured abnormal returns into their respective ranks. The test statistic to test 8): AAR = 0 is given by:

IFGA2 =HI "

-∑-#*22#1J1 27K

(14)

where J7 = 3L)& ∑ (& %∑ J#" (% #'& M3) N'133) (13)

With K" being the number of non-missing returns across firms on day t and

<2 = (3L)& ∑ [((& %∑ J#" (% #'& ) M3) N'133) − J7]$)&/$ (14)

(15)

Table 5. Descriptive statistics of daily abnormal returns in the event window (-30, 30)

Day AAR CAAR % pos % neg

Standardized cross-sectional t test Cowan‘s generalized sign test Corrado rank test -30 -0.01% -0.01% 48.30% 51.70% -0.11619 0.35073 -0.57912 -25 -0.11% -0.35% 45.75% 54.25% -1.78187* -1.31942 -1.66478 -20 0.07% -0.49% 50.36% 49.64% 0.57144 1.25539 0.46549 -15 0.10% -0.37% 49.51% 50.49% 0.92101 0.90745 0.2641 -10 0.10% -0.56% 50.73% 49.27% 1.70357* 1.53375* 0.27817 -9 0.16% -0.41% 52.91% 47.09% 1.81169* 2.71677*** 1.52331 -8 -0.07% -0.48% 49.27% 50.73% 0.10537 0.83786 0.28464 -7 0.02% -0.46% 51.09% 48.91% -0.03474 1.67293** 0.13163 -6 -0.06% -0.53% 49.64% 50.36% -0.97385 0.97704 -0.82353 -5 -0.03% -0.55% 48.06% 51.94% -1.1019 -0.1364 -1.04319 -4 0.13% -0.43% 49.64% 50.36% 1.89109* 1.04662 0.97147 -3 0.01% -0.41% 49.51% 50.49% 0.71363 0.97704 0.23795 -2 -0.02% -0.43% 48.67% 51.33% 0.21618 0.42032 -0.09732 -1 0.02% -0.42% 49.03% 50.97% -0.39202 0.5595 -0.29335 0 0.77% 0.35% 54.13% 45.87% 4.54319*** 3.48225*** 3.78408*** 1 0.35% 0.70% 53.28% 46.72% 2.82751*** 2.99513*** 2.93326*** 2 0.02% 0.71% 48.91% 51.09% -0.09434 0.35073 -0.41092 3 -0.04% 0.67% 50.97% 49.03% -0.0405 1.60334* 0.21516 4 0.00% 0.68% 50.61% 49.39% 0.59474 1.39457* 0.32964 5 -0.04% 0.63% 47.82% 52.18% -0.58292 0.00278 -0.51527 6 -0.03% 0.61% 47.69% 52.31% -0.25976 -0.1364 -1.29099 7 -0.04% 0.57% 50.12% 49.88% 0.59696 1.32498* 0.33048 8 0.00% 0.57% 49.03% 50.97% -0.26148 0.62909 -0.53861 9 0.07% 0.64% 48.79% 51.21% 0.79879 0.35073 0.18648 10 0.03% 0.67% 50.49% 49.51% 0.42359 1.25539 0.4427 15 0.06% 0.75% 49.51% 50.49% 0.32938 0.76827 0.32401 20 -0.10% 0.36% 48.06% 51.94% -0.88379 -0.20598 -1.00579 25 -0.04% 0.31% 49.82% 50.18% 0.34213 0.97704 -0.24471 30 0.04% -0.02% 49.82% 50.18% 0.52931 0.76827 0.21374

***- Asterisks indicate the significance level: * significant at the 10% level, ** significant at the 5% level, *** significant at the 1% level.

AAR is defined as the average abnormal return, CAAR is defined as the cumulative average abnormal return, %pos is defined as the fraction of positive abnormal returns at that event day, %neg is defined as the fraction of negative abnormal returns at that event day

(16)

the generalized sign test also shows significance at the 10% level for day three and four after the event. Due to the fact that this is the only test it is not enough evidence to include these days in the event period used in the regression. Next to that different subsamples, which are discussed later, are individually tested for significance. Therefore, the same significance tests as for the overall sample are done. While the subsample with at least one female director does not result in significant returns on day one after the event, the subsample with a fraction above the threshold gives inconsistent results for the different tests. The other two subsamples give the same results as for the overall sample. Based on this the CAR (0, +1) is kept and used for the regression. The daily abnormal returns for the different subsamples are presented in the Appendix B, C, D and E. Following the results of all three significance tests, the two-day event window (0, +1) starting at the event date is chosen for the regression. Based on the above-mentioned tests and their result it can be assumed that there are no leakage effects or post announcement drifts.

