How is firm value related to CEO tenure for technological and non-technological companies?

How is firm value related to CEO tenure for technological

and non-technological companies?

Name: Jeffrey Bloemen Student number: 10381287 Thesis supervisor: Pepijn Trietsch

Faculty: Faculty of Economics and Business, University of Amsterdam Educational program: Economics and Business

Specialization: Finance and Organization Thesis subject category: Finance

Abstract

This study investigates the relationship between CEO tenure and firm value for technologic and non-technologic firms. Previous research was unable to reach consensus regarding such relationship, although a curvilinear relationship is suggested. Two research models were used for this research. In the first model, tenure per year is regressed on the abnormal returns of firm’s stock, which is the measure for firm value. In the second model, tenure length is put into three categories (‘seasons’), and the relationship between such an individual season and firm value is regressed. The results indicate an increasingly negative relationship between CEO tenure and tech firm value, whereas this relationship is increasingly positive for non-tech firms. However, most of the results appear insignificant, which hinders the external validity of this research.

Statement of Originality

This document is written by Student Jeffrey Bloemen who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

Table of content

2.1 Abnormal stock return as a measure of firm value 6 2.2 CEO tenure related to abnormal returns on firm’s stock 7 2.3 CEO age related to abnormal returns on firm’s stock 9 2.4 Firm size related to abnormal return on firm’s stock 10 2.5 Female executive related to abnormal return on firm’s stock 11

3. Methodology 13

3.1 Model setup 13

3.2 Data description 15

4. Results & analysis 17

4.1 Tenure results 17

4.2 Age results 19

4.3 Company size results 20

4.4 Female dummy results 20

5. Conclusion 21

1. Introduction

1.1 Background

The chief executive officer is ultimately responsible for managing a for-profit organization. Unarguably, performing well at this position requires a high amount of firm- and industry specific knowledge that is primarily obtained through experience. The first years of CEO tenure, the 'honeymoon period', are usually described as the period where a CEO has much to learn about his new job (Hambrick and Fukotomi, 1991). As a result, the CEO is unlikely to perform at his optimal level just yet, which explains the high level of CEO layoffs during these first years of tenure (Frederickson et al, 1988). As a CEO's tenure increases, his quality of decision making should increase as well, up until a certain point when the cons start outweighing the pros of extended tenure (Hambrick and Fukotomi, 1991).

In this thesis I research the relationship between CEO tenure and firm value.

Abnormal return on firm’s stock (read: common shares) is calculated for a sample period of five years, where these abnormal returns indicate whether firm value has increased or decreased during this period. Higher abnormal returns on stock should imply higher firm valuation compared to the market benchmark. The increased quality of decision making, presumably positively related with a CEO's tenure until a certain point, should have a positive relationship with a firm's value until this point is reached. This brings us to the first research question:

"CEO tenure is positively related to a firm's value and this relationship is decreasing after a certain length of tenure"

Technologic (tech) firms deal with a rapidly changing business environment where

abovementioned knowledge is highly important in order to stay ahead of competing firms. For this reason, the relationship between CEO tenure and firm value is expected to be more

positive for tech firms than for non-tech firms. For this research, a tech firm is defined as 'a

firm whose main products or services are dependent on and/or connected through the internet', where the distribution channel of the firm's products and/or services are irrelevant

for the definition of the firm. Therefore, the second research question, which is built on the first research question, is worded as:

"The relationship between CEO tenure and firm value is more positive for tech firms than for non-tech firms, and this relationship is less decreasing after the certain length of tenure"

1.2 Research method

In order to provide a conclusion for the research questions, linear regressions are performed, using panel data from 38 tech- and 25 non-tech, American-based firms over a period of five years. Previous literature investigating the relationship between CEO tenure and firm value has been unable to provide a model to estimate such a relationship. For this reason, the

research in this thesis makes use of dummy variables, where each of these variables represents a certain period of CEO tenure. In addition to these dummy variables, a variety of control variables are included to ensure the validity of the research model. CEO age and firm size are expected to have a negative relationship with firm value (Barker et al, 2012; Keim, 1983), whereas a female CEO is expected to have a positive influence on firm value (Elsaid & Ursel, 2011). The relationship of all of these variables is regressed on the dependent variable

abnormal return on firm's stock, which is calculated as the difference between the yearly

return on stock of a firm compared to the yearly return on stock of a benchmark market index.

