• No results found

Author: Tom Bleijenberg Supervisor: Dr. Swarnodeep Homroy Thesis Msc Finance

N/A
N/A
Protected

Academic year: 2021

Share "Author: Tom Bleijenberg Supervisor: Dr. Swarnodeep Homroy Thesis Msc Finance"

Copied!
43
0
0

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

Hele tekst

(1)

INVESTOR REACTION TO A CEO JOINING AN

OUTSIDE BOARD – AN EVENT STUDY

Author: Tom Bleijenberg

Supervisor: Dr. Swarnodeep Homroy

Thesis Msc Finance

S2353687

(2)

2 Abstract

(3)

3

1. INTRODUCTION

When appointing a new director, many firms choose an independent director outside of the firm. An independent director is assumed to be unbiased regarding the financial decisions and termination of managers. This has led to new regulations for the NYSE and NASDAQ. More specifically, firms are required to have a majority of independent directors on their board.1

Furthermore, the appointment of an independent director increases firm value, regardless of how many independent directors are already on the board (Rosenstein and Wyatt, 1990). Many of these independent directors are CEO of another, unrelated firm.

According to the Spencer Stuart Board Index, CEO and female directors are the two types of independent directors in highest demand.2 Previous research has argued that when a CEO with a

good reputation joins the board of a firm, this is generally perceived as a good signal (see e.g., Dahya and McConnell, 2005; Fahlenbrach, Low and Stulz, 2010). Specifically, the stock market reacts more favourably to the appointment of a CEO as outside director compared to the appointment of a non-CEO director (Fahlenbrach, Low and Stulz, 2010). They theorize that the positive price reaction is due to CEOs having a good reputation, signalling that the appointing firm is going in the right direction. They argue that a CEO with a good reputation will only choose firms that are on the right track in the eyes of the CEO. These arguments are all based on how a shareholder will perceive this news. However, the CEO’s firm also has shareholders. This raises the question, do these effects also apply on the opposite side? For example, when the CEO of firm A gets appointed on the board of firm B, do the shareholders of firm A perceive this as a positive or negative signal?

Various factors have to be taken into account in answering this question. Among firms and industries there are vast differences which all potentially influence the shareholders’ reaction to the CEO joining an outside board. These differences lead to two follow-up research questions: Are

(4)

4 some firms more sensitive to a CEO joining an outside board? Also, what firm/CEO specific characteristics influence the CEO’s decision whether to accept an outside position or not?

It is highly relevant for a CEO to know what the shareholder reaction to joining an outside board is. When the stock performance goes down, CEOs will be more likely to lose their job (Warner, Watts and Wruck, 1998). Therefore a CEO has to take into consideration how his/her shareholders perceive such a move. There are several reasons to think of why it would be a good signal that the CEO of your firm joins the board of another. One could argue that if operations at the firm of the CEO are going smoothly, there is time available to engage in other activities outside the firm. If the firm of the CEO is in distress, the CEO most likely would not engage in activities outside the firm. Furthermore, activities such as network building and technological spill overs may even increase the home firm value (Fama, 1980; Fama and Jensen, 1983). However, a shareholder may also perceive this action as negative. The reputation building may lead focus away from the main job, which is leading the company of which the person is appointed CEO.

When searching for reasons why a CEO joins an outside board, various factors have to be taken into account. Across firms and industries, there are large differences. For firms where the CEO’s time is highly valuable, it may not even be possible for top-level executives to sit on outside boards, and for others it may even be the standard. Furthermore, personal characteristics differ among executives too. An older, more experienced CEO likely has a better reputation than a young inexperienced one and is therefore in higher demand for outside board positions (Weisbach, 1988) These factors all potentially influence whether a CEO joins an outside board.

(5)

5 I will do so by performing an event study, using the method provided by MacKinlay (1997), Fama and French (1993) and Carhart (1997). As Fahlenbrach, Low and Stulz (2010) did, I will perform regression analysis and check for firm performance. However, whereas they checked which firm characteristics determined whether a CEO would join an appointing board, I will check the characteristics of the CEO and his/her own firm. A regression analysis will be performed on all the cumulative abnormal returns (CARs) yielded by the event study to identify which factors influence the shareholder reaction of firms. Furthermore a large, separate sample will be constructed to perform a regression analysis with the goal of identifying factors which influence the number of outside directorships a CEO holds.

The results of this research show no initial investor reaction to a CEO joining another board for the first time since assuming his/her role as CEO. However, subsample analysis show that for small firms, firms with high growth and firms with a female CEO, there are significant negative returns. The results regarding small and high growth firms were to be expected (Booth and Deli, 1996). Through the separate regression analysis it is proven that firm size, CEO duality, and CEO age have a positive influence on the number of outside directorships a CEO holds. It is also proven that female CEOs hold more outside directorships than male CEOs.

This research starts with a review of existing literature in section 2. In this section existing literature is presented and hypotheses are formed based on the literature. Section 3 provides a step-by-step explanation on how the samples for this research are formed. In section 4 the methodology of the event study and regression analysis is explained. Section 5 provides an overview of the results, in section 6 the results are discussed and linked to the previous research and in section 7 a conclusion is drawn including some recommendations for further research.

2. LITERATURE REVIEW

(6)

6 appointing firm, the third subsection offers the perspective of the CEO accepting the position and the fourth subsection offers the perspective of the home firm of the CEO. The literature review concludes with hypotheses based on the literature presented.

2.1 Agency theory and the role of the board

One of the most well-known theories regarding the relationship between the directors and executives is agency theory. Agency theory states that managers will always act in their self-interest regardless of the shareholder value created or destroyed. The main task of the board of directors is to minimize agency costs by monitoring the management. These boards have the power to fire, hire and compensate top-level management (Fama and Jensen, 1983). This gives directors the responsibility to monitor and shape the management, as well as design compensation schemes which minimize agency costs (Jensen and Meckling, 1976). However, CEOs are often also chairperson of the board, meaning they essentially monitor themselves. This gives CEOs a tremendous amount of power within the company (Jensen, 1993).

(7)

7 Hermalin and Weisbach (1998) confirm this view. Their model implicates that the proportion of independent directors negatively influences the CEO’s power. Therefore, a large percentage of outside directors limits the CEO’s power, making it harder for the CEO to work in his/her self-interest as opposed to the shareholders’ self-interest. In monitoring the board, independent directors are more efficient than inside directors. Weisbach (1988) also found that when a CEO performs poorly, a board dominated by outside directors is more likely to force the CEO’s resignation.