A potential problem could be caused by the presence of other acquisitions of the same acquirer in the same event window. However, excluding the events that could be potentially biased and repeating the significance tests, gives the same result for significance of the abnormal returns and therefore the same event period (0, +1). The results of this sample can be found in Appendix F. Based on this analysis the effected events are not excluded from the dataset in order to keep a sample that is as big and representative as possible.

(17)

Table 6. Description of variables

Variable Explanation

Dependent variable

CAR (0, +1) Cumulative abnormal return over the

0, +1 event window

Independent variables

Firm characteristics

FFDB

FCEO

Fraction of female directors on the board of the bidder

Dummy (1 if bidder CEO is female, 0 otherwise)

MC Market capitalization (bidder)

BBS

ROA

Number of directors serving on a corporate board (bidder)

Return on Assets (bidder)

Deal characteristics MP Year DV &#

Method of payment; Dummy (1=cash, 0 otherwise)

Year of announcement (dummy for each year)

Deal Value (Price paid in acquisition)

Error term with zero mean and constant variance

Tables 7 and 11 show the descriptive statistics for the full sample, the subsample with no female directors and the subsample with at least one female director. There seem to be differences in the continuous variables between the different subsamples. To test if these differences are significant, propensity-score matching (PSM) is conducted.

(18)

to reduce the bias that could be caused by other variables when just comparing the CARs and the fraction of female directors. PSM matches each observation from one subsample to one observation from the other subsample. In these matches the observations are similar for each variable except for the CAR and the fraction of female directors. Comparing these matches shows whether there are significant differences between the subsamples and if these differences are due to the fraction of female directors. The results of this test are shown in Table 8. Based on the p-value of 0.234 it can be said that board with no female directors do not generate significantly different returns than boards with at least one female director.

(19)

Table 7. Descriptive statistics of continuous variables for the full sample and subsamples

Full sample No female director At least one female director

(N=824) (N=166) (N=658)

Mean Std.dev. Median Mean Std.dev. Median Mean Std.dev. Median

Continuous variables

CAR (0, +1) 0.0112 0.0644 0.0037 0.0299 0.1056 0.0110 0.0065 0.0478 0.0027

Fraction female directors 0.1435 0.1007 0.1429 0 0 0 0.1797 0.0786 0.1668

Bidder board size 9.66 2.46 9.00 7.78 1.92 8.00 10.14 2.35 10.00 Deal size (m USD) 591.42 2,809.38 200.805 378.30 710.26 134.00 645.18 3,121.82 217.50 Market cap (m USD) 32,728.84 88,461.56 3,615.92 4,059.42 12,017.51 994.67 39,961.55 97,500.35 4,831.09

ROA 0.1245 0.1412 0.1330 0.0863 0.2491 0.1170 0.1340 0.0953 0.1360

Table 8. PSM between the no female against at least one female director sample

AI Robust

CAR01 Coef. Std. Err. z P>|z| [95% Conf. Interval]

Zerofemale

(20)

Table 9. PSM between the above and below threshold sample

AI Robust

CAR01 Coef. Std. Err. z P>|z| [95% Conf. Interval]

Threshold

(1 vs 0) .0073216 .0032867 2.23 0.026 .0008797 .0137635

Table 10. Descriptive statistics of continuous variables for the subsamples

Fraction female directors above threshold Fraction female directors below threshold

(N=100) (N=724)

Mean Std.dev. Median Mean Std.dev. Median

Continuous variables

CAR (0, +1) 0.005 0.05 0.00 0.012 0.065 0.00

Fraction female directors 0.32 0.05 0.3 0.12 0.079 0.13

Bidder board size 10.77 2.36 10.50 9.51 2.44 9.00 Deal size (m USD) 814.45 1,492.06 300.00 560.61 2,994.75 195.00 Market cap (m USD) 56,777.67 99,545.99 12,334.41 29,407.18 86,370.57 3,230.13

(21)

Table 11. Descriptive statistics of dummy variables for the full sample and subsamples

Full sample No female director At least one female director Fraction above threshold Fraction below threshold

(22)

To test the variables for multicollinearity both an informal and a formal test are conducted. A correlation matrix is created and the variance inflation factor (VIF) is calculated. Table 12 shows the correlation between the continuous variables in the full sample.