1.3 Thesis structure

Chapter two of this thesis summarizes the findings of existing literature that are deemed relevant to the research of this thesis. In addition, hypotheses are provided for each variable that is part of the research models. Chapter three further describes the data, the model setup and the research method. Chapter four provides the results and elaborates on their

interpretation. Lastly, chapter five concludes the research and compares its outcome with that of related literature.

2. Related literature

This chapter summarizes the literature that is relevant for the research. Each paragraph describes relevant information for a specific individual variable that is included in the research model, ending with a to-be-tested hypothesis. Next, chapter three describes how exactly these variables are incorporated into the research models.

2.1 Abnormal stock return as a measure of firm value

Calculating the abnormal return (AR) on a firm's stock outstanding is a popular method of measuring firm performance. As described previously, abnormal return on a firm's stock is calculated as the difference between the return on this stock compared to a benchmark market index. Kim et al (2013) found that partnership announcements by Korean firms lead to firm value creation. Such value creation was measured and concluded by comparing the abnormal returns before the announcement to the abnormal returns after the announcement. Elsaid et al (2011) also used the AR method and found that the stock market initially reacts positively to the announcement of hiring an 'outsider' CEO (who is new to the company in question), but that there is no lasting effect over time. The research performed in these two articles made use of an event study, which is not the method used in this thesis. However, Barber et al (1999) find that there are two long-run abnormal return methods that provide valid t-test outcomes for a specific firm. The first method is the abovementioned event study, the second method uses abnormal returns with a time-series t-statistic. The research of this thesis is conducted on multiple firms rather than on one firm, which means that a panel data t-value is used rather than the time-series t-value implied by the article. Time-series analysis usually consists of a low value of N (sample size) and a high value of T (data observations per individual), whereas panel data analysis is tied to a high N and a low T. However, as the research in this thesis consists of a relatively low value of N combined with a high value of T, the panel data analysis provides a valid t-value, in line with the work of Barber et al (2009).

2.2 CEO tenure related to abnormal returns on firm stock

Previous research has been unable to reach consensus regarding the relationship of CEO tenure and firm value. Hermanlin and Weisbach (1998) provide a model in which the threshold below which a CEO would be fired, increases over time. Thus, in such a model, only high ability CEO's manage to stay appointed for a long period of time. If such a model proves to be true in the real world, this would mean that companies led by the same CEO for a significant amount of time must believe that this person is performing well, as else it would become increasingly beneficial for the firm to fire him. This model appears somewhat in line with the results of Brookman and Thistle (2009), who found that the risk of CEO termination increases for about thirteen years before finally decreasing slightly with CEO tenure. They find that 82% of CEO's have a tenure of less than thirteen years and that improved firm performance leads to longer CEO tenure.

However, the abovementioned results appear contradictory to the results of Eitzen and Yetman (1972), who tested the relationship between coaching tenure and the success of a basketball team. They found that this relationship is curvilinear: team success increases with coaching tenure, but effectiveness decreases after approximately thirteen years of tenure. Here, it makes no sense for the risk of termination to increase with tenure (as predicted by

Brookman and Thistle), as tenure's relationship with team performance is positive. In addition, Eitzen and Yetman would predict the risk of termination to increase after thirteen years rather than decrease, as they consider this to be the point where the effectiveness of increased tenure starts decreasing.

Hambrick and Fukotomi (1991) propose a five-season model to describe a CEO's tenure. The five seasons, in chronological order, are: response to mandate, experimentation, selection of an enduring theme, convergence and dysfunction. However, the article does not assign an approximate time period to any of the seasons, which hurts the usefulness of the research enormously. The authors describe the last season in the five-season model, the Dysfunction season, as the period during which the negative effects of a CEO's continuing tenure outweigh the positive effects (which is mostly increased task knowledge). Their

reasoning is that the CEO characteristics that were most prevalent during the first season, such as high task interest and use of diverse information sources, have steadily declined. However, the power of the CEO has increased, which means that the CEO might stay appointed to his position even when replacement might be the better option for the firm.

Frederickson et al (1988) collected data of 43 CEO's in the food processing industry. The data (Figure 1) show that CEO's have the highest chance of losing their job within the first three years of their tenure, which reflects the idea of a honeymoon period. During this first period of a CEO's tenure, the CEO is highly committed but has many things to learn, which increases the risk of termination. Shen (2003) also found that many CEO's lose their job within three years and recognizes that this is barely enough time for CEO's to complete the process of taking charge.