2.2 Why appoint outside directors?

In search for new board members, CEOs are in high demand. A CEO is associated with knowledge and experience and therefore can have a leading role in monitoring the management of a firm. (Adams and Ferreira, 2007). They argue that a CEO brings an unusual amount of authority and knowledge. Therefore, appointing a CEO as outside director will give the board more power in standing up to the CEO of the company. Shivdasani and Yermack (1999) agree with this. However, they argue that some firms benefit more than others from this. Firms with misaligned managerial incentives will benefit most from a CEO director, since the high level of expertise and authority will have the most impact on such a firm.

Shareholders seem to agree that CEOs are good directors. Appointing a CEO as independent director is perceived as a positive signal by shareholders. Fehlenbrach, Low and Stulz (2010) proved this by showing the stock market reacts more positively to the appointment of a CEO independent director compared to a non-CEO independent director appointment. However, firm performance was unaffected by the appointment of a CEO. Therefore, the main driver behind this seems to be the good reputation a CEO brings to the table.

(8)

8 bring a large network to the table which could benefit the firm greatly in establishing itself among competitors.

Among investors, there seems to be a belief that outside directors bring more knowledge to the table. This also applies to the CEO. Dahya and McConnell (2005) found that the announcement of an outside CEO appointment yields higher stock returns compared to the announcement of an inside CEO. Boards with more outside directors were also found to be more likely to appoint outside CEOs. This view is also confirmed by Fama and Jensen (1983).

The field of expertise of a potential outside director is of importance for the appointing firm. When a certain division of the company is not functioning well, appointing an expert in this field as director might solve this problem. Booth and Deli (1999) find that appointing bankers as directors positively influence bank debt, implicating that independent directors can directly influence a firm’s financial policy. Other papers support these findings (see e.g. Güner, Malmendier and Tate 2008; Linck, Netter and Yang, 2009).

Apart from the reputational and expertise argument, shareholders can also expect an increased firm value during tender offers when appointing outside directors. A study by Cotter and Shivdasani (1997) found that during tender offers, more independent directors will increase shareholder returns. Even when the offer is ultimately resisted, the shareholder gains are higher when a large fraction of the board consists of independent directors.

(9)

9

2.3 Director incentives

One of the most obvious incentives to join another board is monetary gain. Jensen and Murphy (1990) report that there is a lack of strong-pay-for-performance incentives for CEOs. If the CEOs are do not feel incentivized to increase the performance of their firm, they may seek additional opportunities for monetary gain. When planning their retirement, the monetary gain is an important factor in accepting an outside position (Brickley, Coles, Linck and 1999). However, reputational incentives may even outweigh the monetary incentives.

The appointment of a CEO as independent director is not only good news for the appointing firm. CEOs that are perceived as better managers tend to become outside directors (Kaplan and Reishus, 1990). Their results suggest that outside directors are hired because they ‘look good’ based on the performance of their own firms. CEOs of large successful firms will have This works both ways, not only does the firm look good for appointing a good manager, but the manager looks good because the managers’ qualities are recognised. Therefore this provides a reputational incentive for a CEO to join another board.

The importance of reputation is further emphasized by Masulis and Mobbs (2014). They find that directors who hold multiple directorships exert more effort where there is more reputational benefit. The directors attend more board meetings of their more prestigious directorships and are willing to spend more time in committees of their relatively more prestigious directorships. The implications of their research are that reputational benefits are highly regarded by directorships. Because the director wants to look good, firms that do not perform or are involved in scandals, will be avoided because of the negative reputational consequences of sitting on the board of such a firm (Fich and Shivdasani, 2007). This shows what the directors perceive to be most beneficial to them, namely, their reputation.

(10)

10 not negatively influence the directors’ ability to do his/her job. In fact, a positive relationship between number of directorships held and director quality is reported. Brickley, Coles and Linck (1999) and Kaplan and Reishus (1990) found the same positive relationship. For a director it is therefore beneficial to accept multiple directorships since it will signal the director is good at its job and will therefore be beneficial to the directors’ reputation.

2.4 Perspective of home firm CEO

The time of a CEO is valuable, and allocating your time to another firm may not fare well with shareholders. Perry and Peyer (2005) found evidence that shareholders react negatively to a CEO joining an additional outside board when already holding prior outside directorships. This emphasizes the fact that a CEO’s time is valuable. The shareholders believe that allocating time between the main job as CEO and multiple directorships will eventually lead to lower performance of the CEO and therefore react negative to the news.

The nature of the home firm also strongly influences the number of outside directorships held by CEOs. Booth and Deli (1996) found that firms where a large amount of the value is made up of growth opportunities and where the marginal value of the CEO’s time is high will not favour a CEO holding outside directorships. The CEO will need to spend his time pursuing growth opportunities instead of sitting on an outside board.

A theoretical model developed by Conyon and Read (2006) finds that CEOs will decrease their home firm value by accepting outside directorships. They argue that time spent at another firm is time not spend on value creating activities at the home firm. According to the model, directors accept more directorships than optimal and will therefore spend too much time outside of their home firm, essentially wasting valuable time. The benefits of accepting an outside position are often met with costs similar or higher according to the model.

(11)

11 firm can benefit of this (Fama, 1980; Fama and Jensen, 1983). Furthermore, Booth and Deli (1996) suggest that the home firm can also profit of this outside directorships through the CEO building relationships and knowledge. For example different management styles and strategies. Especially for younger CEOs, this can be highly beneficial since they can learn from more experienced CEOs. This can potentially increase the CEO’s home firm’s value.

2.5 Hypotheses

The existing literature provides various benefits and costs to accepting outside positions. However, how shareholders perceive these benefits and costs are most important. Based on the literature, I expect the shareholder reaction to be slightly negative to a CEO joining another board as independent director. This effect is expected to be more severe for smaller firms with relatively high growth rates. Furthermore, large firms are expected to have CEOs with more outside directorships compared to smaller firms. This is mainly because firms are expected to have directors with larger networks, more experience and less growth opportunities. It is also expected that CEOs with more power (CEO duality – also being chairperson) are expected to hold more outside directorships. The research questions and hypotheses are formatted as follows:

RQ1: ‘How do the shareholders of firm A react when their CEO gets appointed to the board of firm B?’

H1. The shareholders perceive this as a negative signal, therefore the stock abnormal returns during the event window will be negative.

RQ2: ‘What firms are more prone to a heavy shareholder reaction compared to other firms?’

H2a. Small, firms will have more severe negative abnormal returns.

(12)

12

RQ3: ‘What type of firms/CEO’s usually hold one or more outside directorships?’.

H3a. Firm size has a positive influence on the number of boards a CEO is sitting on. H3b. CEO age will have a positive influence on the number of boards sitting on. H3c. CEO duality will have a positive influence on the number of boards sitting on.

3.