Table 12. Correlation between continuous variables

Fraction of female

directors Bidder boards size Market cap Deal value ROA Fraction of female directors 1 Bidder boards size 0.03 1 Market cap 0.03 0.17 1 Deal value 0.01 0.08 0.05 1 ROA 0.01 0.07 0.18 0.02 1

In order to make sure that none of the outliers biases the analysis, all continuous variables are either winsorized or the natural logarithm is taken. For the variables deal size and market capitalization the natural logarithm is taken. The variables for the relevant cumulative abnormal return, the return on assets, the fraction of female directors and the bidder board size are winsorized. For all the continuous variables the skewness is below 0.32 and the kurtosis is between 2.42 and 3.36.

To test the hypothesis that was developed earlier, an OLS regression is run. The regression is presented in equation 15. Based on the setup of the research, an OLS regression is the most appropriate choice for analysing the problem. CAR (0, +1) is the dependent variable and defined as the cumulative average abnormal return over the two-day event window starting at the announcement day. The independent variable of main interest is the fraction of female directors on the board of the bidder. Furthermore, the different controls are added to the regression.

(23)

Findings

Regression results

Table 13 presents the results of the different regressions that have been conducted. First the variables were tested for multicollinearity. Based on the correlation matrix (Table 12) and a maximum variance inflation factor of 2.99, which is lower than the threshold of 5 (sometimes 10), it can be concluded that there is no multicollinearity within the sample. After that a test for heteroskedasticity is conducted. Following Breusch und Pagan (1979) gives a p-value of 0.000, implying that the assumption of homoskedasticity is violated. In order to overcome this problem, the variables are normalized, and robust standard errors are used.

Table 13. Regression results

VARIABLES (1) (2) (3) (4) FFDB -0.034*** -0.014 (0.011) (0.013) BBS -0.001* -0.001 -0.001* (0.001) (0.001) (0.001) FCEO -0.006 -0.006 -0.007 (0.007) (0.006) (0.007) MC -0.003*** -0.003*** -0.003*** (0.001) (0.001) (0.001) DV 0.002* 0.002 0.002* (0.001) (0.001) (0.001) MP 0.004* 0.004* 0.004* (0.002) (0.002) (0.002) ROA 0.050*** 0.051*** 0.050*** (0.017) (0.017) (0.017) Zero female 0.006 (0.003) Threshold 0.002 (0.003) Constant 0.013*** 0.015* 0.011 0.016** (0.002) (0.008) (0.009) (0.008)

Year Dummies No Yes Yes Yes

Observations 824 798 798 798

Adjusted

R-squared 0.010 0.049 0.052 0.049

(24)

As explained in the methodology section, the most suitable approach to investigate the effect of female directors on value creation in acquisitions is an OLS regression.

Regression 1 of Table 13 only looks at the variable of interest and its effect on the cumulative abnormal return around the event date. When only looking at the fraction of female directors and ignoring all the controls, a negative effect which is significant at the 1% level is found. This means that each one percent increase in the fraction of female directors decreases the cumulative abnormal return over the day zero and day one event window by 0.034%. This is contradictory to the hypothesis that was developed based on the existing literature.

When looking at regression 2 one can see that the fraction of female directors becomes insignificant as soon as different control variables are added to the regression. Multiple control variables seem to have a significant influence on value creation. An increase in the size of the board has a negative relation with the value that is created in acquisitions. This is in line with the perception the bigger group of people to make a decision the more opinions play a role and the worse the decision gets (García-Meca, et al., 2015; Marinova, et al., 2016). Surprisingly companies with a female CEO do not generate a significantly different return than companies with a male CEO. This is not in line with the findings of Huang and Kisgen (2013) who find that acquisition announcements made by female CEOs generate an approximately 2% higher return than announcements made by male CEOs. The size of the acquiring company is negatively related to acquirer returns, meaning that bigger companies generate lower returns. One would expect bigger companies to be able to profit from synergies which would be valued by investors. However, this is not supported by the findings of this sample. Furthermore, bigger deals and paying in cash seems to increase the value that is created. Next to that, also multiple years are significantly positively related to the cumulative abnormal return.