Figure 1: Completed tenure in the food industry (Frederickson et al, 1988)

A recent survey conducted by the well-known business magazine Fortune asked the CEO's of so-called 'Top-500 companies' what they believed to be the optimal tenure for a CEO. 60% of the respondents considered a tenure period of five to ten years to be optimally related to a firm's performance.1 _{These results appear in line with the findings of Kroll et al (2007), who}

found that a tenure period of slightly above eight years predicts the highest overall

shareholder returns. Their research also supports the findings of Eitzen and Yetman (1972), stating that 'performance increases as tenure rises from low to moderate levels and

deteriorates in the later years'.

Although the exact relationship between CEO tenure and firm value is still unknown, most of the related literature agrees that the relationship between CEO tenure and firm performance is curvilinear: the relationship is positive until a certain point, after which the

relationship becomes negative. As firm performance should, in turn, have a positive

relationship with the firm's stock price, a firm's value should increase with CEO tenure, until a certain length of tenure where the relationship becomes negative. The point where this

relationship becomes negative is estimated to be at approximately 10 years of tenure:

H1a: CEO tenure is positively related to firm value, until tenure reaches 10 years, after which the relationship becomes negative

For tech firms, firm performance is expected to be more dependent on the quality of CEO decision making, due to the rapidly changing business environment where developing new ideas is especially linked to higher profitability (Chakrabarti, 1990). For this reason, the relationship between CEO tenure and firm value should be more positive for tech firms than for non-tech firms, until the approximate cutoff point (10 years) where the relationship should be more negative for tech firms than for non-tech firms, as a result of the higher importance of CEO decision making:

H1b: CEO tenure is more positively related to firm value for tech firms than for non-tech firms, until tenure reaches 10 years, after which the relationship becomes more negative for tech firms than for non-tech firms

2.3 CEO age related to abnormal returns on firm’s stock

The literature appears to strongly support the statement that CEO age is negatively related to firm value. Despite the fact that CEO age and tenure are often correlated, research shows that the two variables may have different outcomes on a firm's performance and value (Barker and Mueller, 2002), implying a negative relationship between age and firm value. There has been a big decline in overall CEO tenure in the past two decades, while younger CEO's, especially those in technology industries, are enjoying extended tenure (Barker et al, 2012).

Previous literature on this subject has revealed that longer expected CEO tenure leads to the firm being valued higher compared to a situation where this expected decision horizon is low (Antia et al, 2010). The reasoning behind this is that as a CEO's age reflects his career horizon, hence his decision horizon, the decision horizon and expected tenure of a younger CEO are longer than that of an older CEO. Barker et al (2012) argue that CEO's with a relatively long career horizon are more likely to adopt risky strategies that can pay off for the

firm in the future. This is in contrast with older CEO's, who might be reluctant to embrace risky strategies as the expected pay-off of these strategies is more likely to occur after that CEO's retirement; older CEO's are more risk averse compared to their younger counterparts.

Researched performed by Sonnenburg (2015) found that CEO fitness has a positive impact on firm value, where a CEO was considered fit if he had ran a marathon in a given year. Sonnenburg's reasoning is that fitness leads to improved cognitive functioning, stress coping and also to improved performance in general and he found that these characteristics naturally decline with aging. Other psychological research performed by Verhaeghen & Salthouse (1997) complements these findings as they find that speed of processing and primary-working memory are human abilities that decline as a result of aging.

H2: CEO age is inversely related with firm value

2.4 Firm size related to abnormal return on firm’s stock

There appears to be consensus regarding a negative relationship between firm size and

abnormal returns on firm's stock. Keim (1983) examined the relationship between the size and the abnormal returns of NYSE's and AMEX's companies' common stocks. The size of a company is measured by its market capitalization: shares outstanding times the stock price. He found that there is always a negative relation between a firm’s size and its abnormal returns. These results appear somewhat in line with the results of Israel & Moskowitz (2013), who observed movements of international stock market portfolios. One of the conclusions of their research is that returns on equity declines with firm size. However, for the stocks of the largest firms, which is what the research in this thesis consists of, they find that this

relationship is insignificant.