METHODOLOGY

3.1 Event study method

The first part of this research will consider the shareholder reaction to an outside appointment of their CEO. The shareholder reaction will be measured by calculating abnormal returns. The abnormal returns are calculated using two different models to check whether one model yields different results. The first model used is the market model explained by MacKinlay (1997), the second model is the Fama French 4 factor model, based on Fama and French (1991) and Carhart (1997). The significance will be tested using a t-test introduced by Boehmer, Musumeci and Poulsen (1991)

3.1.1 Calculating abnormal returns

(13)

13 windows will be 11 [-5, 5] and 21 [-10, 10] days. The event and estimation window are illustrated below.

Estimation window Event window

|---|---|--0--|

𝜏0− 281 𝜏0− 61 𝜏1 𝜏0 𝜏2

Where 𝜏0it the announcement date, 𝜏1 is the first day of the event window and 𝜏2 the final day of the event window. The short event window is modelled after Fehlenbrach, Low and Stulz (2010) and in line with Chopra, Lakonishok and Ritter (1992), which reason that short event windows are optimal. In this study the expectation is a slightly negative reaction centred around the announcement date, therefore a short event window is the most optimal. Furthermore, all director appointments are communicated through an announcement, increasing the likeliness of a possible investor reaction centred around the announcement date.

In this research, the standard market model will be used along with the Fama French 4 Factor (FF4F) model. The market model is a simple model and the most used event study method. According to Brown and Warner (1980, 1985), the market model does not do a worse job than more complicated models and therefore this model will be the first choice model used in this study. However, to check whether there are more complicated effects applicable for this study, the FF4F model is applied to capture a different risk calculation. The market model theory is mainly based on the MacKinlay (1997) paper and the Fama French model is based on Fama and French (1993) and Carhart (1997). The standard equation of the market model by MacKinlay (1997) is illustrated below:

𝑅𝑖𝜏 = 𝛼𝑖 + 𝛽𝑖𝑅𝑚𝜏+ 𝜀𝑖𝜏 (1)

Where 𝑅𝑖𝜏 is the actual return of firm 𝑖 at period 𝜏, 𝛼𝑖 is the intercept, 𝛽𝑖 is the estimate of the

(14)

14 The FFM model is based on the Fama French three factor model, introduced in Fama and French (1993). The three factor model was extended by Carhart (1997) to additionally contain a momentum factor:

𝑅𝑖𝜏 = 𝛼𝑖+ 𝛽𝑖,𝑚𝑅𝑚𝜏 + 𝛽𝑖,𝑆𝑀𝐵𝑆𝑀𝐵𝜏 + 𝛽𝑖,𝐻𝑀𝐿𝐻𝑀𝐿𝜏 + 𝛽𝑖,𝑀𝑂𝑀𝑀𝑂𝑀𝜏+ 𝜀𝑖𝜏 (2)

Where 𝑆𝑀𝐵𝜏 stands for ‘small minus big’, 𝐻𝑀𝐿𝜏 stands for ‘high minus low’ and 𝑀𝑂𝑀𝜏 is a factor meant to capture the excess return of past winning over past losing stocks Carhart (1997). For a detailed description of the factors refer to Fama and French (1991) and Carhart (1997).

Daily returns are calculated by dividing the price of firm 𝑖 at period 𝜏 by the price at 𝜏 − 1 and log the returns. This yields the following equation:

𝑅𝑖𝜏 = 𝐿𝑜𝑔 ( 𝑃𝑖𝜏

𝑃𝑖𝜏−1) (3)

Where 𝑅𝑖𝜏 is the continuously compounded return of firm 𝑖 at period 𝜏, 𝑃𝑖𝜏 is the price of firm 𝑖 at period 𝜏 and 𝑃𝑖𝜏−1 is the price of the same firm one day prior.

Based on the models introduced in equation (1) and (2), the abnormal returns can be calculated as follows:

𝐴𝑅𝑖𝜏 = 𝑅𝑖𝜏− (𝛼𝑖 + 𝛽𝑖𝑅𝑚𝜏) (4)

𝐴𝑅𝑖𝜏 = 𝑅𝑖𝜏− (𝛼𝑖 + 𝛽𝑖,𝑚𝑅𝑚𝜏 + 𝛽𝑖,𝑆𝑀𝐵𝑆𝑀𝐵 + 𝛽𝑖,𝐻𝑀𝐿𝐻𝑀𝐿 + 𝛽𝑖,𝑀𝑂𝑀𝑀𝑂𝑀𝜏) (5)

(15)

15 applied to both models. The abnormal returns during the event window can be aggregated to form the cumulative abnormal returns (CAR) for each event:

𝐶𝐴𝑅𝑖(𝜏1, 𝜏2) ∑ 𝐴𝑅𝑖𝜏

𝜏2

𝜏=𝜏1

(6)

Where 𝐶𝐴𝑅𝐼(𝜏1, 𝜏2) is the cumulative abnormal return for firm 𝑖 during the event window (𝜏1, 𝜏2). The abnormal return and cumulative abnormal return can be aggregated to form the average abnormal return (AAR). The AARs can be cumulated during the event window to derive the cumulative abnormal returns (CAARs).

3.1.2 Statistical significance test

To test for statistical significance, the standardized cross-sectional t-test is used. This test is introduced by Boehmer, Musumeci and Poulsen (1991). This test is based on the ordinary t-test but addresses the misspecification problem for the ordinary cross-sectional technique (Boehmer, Musumeci and Poulsen, 1991). The abnormal returns are standardized and put in the following formula: 𝑡𝑠𝑐𝑠 = 1 𝑁∑𝑁𝑖=1𝑆𝐴𝑅𝑖𝜏 √ 1 𝑁(𝑁 − 1)∑ [𝑆𝐴𝑅𝑖−𝑁1∑𝑁𝑖=1𝑆𝐴𝑅𝑖] 2 𝑁 𝑖=1 (7)

Where 𝑁 stands for the number of events and 𝑆𝐴𝑅𝑖𝜏 is firm 𝑖’s standardized residual on day 𝜏. For

a more detailed description refer to Boehmer, Musumeci and Poulsen (1991). 3.1.3 Regression analysis of CAR’s

(16)

16 There are various possible explanations for differences between events. To determine which factors influence the investor reaction, a regression is constructed. Since the sample only includes US firms, there are no country differences. To check for cross-sectional heterogeneity, robust standard errors are used. The regression is displayed below:

𝐶𝐴𝑅𝑖[−1;1] = 𝛼 + 𝛽1𝑅𝑂𝐴𝑖 + 𝛽2𝑆𝑖 + 𝛽3𝐶𝐷𝑖 + 𝛽4𝐶𝐴𝑖+ 𝛽5𝐶𝐺𝑖 + 𝛽6𝐵𝐺𝑅𝑖+ 𝛽7𝐴𝐺𝑖 + 𝜀𝑖 (8)

Where 𝐶𝐴𝑅𝑖[−1;1] is the cumulative abnormal return of firm 𝑖 in the event period [-1;1], 𝛼 is the intercept, 𝑅𝑂𝐴 is return on assets, 𝑆 is firm size, 𝐶𝐷 is CEO duality, 𝐶𝐴 is the age of firm 𝑖’s CEO, 𝐶𝐺 is the gender of the firms’ CEO, 𝐵𝐺𝑅 is the board gender ratio, 𝐴𝐺 is the asset growth of the firm in the event year and 𝜀 is the error term.