Regression 3 replaces the fraction of female directors with a dummy, which takes on 1 for homogeneous boards (no female director) and 0 for boards with at least one female director. In line with the findings of Table 8 the dummy is insignificant, meaning that there is no relation between the value that is created and the fact that a company has homogeneous or heterogeneous board. Compared to regression 2, both the variable for bidder board size and the constant become insignificant.

(25)

Based on the F-tests that where conducted it can be said that regression 3 is preferred over the other three regressions.

It can be said that the effect of female directors on the acquirers return that is found in regression 1 is due to the different controls that have been excluded and are in added in regression 2. This means that no relation between the fraction of female directors and the value creation in acquisitions is found.

A reason for that could be the fact that companies with more female directors tend to be less active in M&A’s. The literature suggests that M&A experience increases the value that is created (Canina et al., 2019; Boateng et al., 2010). The fact that women tend to make better decisions (less overconfidence) but have less experience in M&A’s could be the reason why insignificant results are found. Both mentioned effects could balance each other out.

(26)

Conclusion

The question whether women should play an increasing role in today’s business world is around for quite a while and is still gaining a lot of attention. This is due to the perception that women tend to make better decisions than man, who are more likely to be overconfident. A big part of the literature tries to analyse the effect of women on different aspects of the business. The majority of these papers investigates whether women have an effect on the social or financial performance of the firm. There is only little literature that investigates the effect of female directors on the performance in M&As. The aim of this research is to provide an answer to this question while trying to close the gap in the literature.

In a first step, the acquisitions performance for bidders is analysed by means of an event study. During the next step, the relationship between the fraction of female directors and the value that is created in acquisitions is analysed using OLS regression. It is found that on average acquirers generate significant positive returns in deals.

Overall this study reveals that the value creation within this sample is not due to the differences in the fraction of female directors leading to the conclusion that there is no relation between these two variables. Next to that there is also no relation between value creation and the fact whether a company has a gender-homogeneous board or a gender-heterogeneous board. Lastly, there are significant differences between boards with more or less than 25% female directors. However, the value that is created is not influenced by the fact whether a board has more or less than 25% women. The findings contradict with what is expected based on the literature. Women who are less overconfident than their male counterparts are expected to make better decisions. The unexpected insignificant results could be due to the fact that an increased fraction of female directors is associated with less M&A activities, which in turn means lower acquisition announcement returns. The effect of better decision making by women (less overconfidence) and the effect that comes from less experience in M&A’s could balance each other out, leading to insignificant results. Next to that women who self-select themselves in such positions are different from women in the general population. Female directors demonstrate risk loving behaviour, possibly leading to more negative net present value projects. This could be another reason for the insignificant results.

(27)

Limitations and direction for future research

(28)

Reference list

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

Adams, R., Funk, P., 2012. Beyond the Glass Ceiling: Does Gender Mattter?. Management Science 58 (2), 219-235.

Alexandridis, G., Fuller, K. P., Terhaar, L., Travlos, N. G., 2013. Deal Size, Acquisition Premia and Shareholder Gain. Journal of Corporate Governance 20, 1-13.

Atkinson, S. M., Baird, S. B., Frye, M. B., 2003. Do female mutual fund managers manage differently?. The Journal of Financial Research 26 (1), 1-18.

Barber, B. M., Odean, T., 2001. Boys will be boys: Gender, overconfidence, and common stock investment. Quarterly Journal of Economics 116 (1), 261-292.

Bennouri, M., Chtioui, T., Nagati, H., Nekhili, M., 2018. Female board directorship and firm performance: What really matters?. Journal of Banking and Finance 88, 267-291.

Boateng, A., Du, M., Bi, X., Lodorfos, G., 2019. Cultural distance and value creation of cross-border M&A: The moderating role of acquirer charachteristics. International Review of Financial Analysis 63, 285-295.

Boehmer, E., Masumeci, J., Poulsen, A. B., 1991. Event-study methodology under conditions of event-induced variance. Journal of Financial Economics 30 (2), 253-272.

Breusch, T. S., & Pagan, A. R., 1979. A simple test for heteroscedasticity and random coefficient variation. Econometrica 47 (5), 1287-1294.

Campbell, C., Cowan, R., & Salotti, V., 2010. Multi-country event-study methods. Journal of Banking & Finance, 34 (12), 3078-3090.

Canina, L., Kim, J., Ma, Q., 2010. What We Know about M&A Success. Cornell Hospitality Quaterly 51 (1), 81-101.