In addition, Fama and French (1992) attempted to regress the effect of bèta (β) on a firm's returns, where beta represents the company's level of volatility compared to the level of volatility of the rest of the benchmark market. They find that β is correlated with a firm's size and that it is related to a firm's returns, but only through the correlation with size. They state that the proper inference seems to be that there is a relation between size and average returns, but controlling for size, there is no relation between the bèta (systematic risk) of a firm and its average return. Even though the article does not indicate whether this relationship is actually negative, they refer to the work of Banz (1981), who concludes a negative relationship

between size and stock returns. However, Banz concludes that this does not have to be a causal effect, as firm size could be a proxy for other variables that predict stock returns.

Lastly, Brown et al (1983) show that the relation between excess returns and firm size can be regarded as linear in the log of size. Next they show that excess returns predicted with size are not constant through time. However, they also show that different estimation

methodologies lead to different conclusions about the effect of size, which is trivial for the external validity of their first two conclusions.

H3: Company size is inversely related with firm value

2.5 Female executive related to abnormal return on firm’s stock

Firms led by a female CEO might have an advantage over firms led by male CEO's. Mersland and Strøm (2009) found that a female CEO was one of the factors responsible for higher performance in microfinance institutions. They argue that a female CEO might know better what products women want and is therefore able to reduce the company's information asymmetry related to customers more than a male CEO would be able to do.

Abovementioned results confirm the findings by Welbourne (1999), who found that having women on the top management teams has a positive effect of approximately 7% on the firm's short-term performance, growth in earnings per share and three-year stock price growth. In addition, Campbell & Minguez (2010) researched whether female board appointments in Spanish boardrooms led to higher firm value. They found that both short- and long-term firm value indeed increased when a woman was appointed to the board.

It appears that having women as a part of the company has a positive impact on a firm's value. However, the reason behind this effect remains relatively unknown. According to Elsaid & Ursel (2011), female CEO's are significantly more likely to bring a risk-reducing leadership policy to a firm. This could prove to be beneficial for firms that take on excessive risky behavior.

H4a: Having a female CEO has a positive effect on firm value

However, as was brought to light in chapter 2.3, lower risk is not necessarily a good thing (Barker et al, 2012). A risk-taking policy might prove to be the optimal strategy in certain industries in order to stay or get ahead of competing firms. Especially in tech firms, taking

risks seems especially necessary in an attempt to get a competitive edge on competitors. For this reason, having a female CEO is expected to negatively impact the value of a technologic firm.

3. Methodology

This section describes the model setup, the data and the research method. Two research models are used to improve the validity of the research.. The first model predicts with tenure periods categorized per year. The second model predicts with tenure periods characterized per ‘season’, which is a longer period of tenure. The next chapter shows the results.

3.1 Model setup

Previous research has suggested a curvilinear relationship between CEO tenure and firm value (Eitzen and Yetman, 1972). As tenure increases, the CEO increases his task knowledge and efficiency, until the point where extended tenure is no longer beneficial to the firm (Hambrick and Fukotomi, 1991). The research for this thesis was conducted with two different models.

The first model:

Model 1: ABNORMALRETURN = α + β1TENURE1-2 + β2TENURE2-3 + β3TENURE3-4 + β4TENURE4-5 + β5TENURE5-6 + β6TENURE6-7 + β7TENURE7-8 + β8TENURE8-9 + β9TENURE9-10 + β10TENURE10+ β11AGE + β12MARKETCAP + β13FEMALE + ε

where:

 ABNORMALRETURN = abnormal return on firm's stock

 TENURE1-2 = dummy variable, 1 if: 1 years ≤ CEO tenure < 2 years

 TENURE2-3 = dummy variable, 1 if: 2 years ≤ CEO tenure < 3 years, etcetera

 TENURE10+ = dummy variable, 1 if: CEO tenure ≥ 10 years

 AGE = CEO age in years

 MARKETCAP = market capitalization of CEO's firm in billions of dollars

 FEMALE = dummy variable, 1 if female

The idea behind this model is to test whether different lengths of CEO tenure, taken per year, are related to the abnormal return on a given firm's stock, controlling for other variables that could potentially impact this relationship (age, market capitalization and a female CEO). Ideally, a dummy variable for year one (tenure period from day one to year one; day 1 until day 364) would be included in the model. However, this would cause a dummy variable trap,

leaving a regression useless. As a result, the TENURE dummy variables indicate the relationship between “that amount of tenure - firm value”, compared to the relationship of

“less than 1 year of tenure - firm value”.