This regression signals if and which factors possibly influence the investor reaction to the news of their CEO joining another board. This is an important part of the research because it offers a possible explanation for differences between event results.

3.2 Regression analysis on outside positions

To find out what factors influence a CEO to join another board, a regression analysis is performed. This short analysis will contain two almost identical regressions. The first regression will be based on the number of board positions. The second regression is based on a dummy variable which takes 1 if the CEO holds an outside position and 0 if not. The sample constructed includes yearly data for individual firms, therefore the regression analysis has to be performed using panel data appropriate tests The data is collected across time and firms and is therefore characterized as panel data. Firms are fixed across time and therefore no random effects are applicable. The regression used will be a fixed effects panel data regression. The regression is constructed as follows:

𝑁𝑂𝑃𝑖 = 𝛼𝑖 + 𝛽1𝑅𝑂𝐴𝑖𝑡+ 𝛽2S𝑖𝑡+ 𝛽3 𝐵𝐺𝑅𝑖𝑡+ 𝛽4𝐵𝑁𝑀𝑖𝑡+ 𝛽5𝐵𝑆𝑖𝑡+ 𝛽6𝐶𝐷𝑖𝑡+ 𝛽7𝐶𝐺𝑖𝑡+ 𝛽8𝐶𝐴𝑖𝑡

(17)

17 𝑂𝑃𝑖 = 𝛼𝑖 + 𝛽1𝑅𝑂𝐴𝑖𝑡+ 𝛽2S𝑖𝑡+ 𝛽3 𝐵𝐺𝑅𝑖𝑡+ 𝛽4𝐵𝑁𝑀𝑖𝑡+ 𝛽5𝐶𝐷𝑖𝑡+ 𝛽6𝐶𝐺𝑖𝑡+ 𝛽7𝐶𝐴𝑖𝑡

+ 𝜀𝑖 (10)

Where 𝑁𝑂𝑃 is number of board positions for firm 𝑖’s CEO, 𝑂𝑃 is a dummy which takes 1 if the CEO holds an outside position and 0 if not, 𝑅𝑂𝐴 is return on assets, 𝑆 is firm size, 𝐵𝐺𝑅 is board gender ratio, 𝐵𝑁𝑀 is board nationality mix, 𝐶𝐷 is CEO duality, 𝐶𝐺 is CEO gender, 𝐶𝐴 is CEO age, and 𝜀 is the error term. Like the regression in equation 8, there are no country differences.

Table 1:

Definitions and sources of all variables used

Variable Description Database

Daily return data Daily price of security CRSP

Return on Assets Net Income / Total Assets Compustat

Size Ln (Total Assets) Compustat

Asset Growth in year t (Total Assets)/(Total Assets t-1) Compustat

CEO Age CEO age per year BoardEx

CEO Gender Dummy variable (1 = male, 0 = female) BoardEx

CEO Duality Dummy variable (1 = CEO is also chairperson, 0 = not) BoardEx

Board Nationality Mix No. of different nationalities / 1 BoardEx

Board Gender Ratio Number of males on board / board size BoardEx

4 DATA

4.1 Databases

(18)

18

4.2 Event study data

The event study sample is constructed by using the announcement list provided by BoardEx. The list starts in 2003 and contains 205,832 board mutation announcements. Because a large sample is preferred, the period over which events are be collected is 2003 until 2018. The events list is created by filtering the announcements to only show independent director appointments. This leaves a list of 20,340 announcements. To perform an event study, only firms with daily return data can be used. Therefore, I use board data for SP1500 firms between 2003 and 2018 to determine whether an independent director was CEO of an SP1500 firm at the day of announcement or not. This is done by matching director ID’s of independent director announcements with director ID’s of the complete list of SP1500 CEO’s between 2003 and 2018. Subsequently, dates are checked whether the announcement coincided with the directors’ employment as CEO.

(19)

19

Figure 1:

Distribution of events per year. For comparison a distribution of total independent director announcements by BoardEx is included in the appendix.

Daily return data is provided by the CRSP database, the company tickers associated with the company of the CEO are linked to the CRSP database which provides daily return data for securities.

After the event list is completed, several variables are linked from Compustat and BoardEx to each individual event. The company ID’s and tickers are matched across the different databases. The CEO age, CEO gender, Board Gender Ratio, and CEO duality are retrieved from BoardEx and linked to the company IDs. Financial data is retrieved from Compustat and for each event the return on assets is calculated in the event year, and asset growth for each event year is calculated. The financial data provided by Compustat is also linked to the company IDs. The descriptive statistics are shown below in table 2:

12 12 24 36 17 29 33 19 10 9 6 8 2 0 0 5 2 0 0 6 2 0 0 7 2 0 0 8 2 0 0 9 2 0 1 0 2 0 1 1 2 0 1 2 2 0 1 3 2 0 1 4 2 0 1 5 2 0 1 6 N U MBE R O F E V EN T S YEAR

(20)

20

Table 2:

Summary statistics of variables used in CAR regression

Variable Obs Mean Std. Dev. Min Max

Total Assets 215 40624.58 182885.9 78.537 2187631

Asset Growth (Event Year) 215 0.11041 0.4182499 -0.441639 5.224293

ROA 215 0.046313 0.0746671 -0.441195 0.2250363

Board Gender Ratio 211 0.856005 0.0921953 0.571 1

CEO Gender 215 0.934884 0.2473067 0 1

CEO Age 215 54.14884 6.002818 34 71

CEO Duality 215 0.404651 0.4919699 0 1

These variables will be included as independent variables in the regression analysis of the CARs for both the market model and the Fama French 4 factor model.

4.3 Regression analysis data

The second sample is created to determine what drives a CEO to accept outside positions. It is constructed by linking a sample of all SP1500 firms between 2003 and 2017 to their board and financial data by using company IDs. First, all incomplete data is removed from the sample. After this, all board data excluding CEO data is removed. This leaves a sample of 11,492 observations consisting of 1,490 unique firms. Included in the sample are all variables used in the regression analysis. The board data consists of the board gender ratio, board nationality mix, CEO gender, CEO age, the total number of directorships held by the CEO and a dummy variable which indicates whether a CEO holds more than one directorship. The financial data included total assets, asset growth, and return on assets.

A table displaying the yearly means of the sample shows several trends such as an increase in female representation in the directorship labour market, a decrease in CEO duality and an increase in the average age of CEOs. The table can be found in the appendix alongside the descriptive statistics of the sample.