Chapple, L., Humphrey, J. E., 2014. Does board gender diversity have a financial impact? Evidence using stock portfolio performance. Journal of Business Ethics 122 (4), 709-723.

(29)

Chen, J., Leung, W. S., Song, W., Goergen, M., 2019. Why female board representation matters: The role of female directors in reducing male CEO overconfidence. Journal of Empirical Finance 53, 70-90.

Corrado, C., 1989. A nonparametric test for abnormal security-price performance in event studies. Journal of financial economics 23 (2), 385-395.

Corrado, C., 2011. Event studies: A methodology review. Accounting and Finance 51 (1), 207-234.

Cowan, A., 1992. Nonparametric event study tests. Review of Quantitative Finance and Accounting 2 (4), 343-358.

Dowling, M., Aribi, Z. A., 2013. Female directors and UK company acquisitiveness. International Review of Financial Analysis 29, 79-86.

Dwyer, D. P., Gilkeson, J. H., List, J. A., 2002. Gender differences in revealed risk taking: evidence from mutual fund investors. Economics Letters 76 (2), 151-158.

Galbreath, J., 2018. Is board gender diversity linked to financial performance? The mediating mechanism of CSR. Business & Society 57 (5), 863-889.

García-Meca, E., García-Sánchez, I. M., Martínez-Ferrero, J., 2015. Board diversity and its effects on bank performance: An international analysis. Journal of Banking and Finance 53, 202–214.

Golubov, A., Xiong, N., 2020. Post-acquisition performance of private acquirers. Journal of Corporate Finance 60, 101545.

Hagendorff, J., Keasey, K., 2009. Post-merger strategy and performance: evidence from the US and European banking industries. Accounting and Finance 49 (4), 725-751.

Huang, J., Kisgen, D. J., 2013. Gender and corporate finance: Are male executives overconfident relative to female executives?. Journal of Financial Economics 108 (3), 822-839.

Horak, S., Cui, J., 2017. Financial performance and risk behaviour of gender-diversified boards in the Chinese automotive industry. Personnel Review 46 (4), 847-866.

(30)

Levi, M., Li, K., Zhang, F., 2014. Director gender and merger and acquisitions. Journal of Corporate Finance 28, 185-200.

Li, H., Chen, P., 2018. Board gender diversity and firm performance: The moderating role of firm size. Business Ethics 27 (4), 294-308.

Liu, Y., Wei, Z., Xie, F., 2014. Do women directors improve firm performance in China?. Journal of Corporate Finance 28, 169-184.

Loderer, C., Martin, K., 1990. Corporate Acquisitions by Listed Firms: The Experience of a Comprehensive Sample. Financial Management 19 (4), 17-33.

MacKinlay, A. C., 1997. Event Studies in Economics and Finance. Journal of Economic Literature 35 (1), 13-39.

Malmendier, U., Tate, G., 2008. Who makes acquisitions? CEO overconfidence and the market's reaction. Journal of Financial Economics 89 (1), 20–43.

Marinova, J., Plantenga, J., Remery, C., 2016. Gender diversity and firm performance: evidence from Dutch and Danish boardrooms. International Journal of Human Resource Management 27 (15), 1777–1790.

Martinez-Jimenez, R., Hernandez-Ortiz, M. J., Cabrera Fernandez, A. I., 2020. Gender diversity influence on board effectiveness and business performance. Corporate Governance 20 (1), 307-323.

Owen, A. L., Temesvary, J., 2018. The performance effects of gender diversity of bank boards. Journal of Banking and Finance 90, 50-63.

Palvia, A., Vähämaa, E., Vähämaa, S., 2015. Are female CEOs and chairwoman more conservative and risk averse? Evidence from the banking industry during the financial crisis. Journal of Business Ethics 131 (3), 577-594.

Pucheta-Martinez, M. C., Bel-Oms, I., Olcina-Sempere, G., 2018. Female institutional board directors on boards and firm value. Journal of Business Ethics 152 (2), 343-363.

(31)

Vafaei, A., Ahmed, K., Mather, P., 2015. Board diversity and financial performance in the top 500 Australian firms. Australian Accounting Review 25 (4), 413-427.

Wangering, D., 2019. M&A due diligence, post-acquisition performance, and financial reporting for business combinations. Contemporary Accounting Research 36 (4), 2344-2378.