According to previous research, abnormal return on company stock is an appropriate measure of firm performance during that time period (Barber et al, 1999). In order to calculate the abnormal return over a time period, the expected return over that time period has to be calculated first. This model is listed in the article of Elsaid et al (2011):

Rt = α + β(Rmt) + ε

where:

 Rt = Expected return of firm’s stock at time t

 β = Sensitivity of the expected excess asset returns compared to the expected excess market return

 Rmt = Expected return of the market at time t

Lastly, in order to calculate abnormal returns, I use another model that is provided by Elsaid et al (2011). Here, the S&P500 is taken as the benchmark market index.

ARt = Rt - (α + βRmt)

where:

 AR = Abnormal return of firm’s stock at time t  R = Expected return of firm's stock at time t  Rm = Expected return of the market at time t

The second model that will be used for the research is based on the season-model provided by Hambrick and Fukotomi (1991). The length of tenure can be represented by one of three ‘tenure seasons’. Season one represents the tenure period from 1 to 5 years, season two represents the tenure period from 5 to 10 years and season three represents the tenure period from 10 years and onwards. Again, the tenure period from 0 to 1 year was omitted to prevent a dummy variable trap from occurring. Again, assuming a curvilinear relationship between tenure and abnormal returns, the SEASON2 dummy is expected to be more favorable (more positive or less negative) than the dummy of SEASON1, and the SEASON3 dummy is expected to be less favorable than the dummy of SEASON2.

The second model:

Model 2: ABNORMALRETURN = α + β1SEASON1 + β2SEASON2 + β3SEASON3 + β4AGE + β5MARKETCAP + β6FEMALE + ε

where:

 ABNORMALRETURN = abnormal return on firm's stock

 SEASON1 = dummy variable, 1 if: 1 years ≤ CEO tenure < 5 years  SEASON2 = dummy variable, 1 if: 5 years ≤ CEO tenure < 10 years

 SEASON3 = dummy variable, 1 if: CEO tenure ≥ 10 years

 AGE = CEO age in years

 MARKETCAP = market capitalization of CEO's firm in billions of dollars

 FEMALE = dummy variable, 1 if female

3.2 Data description

The research in this thesis makes use of panel data, using the STATA12 program to perform linear regressions. In order to do so, the yearly abnormal returns are calculated for every firm that is part of the research. A time period of five years is used: January 2009 until December 2013. The NASDAQ-100 technology index (^NDXT) is used to represent tech firms. This index consists of the 38 components of the NASDAQ-100 index that are classified as

'technologic companies', such as Google and Nvidia. Next, the Dow Jones 30 index (^DJI) is used to represent 'non-technologic' companies. This index consists of thirty large, publicly owned components and is arguably the most known index worldwide. In total, five companies from this index are omitted as they are already part of the NASDAQ-100 technology index (Microsoft, Cisco, Apple, Intel) or also classified as a technologic ('a firm whose main

products or services are dependent on and/or connected through the internet') company

(IBM). This leaves a sample size of 25 companies, for a total of 125 DJI observations (5 years) compared to 190 observations for the NDXT.

The needed data is collected from the Wharton Research Data Services (WRDS) website.2 _{The website provides data regarding CEO's age, tenure and sex, in addition to}

providing data about a company's historical adjusted stock closing price. In a few cases, data for a certain time period is missing, after which the entire data observation for that time period was omitted from the sample.

Five data observations in the tech sample were considered outliers. After omission of these observations, the maximum length of tenure for the tech company data is 319 months, with a minimum of 3 months. For the non-tech sample this is a 155 month maximum with a 4 month minimum. The average period of tenure in the tech firm sample is 94.2 tenure months, whereas it is only 58.8 months in the non-tech sample. The average age of CEO’s is 53.2 for the tech sample and 57.9 for the non-tech sample. This is in line with the work of Barker et al (2012), who found that younger CEO’s are enjoying longer tenure, especially in the

technology sector. The non-tech sample does not consist of any CEO’s with tenure over 13 years. As a result, this thesis uses ten years as the approximate cutoff point where the curvilinear relationship between tenure and firm value becomes negative.

After omission of outliers and incomplete data observations, the NDXT sample consists of 181 data observations and the DJI consists of 123 data observations. The results of the research are described in the next chapter.