(21)

21

5.1 Event study results

The event study results are divided into three separate parts. The first part is dedicated to the main results, the second part is dedicated to the CAR regression results, the third part consists of the results of subsample analysis and the final part is presents the results of the panel data regression.

5.1.1 Main results

The main results are shown below in table 3. Both the market model and the Fama French 4 factor model are displayed. The CAARs for three separate event windows are displayed for both models:

Table 3: Main results

This table displays the full sample market-model and Fama French 4 factor model results. Only CEOs leading an SP1500 firm joining an outside board for the first time since becoming CEO are included.

Full sample

Market Model Fama French 4 Factor Model Event Window N CAAR T-value p-value CAAR T-value p-value

[-1, 1] 215 -0.045% -0.160 0.873 -0.005% 0.093 0.926

[-5, 5] 215 0.417% 0.771 0.442 0.454% 0.798 0.426

(22)

22 Table 3 shows the main results of the regression. In the 3 day event window [-1, 1] the returns are negative for both models. This is in line with the first hypothesis. However, the p-value for both models is very large and well above 0.10, meaning that the CAARs cannot be proven significant. For the larger event windows, the CAARs are positive. However, as is the case with the 3 day event window, both the longer event windows are insignificant and therefore it cannot be proven that the CAARs are different from zero.

Figure 2 displays the CAAR development during the event window. It shows that the day after the announcement, the CAAR decreases for both models. For comparison, the CAAR development during the event window of 11 and 21 days can be found in the appendix.

5.1.2 CAR regression results -0.04% 0.00% 0.04% 0.08% 0.12% 0.16% -1 0 1 C A A R

Days relative to event

CAAR in event window [-1, 1]

MM CAAR FF CAAR

Figure 2:

(23)

23 After calculating all CARs and CAARs for both model, the regression analysis shown in equation 8 is performed to check if and which factors influence the CARs of firms. The results are shown in table 4 below:

Table 4: Full event study sample regression results

The following table displays the results of a robust standard error regression on the CARs of the 3 day event period [-1, 1] for both models.

Market Model Fama French 4 Factor Model

Variables Coef. T-value p-value Coef. T-value p-value

Ln(Total Assets) 0.0000503 0.04 0.967 -0.00025 -0.18 0.861

Asset Growth (Event Year) -0.0064463 -1.62 0.108 -0.00503 -1.23 0.218

ROA 0.0500234 0.81 0.42 0.064976 1.15 0.254

Board Gender Ratio -0.0156773 -0.66 0.511 -0.00888 -0.36 0.717

CEO Gender 0.0127953 1.49 0.138 0.006214 0.65 0.516 CEO Age 0.0005658 1.19 0.236 0.000706 1.49 0.138 CEO Duality 0.0012074 0.2 0.842 0.000362 0.06 0.953 Constant -0.031947 -1.17 0.244 -0.03695 -1.26 0.208 F Value (Prob > F) 2.1 (0.0447) 1.59 (0.1394) R-Squared 0.0347 0.04023

For both models, total assets, asset growth, and gender ratio have negative coefficients. However, these coefficients are very small are not proven significant. The same is true for the positive coefficients. Furthermore, the R-squared for both models is very small, meaning that the variables explain only a small percentage of the effect on CARs for each firm. However, the regression of the market model is not completely unfitting. The Prob(F) value of 0.0447 suggests that the regression equation does have some validity in fitting the data. Therefore, the independent variables are not purely random with respect to the dependent variable. For the FF4F model, the Prob(F) value exceeds 0.1 and therefore it cannot be disproved that all independent coefficients are zero.

(24)

24

5.1.3 Subsample analysis

To check for in-sample differences, three small, separate event studies have been performed. The three groups formed include a group consisting of firms which have all grown 20% or more during the event year, a group which includes only the events with female CEOs and a group with the 24 smallest firms. The results of these tests are shown in table 4.

Table 5: Subsample analysis results

This table displays the results of three subsamples. Both the market-model and Fama French 4 factor model CAARs and corresponding p-values are shown.

Female CEOs

Market Model Fama French 4 Factor Model Event Window N CAAR T-value p-value CAAR T-value p-value

[-1, 1] 14 -1.48%** -2.007** 0.046** -0.93% -1.223 0.223

[-5, 5] 14 -1.49% -1.041 0.299 -1.53% -1.041 0.299

[-10, 10] 14 2.05% 0.833 0.406 2.61% 0.999 0.319

High Growth Firms

Market Model Fama French 4 Factor Model

Event Window N CAAR T-value p-value CAAR T-value p-value

[-1, 1] 23 -1.11%* -1.767* 0.079* -1.00% -1.637 0.103

[-5, 5] 23 -0.35% -0.204 0.839 -0.21% -0.138 0.891

[-10, 10] 23 0.60% 0.587 0.558 0.56% 0.340 0.734

Smallest firms

Market Model Fama French 4 Factor Model Event Window N CAAR T-value p-value CAAR T-value p-value

[-1, 1] 24 -1.23%** -2.228** 0.027** -0.70% -1.065 0.288

[-5, 5] 24 -0.38% -0.135 0.893 0.01% -0.008 0.994

[-10, 10] 24 0.86% 0.429 0.668 0.81% 0.417 0.677

(25)

25 Table 5 shows that for all samples and models the 3-day event period has a negative CAAR. However, not all event windows show negative returns, this was also the case for the full sample results. The Fama French 4 Factor model seems to hold the same trend regarding returns as the market model does. However, the FF4F Model does not show any significance, while the market model does show significance for all subsamples in the 3-day event period.

The difference in both models is explained by the risk calculation method, as can be seen in equation 4 and 5. The FF4F Model uses additional risk factors which, in this case show CAARs closer to zero than the market model. It is highly likely that small, growing firms are more prone risk based on size (SMB) and on value (HML). For further reference, see Fama and French (1991).

(26)

26

5.2 Panel data regression analysis

The results of the regression displayed in equation 9 are shown in table 6:

Table 6:

Number of outside positions regression analysis

The table below displays the results of a fixed-effects (within) regression. The dataset consists of data across time and firms. In total there are 11492 observations and 1490 groups. The dependent variable is the total number of directorships held by a CEO.