(32)
(33)

Appendix B

Subsample with no female directors

Day AAR CAAR

Standardized cross-sectional t test Cowan‘s generalized

sign test Corrado rank test

-30 0.14% 0.14% 0.7044 0.98137 0.31054 -25 0.18% 0.08% 1.25527 0.98137 0.79128 -20 0.04% -0.12% 0.05772 0.51553 -0.13825 -15 0.24% 1.33% -0.2632 0.20497 -0.61552 -10 0.19% 1.10% 1.65353* 0.51553 0.4606 -9 -0.05% 1.05% -0.73933 1.13665 -0.21397 -8 -0.06% 0.99% -0.01151 0.98137 0.27928 -7 0.15% 1.14% 1.18822 1.75776** 0.68221 -6 0.09% 1.23% 0.47467 0.82609 0.0132 -5 -0.04% 1.19% -0.02875 -0.57143 -0.487 -4 0.25% 1.45% 1.26368 0.82609 0.88924 -3 -0.17% 1.28% -0.56475 -1.81366 -1.4589 -2 -0.01% 1.27% 0.44301 1.13665 0.72598 -1 0.19% 1.46% 1.18081 1.4472* 1.08931 0 1.97% 3.43% 2.67914*** 1.91304** 2.48222** 1 1.12% 4.56% 3.54865*** 3.15528*** 3.91334*** 2 0.03% 4.59% 0.26741 0.51553 -0.15075 3 0.13% 4.72% 0.55988 1.29193* 0.36056 4 0.15% 4.87% 1.83595* 2.53416*** 1.56867 5 -0.05% 4.81% -0.80829 -0.10559 -0.58287 6 -0.14% 4.67% -0.82928 -1.03727 -1.52004 7 -0.01% 4.66% 0.71421 -0.10559 -0.05975 8 -0.09% 4.56% -1.31581 -0.57143 -0.58495 9 0.03% 4.59% 0.57153 0.51553 0.46963 10 0.13% 4.72% 0.9246 0.67081 0.23551 15 0.24% 4.74% 0.42403 0.20497 0.58078 20 -0.17% 4.37% -0.30486 -0.26087 -0.74543 25 -0.24% 3.42% -0.9137 -1.03727 -1.50199 30 -0.01% 2.43% -0.6249 -0.10559 -0.42844

***- Asterisks indicate the significance level: * significant at the 10% level, ** significant at the 5% level, *** significant at the 1% level.

(34)

Appendix C

Subsample with at least one female director

Day AAR CAAR

Standardized cross-sectional t test Cowan‘s generalized

sign test Corrado rank test

-30 -0.03% -0.03% -0.40141 -0.09969 -0.82305 -25 -0.19% -0.50% -2.5966*** -1.96885 -2.31806** -20 0.07% -0.61% 0.59482 1.14642 0.60891 -15 0.07% -0.81% 1.09192 0.91277 0.59939 -10 0.06% -1.00% 1.07343 1.45794* 0.10707 -9 0.20% -0.79% 2.10071** 2.4704*** 1.87994* -8 -0.07% -0.86% 0.12309 0.44548 0.20034 -7 -0.02% -0.87% -0.63154 0.99065 -0.16881 -6 -0.09% -0.97% -1.24196 0.67913 -0.96788 -5 -0.04% -1.00% -1.20324 0.13396 -0.98792 -4 0.10% -0.90% 1.45515 0.75701 0.71401 -3 0.07% -0.83% 1.00351 2.00312** 0.96756 -2 -0.02% -0.85% -0.01467 -0.09969 -0.45685 -1 -0.02% -0.88% -1.15036 -0.09969 -0.85753 0 0.47% -0.41% 3.68458*** 2.93769*** 3.24522*** 1 0.15% -0.26% 1.15747 1.76947** 1.57515 2 -0.01% -0.27% -0.27507 0.13396 -0.40857 3 -0.07% -0.34% -0.33428 1.14642 0.08079 4 -0.04% -0.37% -0.456 0.28972 -0.35668 5 -0.03% -0.40% -0.36035 0.05607 -0.32613 6 0.01% -0.39% 0.05638 0.3676 -0.78889 7 -0.04% -0.43% 0.33335 1.53583* 0.41415 8 0.04% -0.36% 0.4028 0.99065 -0.35241 9 0.08% -0.27% 0.6078 0.13396 -0.00427 10 0.00% -0.25% 0.0111 1.06854 0.40561 15 0.01% -0.16% 0.13938 0.75701 0.10378 20 -0.08% -0.56% -0.82978 -0.09969 -0.82206 25 0.02% -0.52% 0.90822 1.61371* 0.42039 30 0.04% -0.69% 0.84548 0.91277 0.45061

***- Asterisks indicate the significance level: * significant at the 10% level, ** significant at the 5% level, *** significant at the 1% level.