4. Results & analysis

This chapter shows the results of the research and provides context for their interpretation. Each paragraph discusses the results per individual variable and concludes whether the relevant hypothesis, as defined in chapter 2, is accepted or rejected.

4.1 Tenure results

H1a: CEO tenure is positively related to firm value, until tenure reaches 10 years, after which the relationship becomes negative

H1b: CEO tenure is more positively related to firm value for tech firms than for non-tech firms, until tenure reaches 10 years, after which the relationship becomes more negative for tech firms than for non-tech firms

Figure 2 shows results of the first, non-season regression model. The output shows that for tech companies, CEO tenure does not have a positive relationship with abnormal returns. Most of the TENURE coefficients are insignificant, and the ones that are significant show a negative relationship between tenure and abnormal returns. However, a tenure period of more than 10 years appears to have a significantly negative relationship with abnormal returns, as predicted by the literature.

For non-tech firms, CEO tenure does appear positively related to abnormal returns. However, tenure above 10 years is positively related with abnormal returns, which is contradictory to literature’s predictions, and now only the ninth year of tenure has a

significant result. This is in line with the results of Kroll et al (2007), who found that a tenure period of just over 8 years yields the highest shareholder returns. However, for tech firms the coefficient of the ninth year of tenure predicts the most negative relationship with abnormal returns: the exact opposite compared to non-tech firms and completely contradictory to what Kroll et al (2007) predicted.

Figure 2: Tenure model per year; model 1

Figure 3 shows results of the second, three-season regression model. The relationship between CEO tenure and abnormal returns appears to be increasingly negative for tech firms, where only the tenure period of 10 years or more has a coefficient significant from zero.

For non-tech firms, this relationship appears to be completely different. Here, the relationship is increasingly positive, even further after 10 years of tenure. From the results it can be concluded that both Hypothesis 1a and 1b are rejected, as the relationship per year or season of tenure and a firm’s abnormal returns appears to be quite random. An interesting finding, however, is that CEO tenure above 10 years has a significantly negative relationship with abnormal returns for tech firms, whereas for non-tech firms this relationship is

insignificantly positive.

Variable Coeff. tech Coeff. non-tech

TENURE1-2 -0.0032 (0.0157) -0.0045 (0.0099) TENURE2-3 0.0047 (0.0169) -0.0108 (0.0106) TENURE3-4 -0.0050 (0.0184) -0.0129 (0.0100) TENURE4-5 -0.0098 (0.0159) -0.0037 (0.0103) TENURE5-6 0.0024 (0.0162) -0.0082 (0.0106) TENURE6-7 -0.0283* (0.0161) -0.0156 (0.0112) TENURE7-8 -0.0140 (0.0173) -0.0041 (0.0113) TENURE8-9 -0.0293* (0.0173) 0.0209* (0.0121) TENURE9-10 -0.0014 (0.0183) -0.0107 (0.0135) TENURE10+ -0.0219* (0.0129) 0.0066 (0.0128) AGE -0.0004 (0.0005) 0.0005 (0.0007) MARKETCAP -0.0004 (0.0004) -0.0005* (0.0003) FEMALE -0.0149 (0.0213) 0.0068 (0.0119) * significant at the 10% level R² = 0.0685

Observ. = 181

R² = 0.1466 Observ. = 123

Figure 3: Tenure model per season; model 2

4.2 Age results

H2: CEO age is inversely related with firm value

The results indicate that CEO age is negatively related with abnormal returns for technologic companies, but positively related with abnormal returns for non-technologic companies. For this reason, H2 is accepted for tech companies and rejected for non-tech companies. The most likely explanation for this outcome lies in the fact that technology is a relatively new

phenomenon. It is common knowledge that technology has become an increasingly important factor in people's lives only over the recent decades. For this reason, younger CEO's are more likely to have more knowledge surrounding technology and are preferred over older CEO's in the technology sector. For CEO's of non-tech companies, however, the additional knowledge and experience that comes with aging is more likely to be relevant for performing well at the job. However, no CEO age coefficients appear to be significant in either data output.