Variable Coef. T-value p-value

ROA 0.0783 0.73 0.4670

Ln(Total Assets) 0.0925*** 4.4*** 0.0000***

Board Gender Ratio 0.2339 1.37 0.1710

Board Nationality Mix -0.0023 -0.02 0.9850

CEO Gender (Male = 1) -0.6486*** -7.33 0.0000***

CEO Duality 0.3770*** 11.96*** 0.0000***

CEO Age 0.0183*** 8.44*** 0.0000***

Constant 0.7900** 2.98** 0.0030**

F Value (Prob > F) 54.00 (0.0000)

R-squared 0.0367

*** = Significant at 1% level, ** = Significant at 5% level, * = Significant at 10% level

(27)

27 The results of the second panel data regression, illustrated in equation 10, are shown below in table 7:

Table 7:

Outside position (y/n) regression analysis

The table below displays the results of a fixed-effects (within) regression. The dataset consists of data across time and firms. In total there are 11492 observations and 1490 groups. In this regression, the dependent variable is a dummy variable which takes 1 if the CEO sits on an outside board and 0 if not.

Variable Coef. T-value p-value

ROA 0.0635* 1.75* 0.0800*

Ln(Total Assets) 0.0477*** 6.72*** 0.0000***

Board Gender Ratio 0.1377** 2.39** 0.0170**

Board Nationality Mix -0.0135 -0.34 0.7360

CEO Gender (Male = 1) -0.2030*** -6.8*** 0.0000***

CEO Duality 0.0958*** 9.01*** 0.0000***

CEO Age 0.0069*** 9.38*** 0.0000***

Constant -0.1205 -1.35 0.1780

F Value (Prob > F) 49.38 (0.0000)

R-squared 0.0778

*** = Significant at 1% level, ** = Significant at 5% level, * = Significant at 10% level

Table 7 generally shows the same results as table 6 does. However, the coefficients are lower, which is to be expected with a dummy dependent variable. Furthermore, firm performance has a significant influence in this regression, meaning that firm performance does have an influence on whether a CEO is appointed to at least one outside position or not.

6

DISCUSSION OF THE RESULTS

6.1 First hypothesis

(28)

28 significant in both models and therefore the null hypothesis cannot be rejected. Meaning that it is not proven that there is an investor reaction to a CEO joining an outside board.

Considering the literature discussed, this was not an unlikely result. The reviewed literature offers different views regarding the costs and benefits of joining an outside board. Conyon and Read (2006) developed a model which predicted the CEO would accept too many outside directorships which would result in a decreased value. This was confirmed by Perry and Peyer (2005). However, these findings were all based on multiple outside directorships. This study only focused on first time outside appointments since assuming the role of CEO.

The literature also offered many potential benefits of joining an outside board. A signal of managerial quality or knowledge spill overs for example (see e.g. Fama, 1980; Fama and Jensen, 1983; Kaplan and Reishus 1990). Booth and Deli (1996) argued in favour of the benefits in the form of network building.

The different views that the previous literature offered seem to balance the costs and benefits. Shareholders do not show a significant reaction to their CEO joining an outside board. Therefore, a CEO would not hurt his/her firms value by accepting an outside position. However, accepting multiple outside positions can still destroy firm value (Perry and Peyer, 2005).

6.2 Second hypothesis

The regression analysis of the CARs did not yield any significant results, which is surprising since earlier literature suggests size and growth does have a significant impact (Perry and Peyer, 2005; Booth Deli, 1996). However, both asset growth and firm size did not have a significant impact on the CARs of each individual firm.

(29)

29 significant results. There were no effects observed in the longer window. An interesting result was the negative CAAR for the sample with female CEOs. Looking at the subsample means in de appendix, it shows that the firms of female CEOs are neither high growth nor small. I was not able to find a conclusive explanation for this. Therefore would provide an excellent opportunity for further research.

The significant negative CAARs show that before accepting an outside position, a CEO would be wise to first consider the characteristics of his/her own firm. Booth and Deli (1996) already argued that within small and growing companies, a CEO’s time is valuable. This subsample analysis confirms this view.

6.3 Third hypothesis

The final hypothesis predicts that large firms will have CEOs holding more outside positions, CEO age will positively influence outside positions and CEO duality will positively influence outside positions. This is confirmed by the regression analysis, where all three characteristics have a significant positive influence on the number of outside directorships. This is in line with Booth and Deli (1996) and Kaplan and Reishus (1990). The argument Kaplan and Reishus present is that CEOs of successful, large firms have a good reputation and are therefore in high demand.

Furthermore, female CEOs tend to hold significantly more directorships, This is likely due to the fact that female directors are in high demand (see Spencer Stuart Board Index 2018). According to the index, the two types of directors in highest demand are females and CEOs. Therefore, it is no surprise that a female CEO is in high demand.

(30)

30 CEOs that are also chairperson of the board also hold more outside directorships. This comes as no surprise, since Jensen (1993) already argued that this double-role gives a CEO a tremendous amount of power. Agency theory suggests that with the power to monitor him/herself the CEO will pursue opportunities which will increase personal gain (Fama and Jensen, 1983). The literature suggests that CEOs hold monetary gains and reputational gains in high regard (see e.g. Masulis and Mobbs, 2014; Jensen and Murphy, 1990).

The second regression only checked what variables influence whether a CEO has an outside position or not. An interesting find is that firm performance (ROA) has a positive influence on the outside directorship dummy. This confirms that CEOs that perform well are in higher demand and are ‘rewarded’ with outside board appointment, as argued by Kaplan and Reishus (1990).

7. CONCLUSION

In this research, an event study was performed to see how investors react to the CEO of their firm joining the board of another. To check for only an initial reaction, only first time outside positions are counted. Two risk methods were used in performing the event study, the market model and the Fama French 4 factor model. Furthermore a robust standard error regression was performed on the CARs of both models to check if size and growth rate influences the CARs of firms during these events. A subsample analysis was performed to check whether certain groups showed a different shareholder reaction than the complete sample. To determine what types of CEOs hold outside directorships, a dataset was created including data across time and firms. A fixed effects regression was executed to check which variables influence the number of outside positions held.

(31)

31 The findings of this paper suggest that shareholders do not react significantly to the appointment of their CEO to an outside board. The CARs for both models were negative, however they were not proven to be significant. Therefore, it is concluded that investors do not react to their CEO accepting an outside position for the first time since assuming the role as CEO.

A regression performed on the CARs did not yield any significant results and therefore no variables were found to influence the investor reaction in the event study sample. Size and other factors did not prove to have any effect on the investor reaction in the complete sample.

However, when performing a subsample analysis, the market model did show significant negative CAARs for all three groups. The subsample analysis showed that investors react negatively to the CEO of a high growth firm joining another board, a CEO of a relatively small firm (compared to the complete sample) and a female CEO joining another board a joining another board. The subsample analysis results are in line with previous literature. However, the negative reaction to female CEOs accepting outside positions is surprising, a more thorough analysis on female CEOs might prove useful.

A separate sample was constructed to check whether firm- and CEO-specific factors influence the number of outside positions held by CEOs. This yielded some significant results. A CEO holds more outside directorships for larger firms, if he is also chairperson of the board of his/her own firm, if the CEO is female and somewhat older.