(35)

Appendix D

Subsample where the fraction of female directors is below the 25% threshold

Day AAR CAAR

Standardized cross-sectional t test Cowan‘s generalized sign test Corrado rank test -30 -0.01% -0.01% -0.201530 0.335810 -0.628080 -25 -0.09% -0.36% -1.195760 -1.150070 -1.495100 -20 0.04% -0.53% 0.099560 0.632990 -0.146470 -15 0.10% -0.33% 0.598760 0.632990 0.155780 -10 0.06% -0.51% 1.010580 1.078750 -0.081610 -9 0.11% -0.40% 1.237460 1.970280** 0.712480 -8 -0.08% -0.48% -0.062780 0.707280 -0.172220 -7 0.02% -0.46% 0.017310 1.450220* 0.113270 -6 -0.09% -0.55% -1.762300* 0.261520 -1.418140 -5 -0.02% -0.57% -0.505550 -0.407130 -1.132960 -4 0.14% -0.43% 2.095370** 1.301630* 1.234430 -3 0.01% -0.41% 0.339290 1.078750 0.181530 -2 0.00% -0.42% 0.493830 0.558690 0.178430 -1 0.04% -0.38% 0.037820 0.781580 -0.019550 0 0.79% 0.41% 4.010540*** 3.307580*** 3.598410*** 1 0.39% 0.80% 2.967560*** 2.638930*** 2.618130*** 2 -0.01% 0.79% -0.406410 -0.258540 -0.741650 3 -0.03% 0.76% 0.028120 1.895990** 0.496190 4 0.01% 0.77% 0.789800 1.227340 0.525050 5 -0.03% 0.74% -0.481180 -0.109960 -0.546780 6 -0.02% 0.72% -0.008320 0.038630 -1.141960 7 -0.03% 0.69% 0.620910 1.227340 0.281460 8 0.01% 0.73% -0.474890 0.632990 -0.554530 9 0.10% 0.83% 1.117820 0.707280 0.601080 10 0.03% 0.86% 0.333990 1.078750 0.389140 15 0.07% 0.96% 0.535770 0.855870 0.527540 20 -0.12% 0.57% -1.239650 -0.184250 -1.217370 25 -0.03% 0.40% 0.517340 1.301630* 0.201670 30 0.00% 0.03% -0.083660 -0.332840 -0.626770

***- Asterisks indicate the significance level: * significant at the 10% level, ** significant at the 5% level, *** significant at the 1% level.

(36)

Appendix E

Subsample where the fraction of female directors is above the 25% threshold

Day AAR CAAR

Standardized cross-sectional t test Cowan‘s generalized sign test Corrado rank test -30 0.09% 0.09% 0.28846 -0.09761 -0.03125 -25 -0.28% -0.55% -1.65577* -0.89442 -0.98304 -20 0.22% -0.40% 1.40651 1.69522** 1.89912* -15 0.11% -0.80% 1.22129 0.6992 0.39018 -10 0.31% -1.04% 1.90522* 1.29681* 1.11786 -9 0.44% -0.60% 2.87387*** 2.29283** 2.78573*** -8 0.05% -0.55% 0.40879 0.3008 1.39911 -7 0.03% -0.52% -0.14438 0.6992 0.09196 -6 0.18% -0.34% 1.2887 1.89442** 1.46608 -5 -0.18% -0.52% -1.21775 0.5 -0.05179 -4 0.05% -0.47% -0.1513 -0.69522 -0.46786 -3 0.09% -0.38% 1.1865 -0.29681 0.23304 -2 -0.13% -0.51% -0.83881 -0.49602 -0.82233 -1 -0.12% -0.63% -1.21312 -0.69522 -0.875 0 0.68% 0.05% 2.30846** 0.89841 1.65894* 1 -0.02% 0.03% 0.42752 1.29681* 1.77858* 2 0.08% 0.11% 1.02428 1.49602* 0.82947 3 -0.05% 0.06% -0.21162 -0.69522 -0.74465 4 -0.06% 0.00% -0.59543 0.5 -0.46429 5 -0.05% -0.04% -0.42164 0.10159 -0.0625 6 -0.05% -0.10% -0.72519 -0.69522 -0.8125 7 -0.02% -0.12% 0.037 0.3008 0.23929 8 0.04% -0.08% 0.58152 -0.09761 -0.11429 9 -0.13% -0.21% -0.80499 -1.09363 -1.13751 10 0.00% -0.09% 0.33272 0.5 0.28572 15 -0.06% -0.12% -0.63277 -0.29681 -0.48929 20 0.09% -0.58% 0.73235 -0.29681 0.30982 25 -0.11% -0.68% -0.56746 -0.89442 -1.35536 30 0.26% -0.78% 2.48095** 2.89044*** 2.47858**