Variable Coeff. tech Coeff. non-tech

SEASON1 -0.0037 (0.0120) -0.0079 (0.0085) SEASON2 -0.0143 (0.0120) -0.0044 (0.0091) SEASON3 -0.0220* (0.0129) 0.0064 (0.0131) AGE -0.0004 (0.0005) 0.0005 (0.0006) MARKETCAP -0.0003 (0.0004) -0.0005 (0.0003) FEMALE -0.0145 (0.0204) 0.0070 (0.0122) * significant at the 10% level R² = 0.0364

Observ. = 181

R² = 0.0491 Observ. = 123

4.3 Company size results

H3: Company size is inversely related with firm value

Figure 2 and 3 both show that the size of a firm (market capitalization) is negatively related with a company's abnormal returns, albeit only significant at a 10% level for non-technologic companies. Therefore, it could be assumed that the theories (Banz, 1981; Israel & Moskowitz, 2013) regarding the inverse relationship between company size and abnormal returns are true, which means that hypothesis 3 is accepted but that firm size could be correlated with other factors, explaining the negative relationship with firm value. However, the results may

definitely be biased as both the NDXT and DJI indices contain larger-than-average companies. The lack of small, publicly traded companies in the data makes it so that providing a valid conclusion regarding the relationship between company size and abnormal returns, is not possible.

4.4 Female dummy results

H4: Having a female CEO has a positive effect on firm value

According to the research, a female CEO is negatively related to abnormal returns for tech firms, and positively related to these returns for non-tech companies, although the coefficients are not significant. This negative relationship for tech companies could be explained by the argument that tech firms are more dependent on innovative ideas in order to create a

competitive advantage (Chakrabarti, 1990). A risky leadership policy might be appropriate for such firms, which clashes with the mindset of relatively risk averse women (Elsaid & Ursel, 2011). However, both of indices used for this research contained only one firm with a female CEO. For this reason, there is not enough evidence to confirm a positive relationship, as implied by Mersland (2009) and Welbourne (1999).

5. Conclusion

Related literature found that firm performance, hence firm value, increases as CEO tenure increases (Eitzen and Yetman, 1972). This positive relationship holds on until a certain amount of tenure, after which the relationship becomes negative: a curvilinear relationship (Brookman and Thistle, 2009). For tech firms, the CEO’s quality of decision making is arguably more important in order to get ahead of competition, as innovation is a more pronounced phenomenon in this industry (Chakrabarti, 1990). As a result, the

abovementioned relationship is expected to be more positive for tech firms until the cutoff point, from where on this relationship should be more negative. The research of this thesis asked the question whether the relationship between CEO tenure and firm value is indeed curvilinear and whether this relationship is stronger for tech firms than for non-tech firms.

For tech firms, this relationship appears to become increasingly negative, rather than positive. After the cutoff point of ten years of tenure, the relationship becomes even more negative. For non-tech firms, the relationship becomes increasingly positive. Especially the ninth year of tenure is associated with high abnormal returns, which supports the findings of Kroll et al (2007) that a tenure period of just over 8 years is associated with the highest shareholder returns. However, as most of the tenure coefficients in the models are

insignificant, the applicability of the used models seems limited. This is in line with previous research: no consensus has been reached regarding the relationship between CEO tenure and firm value.

A CEO’s age is negatively related to a technological firm’s value, whereas it is positively related to a non-tech firm’s value. The likely explanation for this result is that younger CEO’s are more valuable for tech firm’s results, whereas older, “experienced” CEO’s have an edge in non-tech businesses. However, the results were not significant. Female CEO’s are related to lower firm value for tech firms, likely due to their risk averse mentality. For non-tech firms, the opposite appears to be true. However, the coefficients are again not significant and only two females were part of the data sample. Lastly, the size of a firm appears to have a negative relationship with a non-tech firm’s value, and this came out as the only significant control variable coefficient.

Future research could make use of a bigger sample size in order to enhance the strength of the models. The use of more (control) variables could enhance the model's

prediction accuracy, although finding relevant variables has proven to be hard. In addition, it is definitely possible that the season-of-tenure model proves not effective enough to predict the complicated relationship between CEO tenure and a company's performance. The main shortcoming of the models is that they lack causality: there are many variables that could correlate with tenure seasons and a tenure season on its own does not explain why the value of a company changes.

References:

Antia, M., Pantzalis, C., Park, J.C., 2010. CEO decision horizon and firm performance: An investigation. Journal of Corporate Finance 16, 288-301.

Atiase, R.K., Bamber, L.S., Tse, S.T., 1989. Timeliness of financial reporting, the firm size effect and stock price reactions to annual earnings announcements. Contemporary

accounting research 5, 526-552.