(32)

32 Furthermore, some factors influencing the number of outside positons of a CEO are identified. The findings suggest it is generally accepted for CEOs of large firms to hold multiple outside directorships. Apart from firm size, it is also proven that several personal characteristics have a positive influence on the number of directorships a CEO holds. Older CEOs, CEOs who are also chairperson and female CEOs are all likely to hold more outside directorships. To gain at least one outside appointment, a CEO would increase his/her chances by maintaining a high firm performance (ROA).

An important limitation of this research is the small sample sizes for the subsample analysis. The conclusions drawn from the subsample analysis are drawn using very small samples. Generally, for an event study to be of any significance it needs at least 100 events (MacKinlay, 1997). Therefore the subsamples used in this event study are very small. The subsamples did however prove that the sample does have in-sample differences. Furthermore the event study sample had to be trimmed down because of missing data for various variables. This made the sample significantly smaller than the originals sample, which would have given room for more thorough subsample analysis if it had been larger.

(33)

33

REFERENCES

Adams, R., Ferreira, D., 2007. A theory of friendly boards. Journal of Finance 62, 217–250. Boehmer, E., Musumeci, J., Poulsen, A., 1991. Event-study methodology under conditions of event-induced variance. Journal of Financial Economics 30(2), 253-272.

Booth, J., Deli, D., 1996. Factors affecting the number of outside directorships held by CEOs. Journal of Financial Economics 40(1), 81-104.

Booth, J., Deli, D., 1999. On executives of financial institutions as outside directors. Journal of Corporate Finance 5, 227–250.

Brickley, J., Coles, J., Linck, J., 1999. What happens to CEOs after they retire? New evidence on career concerns, horizon problems, and CEO incentives. Journal of

Financial Economics 52, 341–377.

Brown, S., Warner, J., 1980. Measuring security price performance. Journal of Financial Economics 8, 205-258.

Brown, S., Warner, J., 1985. Using daily stock returns. Journal of Financial Economics 14, 3-31. Carhart, M., 1997. On persistence in mutual fund performance. Journal of Finance 52(1), 57-82. Chopra, N., Lakonishok, J. and Ritter, J. (1992). Measuring abnormal performance: Do

stocks overreact? Journal of Financial Economics, 31(2), pp.235-268.

(34)

34 Cotter, J., Shivdasani, A., Zenner, M., 1997. Do independent directors enhance target shareholder wealth during tender offers? Journal of Financial Economics 43, 195 – 218.

Dahya, J., McConnell, J., 2005. Outside directors and corporate board decisions. Journal of Corporate Finance 11, 37-60.

Fahlenbrach, R., Low, A., Stulz, R., 2010. Why do firms appoint CEOs as outside directors? Journal of Financial Economics 97, 12-13.

Fama, E., 1980, Agency problems and the theory of the firm. Journal of Political Economy 88, 288–307.

Fama, E., French, K., 1993. Common risk factors in the returns on stocks and bonds. Journal of Financial Economics 33, 3-56.

Fama, E., Jensen, M., 1983. Separation of ownership and control. Journal of Law and Economics 26(2), 301-325.

Ferris, S., Jagannathan, M., Pritchard, A., 2003. Too busy to mind the business? Monitoring by directors with multiple board appointments. Journal of Finance 58(3), 1087-1111.

Fich, E., Shivdasani, A., 2007. Financial fraud, director reputation, and shareholder wealth. Journal of Financial Economics 86, 306–336.

Fich, E., White, L., 2005. Why do CEOs reciprocally sit on each other’s boards? Journal of Corporate Finance 11, 175-195.

(35)

35 Hermalin, B., Weisbach, M., 1998. Endogenously chosen boards of directors and their monitoring of the CEO, American Economic Review 88, 96–118.

Jensen, M., Meckling, W., 1976. Theory of the firm: managerial behavior, agency costs and ownership structure. Journal of Financial Economics 3, 305 – 360.

Jensen, M., Murphy, K., 1990. Performance pay and top-management incentives. Journal of Political Economy 98, 225 – 264.

Jensen, M. 1993. The modern industrial revolution, exit and the failure of internal control systems. Journal of Finance 48, 831-880

Kaplan, S., Reishus, D., 1990. Outside directorships and corporate performance. Journal of Financial Economics 27(2), 389-410

Linck, J., Netter, J., Yang, T., 2009. Effects and unintended consequences of the Sarbanes–Oxley Act on corporate boards. Review of Financial Studies 9, 3287–3328.

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

Masulis, R., Mobbs, S., 2011. Are all inside directors the same? Evidence from the external directorship market. Journal of Finance 66(3), 823-872.

Perry, T., Peyer, U., 2005. Board seat accumulation by executives: A shareholder’s perspective. Journal of Finance 60, 2083-2123.

(36)

36 Shivdasani, A., Yermack, D., 1999. CEO Involvement in the selection of new board members: an empirical analysis. Journal of Finance 54, 1829-1853.

Sila, V., Gonzalez, A., Hagendorff, J., 2017. Independent director reputation incentives and stock price informativeness. Journal of Corporate Finance 47, 219-235.

Warner, J., Watts, R., Wruck, K., 1988. Stock prices and top management changes. Journal of Financial Economics 20, 461–492.

(37)

37

APPENDIX

1. Distribution of all independent director announcements

Figure 3:

Distribution of independent director announcements per year.

2. Descriptive statistics event study sample

1156 1808 2251 2327 2178 2253 2389 1372 712 787 714 766 2 0 0 5 2 0 0 6 2 0 0 7 2 0 0 8 2 0 0 9 2 0 1 0 2 0 1 1 2 0 1 2 2 0 1 3 2 0 1 4 2 0 1 5 2 0 1 6 N O . O F A N N O U N C EMEN T S YEAR

I DPT. DI R. ANNO UNCEM ENM TS

Table 10: Summary statistics CEOs/Firms involved in events

Summary statistics of variables used in CAR regression

Variable Obs Mean Std. Dev. Min Max

Total Assets 215 40624.58 182885.9 78.537 2187631

Asset Growth (Event Year) 215 0.11041 0.4182499 -0.441639 5.224293

ROA 215 0.046313 0.0746671 -0.441195 0.2250363

Board Gender Ratio 211 0.856005 0.0921953 0.571 1

CEO Gender 215 0.934884 0.2473067 0 1

CEO Age 215 54.14884 6.002818 34 71

(38)