***- Asterisks indicate the significance level: * significant at the 10% level, ** significant at the 5% level, *** significant at the 1% level.

(37)

Appendix F

Subsample where suspicious events are excluded (another event in the event window)

Day AAR CAAR

Standardized cross-sectional t test Cowan‘s generalized

sign test Corrado rank test

-30 0.01% 0.01% -0.00123 0.47899 -0.42386 -25 -0.12% -0.38% -1.71403* -1.34174 -1.64698* -20 0.06% -0.52% 0.51622 1.31933* 0.49883 -15 0.10% -0.38% 0.83471 0.82913 0.18488 -10 0.09% -0.56% 1.68523* 1.52941* 0.25505 -9 0.15% -0.40% 1.75214* 2.57983*** 1.47253 -8 -0.08% -0.48% -0.12051 0.7591 0.18459 -7 0.01% -0.47% -0.22207 1.66947** 0.10315 -6 -0.06% -0.53% -1.04702 0.82913 -0.90973 -5 -0.04% -0.57% -1.19539 -0.22129 -1.13519 -4 0.12% -0.46% 1.65827* 0.89916 0.81222 -3 0.03% -0.43% 0.84191 1.03922 0.29338 -2 -0.02% -0.45% 0.12558 0.26891 -0.22292 -1 0.01% -0.44% -0.53724 0.40896 -0.42245 0 0.79% 0.35% 4.60985*** 3.63025*** 3.89002*** 1 0.35% 0.71% 2.90513*** 2.85994*** 2.99748*** 2 0.00% 0.70% -0.19064 0.19888 -0.53152 3 -0.04% 0.67% -0.11347 1.45938* 0.1271 4 0.00% 0.67% 0.56485 1.38936* 0.29704 5 -0.03% 0.63% -0.61001 -0.0112 -0.54054 6 -0.02% 0.61% -0.25424 -0.15126 -1.29949 7 -0.03% 0.58% 0.49851 1.2493 0.25082 8 0.01% 0.62% -0.24888 0.54902 -0.5628 9 0.08% 0.70% 0.86768 0.47899 0.25054 10 0.03% 0.74% 0.5299 1.31933* 0.52898 15 0.07% 0.82% 0.44605 0.89916 0.40244 20 -0.10% 0.43% -0.99257 -0.15126 -1.04162 25 -0.03% 0.27% 0.36561 0.96919 -0.1947 30 0.04% -0.07% 0.65273 1.10924 0.41253

***- Asterisks indicate the significance level: * significant at the 10% level, ** significant at the 5% level, *** significant at the 1% level.

Referenties

GERELATEERDE DOCUMENTEN

vier faktore naam= Iik: (i) ryping: die ryping van fisiese strukture open die moontlikhede tot denkhandelinge; (ii) ervaring, fi= sies sowel as logies-matematies~

tyre treads12, 13, like adding strength at low strains by the use of short fibres.5 Despite the fact that short-cut fibres, ranging in lengths of a few millimetres, can

To furnish insight to the reader and underpin the framework developed in this paper that will be utilized in the future, we begin by brie fly discussing select multiwavelength

A 4-element LO-phase shifting phased-array system with 8-phase passive mixers terminated by baseband capacitors is realized in 65nm CMOS.. The passive mixers upconvert

This paper examined the relation between an auditee’s risk profile, listing status and gender diversity in the boardroom on the level of audit fees and the quality of audit

The data concerning directors’ and CEOs’ skills, CEO power, board size, gender diversity, and, for some companies, other variables was manually collected from the

By performing both, cross sectional and panel data models, this study indicates that although female directors are more highly represented among less valuable firms,

Reminding, that age enhances strategic change, innovation such as R&amp;D investments is of high strategic importance and that R&amp;D is a risky investment, directors need