Banz, R.W., 1981. The relationship between return and market value of common stocks.

Journal of Financial Economics 9, 3-18.

Barber, B.M., Lyon, J.D., Tsai, C., 1999. Improved Methods for Tests of Long-Run Abnormal Stock Returns. The Journal of Finance 54, 165-201.

Barker, V.L., McClelland, P.L., Oh, W., 2012. CEO career horizon and tenure: Future

performance implications under different contingencies. Journal of Business Research

65, 1387-1393.

Barker, V.L., Mueller, G., 2002. CEO characteristics and firm R&D spending. Management

Science 48, 782-801.

Baumrin, S., Frederickson, J.W., Hambrick, D.C., 1988. A Model of CEO Dismissal . The

Academy of Management Review 13, 255-270.

Berzkalne, I., Zelgalve, E., 2014. Intellectual Capital and Company Value. Procedia - Social

and Behavioral Sciences 110, 887-896.

Brookman, J., Thistle, P.D., 2009. CEO tenure, the risk of termination and firm value. Journal

of Corporate Finance 15, 331-344.

Brown, P., Kleidon, A.W., Marsh, T.A., 1983. New evidence on the nature of size-related anomalies in stock prices. Journal of Financial Economics 12, 33-56.

Campbell, K., Minguez, V.A., 2010. Journal of Management & Governance 14, 37-59. Chakrabarti, A. K. 1990. Scientific output of small and medium size firms in high tech

industries. IEEE Transactions on Engineering Management, 48-52.

Eberhart, A.C., Maxwell, W.F., Siddique, A.R., 2004. An Examination of Long-Term Abnormal Stock Returns and Operating Performance Following R&D Increases. The

Journal of Finance 59, 623-650.

Elsaid, E., Wang, X., Wallace, N.D., 2011. Does experience matter? CEO successions by former CEOs. Managerial Finance 37, 915-939.

Eitzen, D.S., Yetman, N.R., 1972. Managerial Change, Longevity, and Organizational Effectiveness. Administrative Science Quarterly 17, 110-116

Fama, E.F., French, K.R., 1992. The Cross-Section of Expected Stock Returns. The Journal

of Finance 47, 427-465.

Hermanlin, B.E., Weisbach, M.S., 1998. Endogenously Chosen Boards of Directors and Their Monitoring of the CEO. The American Economic Review 88, 96-118.

Huson, M.R., Malatesta, P.H., Parrino, R., 2004. Managerial succession and firm performance.

Journal of Financial Economics 74, 237-275.

Israel, R., Moskowitz, T.J., 2013. The role of shorting, firm size, and time on market anomalies. Journal of Financial Economics 108, 275-301.

Karaevli, A., 2007. Performance Consequences of New CEO 'Outsiderness': Moderating Effects of Pre- and Post-Succession Contexts. Strategic Management Journal 28, 681-706

Keim, D.B., 1983. Size-related anomalies and stock return seasonality: Further empirical evidence. Journal of Financial Economics 12, 13-32

Kim, D., Lee, H., Minkyo, S., 2013. Market valuation of marketing alliances in East Asia: Korean evidence. Journal of Business Research 66, 2492-2499.

Kroll, M.J., Walters, B.A., Wright, P., 2007. CEO tenure, boards of directors, and acquisition performance. Journal of Business Research 60, 331-338.

Mersland, R., Strøm, R. Ø., 2009. Performance and governance in microfinance institutions.

Journal of Banking & Finance 33, 662-669.

Verhaeghen, P., Salthouse, T.A., 1997. Meta-analyses of age-cognition relations in adulthood: estimates of linear and nonlinear age effects and structural models. Psychological

bulletin, 231-249

Shen, W., 2003. The Dynamics of the CEO-Board Relationship: An Evolutionary Perspective.

The Academy of Management Review 28, 466-476.

Sonnenburg, F., 2015. CEO fitness and firm value. University of Cologne, Centre for

Financial Research, 1-65.

Taylor, L.A., 2010. Why Are CEOs Rarely Fired? Evidence from Structural Estimation. The

Journal of Finance 65, 2051-2087.

Welbourne, T.M., 1999. Wall Street Likes its Women: An Examination of Women in the Top Management Teams of Initial Public Offerings. CAHRS Working Paper Series 99-07.