38 3. Yearly means for regression analysis sample

Table 11: Yearly means of sample Year n Total Assets ROA Gender Ratio CEO Gender CEO Age Total Directorships CEO Duality Outside Directorship 2003 727 13055.99 0.035 0.91 0.98 54.7 2.45 0.65 0.61 2004 792 13884.86 0.043 0.9 0.98 55.1 2.36 0.6 0.6 2005 805 15014.44 0.047 0.9 0.98 54.91 2.27 0.57 0.59 2006 818 17241.88 0.046 0.9 0.98 55.2 2.33 0.56 0.6 2007 796 16753.61 0.048 0.9 0.97 55.42 2.27 0.55 0.59 2008 762 18673.59 0.014 0.89 0.98 55.57 2.27 0.53 0.6 2009 763 17945.82 0.029 0.89 0.97 55.64 2.23 0.5 0.58 2010 782 17573.62 0.045 0.89 0.97 56.05 2.22 0.48 0.6 2011 772 15487.95 0.048 0.88 0.97 56.34 2.24 0.46 0.61 2012 775 19544.49 0.043 0.87 0.97 56.51 2.32 0.43 0.63 2013 759 18877.16 0.047 0.86 0.96 56.59 2.33 0.44 0.64 2014 751 18564.2 0.04 0.86 0.96 56.99 2.38 0.42 0.66 2015 748 21082.61 0.025 0.84 0.96 56.73 2.31 0.4 0.62 2016 722 22423.42 0.032 0.83 0.96 57.1 2.31 0.37 0.62 2017 701 23238.63 0.038 0.82 0.96 57.25 2.25 0.36 0.6

4. Descriptive statistics of regression analysis sample

5. CAAR trend for full sample 11 day event period

Table 12: Descriptive statistics of independent variables used in second regression

Variable Obs Mean Std. Dev. Min Max

ROA 11,492 0.0390526 0.1089359 -2.511917 0.642344

Ln(Total Assets) 11,492 7.856022 1.780195 2.292737 14.63305

Board Gender Ratio 11,492 0.8770226 0.102789 0.333 1

Board Nationality Mix 11,492 0.0808562 0.1545606 0 0.8

CEO Gender (Male = 1) 11,492 0.9702402 0.1699315 0 1

IsalsoChai~n 11,492 0.4900801 0.4999233 0 1

(39)

39

Figure 4:

CAAR development for both models in 11 day event window

-0.20% -0.10% 0.00% 0.10% 0.20% 0.30% 0.40% 0.50% 0.60% -5 -4 -3 -2 -1 0 1 2 3 4 5 C A A R

Days relative to event

CAAR in event window [-5, 5]

(40)

40 6. CAAR trend for full sample 21 day event period

Figure 5:

CAAR development for both models in 11 day event window

-0.80% -0.60% -0.40% -0.20% 0.00% 0.20% 0.40% 0.60% -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 C A A R

Days relative to event

CAAR in event window [-10, 10]

(41)

41 7. Summary statistics for subsamples

Table 13: Summary statistics for subsamples

High Growth Firms:

Variable Obs Mean Std. Dev. Min Max

CEO Duality 24 0.333333 0.481543 0 1

CEO Age 24 53.95833 8.227627 38 71

CEO Gender 24 0.958333 0.204124 0 1

Board Gender Ratio 23 0.859 0.089806 0.727 1

Board Nationality Mix 23 0.859 0.089806 0.727 1

Ln(Assets) 24 8.200047 1.87687 4.36357 11.65185

ROA 24 0.041266 0.112742 -0.37497 0.204992

Asset Growth Evt. Yr. 24 0.699088 1.072249 0.196494 5.224293 Female CEOs:

Variable Obs Mean Std. Dev. Min Max

CEO Duality 14 0.071429 0.267261 0 1

CEO Age 14 51.71429 4.936965 42 59

CEO Gender 14 0 0 0 0

Board Gender Ratio 13 0.793077 0.099472 0.6 0.9

Board Nationality Mix 13 0.793077 0.099472 0.6 0.9

Ln(Assets) 14 9.217664 2.800914 4.496081 13.52769

ROA 14 0.05352 0.065314 -0.03519 0.201266

Asset Growth Evt. Yr. 14 0.039389 0.161786 -0.44164 0.2

Smallest Firms:

Variable Obs Mean Std. Dev. Min Max

CEO Duality 25 0.4 0.5 0 1

CEO Age 25 53.24 5.939136 38 62

CEO Gender 25 0.92 0.276888 0 1

Board Gender Ratio 24 0.899 0.125126 0.571 1

Board Nationality Mix 24 0.899 0.125126 0.571 1

Ln(Assets) 25 5.521645 0.705526 4.36357 6.415159

ROA 25 0.014973 0.113582 -0.37497 0.201266

Asset Growth Evt. Yr. 25 0.247558 1.049515 -0.44164 5.224293

8. Correlation Table

(42)

42 Ln(Total Assets) Asset Growth (Event Year) ROA Board Gender Ratio CEO Gender CEO Age CEO Duality Ln(Total Assets) 1.0000

Asset Growth (Event

Year) -0.0946 1.0000

ROA 0.0599 -0.0672 1.0000

Board Gender Ratio -0.2557 0.0859 -0.1268 1.0000

CEO Gender -0.1680 0.0198 -0.0035 0.1753 1.0000

CEO Age 0.1980 -0.0920 0.1087 -0.0656 0.0973 1.0000

CEO Duality 0.0323 -0.0684 0.0346 -0.2488 0.1724 0.0258 1.0000

9. Correlation table

Table 15: Correlation table for all independent variables used in the separate panel data regression analysis.

ROA Ln(Total Assets) Board Gender Ratio Board Nationality Mix CEO Gender (Male = 1) CEO Duality CEO Age ROA 1 Ln(Total Assets) 0.0505 1

Board Gender Ratio -0.0445 -0.3155 1

Board Nationality Mix -0.0241 0.127 -0.0457 1

CEO Gender (Male = 1) 0.0083 0.0078 0.2386 0.0108 1

CEO Duality 0.0549 0.1005 -0.0196 -0.0155 0.0764 1

(43)

Referenties

GERELATEERDE DOCUMENTEN

In terms of the explanatory power of control variables, only leverage is negatively related to R&D expense to sales ratio at 10% significance level in firm fixed effect

The significantly higher returns can be explained by the fact that a takeover premium is paid over the market value of the target company, which is beneficial for the shareholders

Besides this, when looking at the regression results, statement 1, “I think it is more important to have safe investments and guaranteed returns, than to take risk to have a

Excessive optimism as an indicator for overconfidence in this thesis, is tested by making an estimation of the economic climate which is subtracted from the subcategory of

Lagged NPL is impaired loans over gross loans at time t-1, lagged reserve ratio is the loan loss reserves over impaired loans at time t-1, Slope EU/US is the yield curve

Table 8: The effect of the four components of Corporate Social Responsibility on Corporate Financial Performance as measured by return on assets for European companies from the

Above all, the disaggregated analysis implies that in subgroups of female and high-trust respondents, the happiness positively affects their holding of risky

To elaborate more on the performance of the smart beta portfolio based on employees table 11 gives the overall period return as well as the standard deviation of the returns,