The Effect of CEO characteristics on CEO compensation

The Effect of CEO Characteristics on CEO Compensation

Student ID: 11444223

Programme: Economics and Business Economics Specialization: Finance

Abstract

The aim of this study is to examine the relationship between CEO human capital and total CEO compensation. I focus on the educational aspects of human capital. Specifically, I include the CEO’s level of education, i.e., CEOs that hold a graduate degree, and the educational institution, i.e., CEOs that graduated from an Ivy League institution. The objective is to determine whether higher human capital of the CEO can result in higher total CEO compensation. The sample used in this thesis consists of 100 CEOs of S&P 1500 firms. The findings of this thesis indicate a positive relationship between CEOs who hold a graduate degree and the total compensation. Additionally, CEOs who graduated from Ivy league institutes are also entitled to higher compensation. These results, however, are not significant, possibly due to the data sample limitations. As a result, there is inconclusive evidence to claim that CEO human capital, when represented by educational characteristics, influences CEO compensation.

Statement of Originality

This document is written by Muhammad Muzzammil Hasan 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.

Contents

1. Introduction………. 5

2. Literature Review………...8

2.1. CEO Compensation: Cash vs. Equity Based………....8

2.1.1. Components of CEO Pay……….9

2.1.2. CEO Compensation and its Dynamics Over Time………...9

2.2. CEO Characteristics and Skills and Their Relation to CEO Compensation…………...11

2.2.1. Knowledge, Experience and Attributes Gained (Human Capital) by CEOs………....11

2.2.2. Education as a Measure of Human Capital………..12

3. Methodology………..13 3.1. Independent Variable………13 3.2. Dependent Variable………...14 3.3. Control Variables....………..14 3.4. Data Sample……....………...15 3.5.Descriptive Statistics...………...16 3.6. Research Design…..………...16 4. Results………18 4.1. Correlation Matrix..………...18

4.2. OLS Regression Results…..………...19

5. Conclusion and Discussion………...23

6. References………..26

1. Introduction

Gaines-Ross (2000) highlight the importance of CEOs to firms by reporting that “CEO reputation can represent a staggering 45 percent of a company’s reputation” (p. 365). Therefore, attracting and sustaining competent CEOs is one of the most important aims of a firm. To this end, high levels of executive compensation are often used by firms. As a result, executive compensation has captivated the interests of researchers and practitioners alike. Murphy (1999) states that literature on the subject has witnessed a sharp increase since the 1990’s. The rationale for this is the empirical observation of an excessive increase in CEO compensation during that decade, and since then, CEO compensation has only increased further.

Figure 1: Trends of executive compensation between the years 1936 - 2005 (Frydman and Jenter, 2010)

Frydman and Jenter (2010), in Figure 1, demonstrate the trends of executive compensation between the years 1936 and 2005. Their data shows a decline in executive compensation during World War II, before a slow increase followed by a sharp rise from the mid 70s onwards. This

thesis aims to build on this previous research by exploring the justification behind CEO compensation increase. The justification of interest in this thesis is CEO human capital, specifically their educational characteristics, and how it relates to the variation in executive compensation.

By creating a link between CEO human capital (represented by educational characteristics) and CEO compensation, I attempt to analyse the increasing trend in the latter over the years. According to the market-based theory, the human capital accumulated by CEOs over the years increases the demand for their services. As per the law of demand and supply, then, an excess demand results in increased factor price (Schweinberger, 1979). A higher demand for CEO services therefore should entitle them to higher compensation. Thus, it can be argued that a CEO with a higher level of human capital must be accordingly compensated. In this thesis, I use educational characteristics in the form of first, educational level, and second, graduation from an Ivy League institution, as a construct for human capital. The Ivy League refers to a group of elite and highly selective higher education institutions in the United States, comprising 8 renowned universities (Miller et al, 2014). According to Miller et al. (2014), Ivy League students are fostered in an environment that aims to create human capital, hence they find Ivy League graduates contribute to better firm performance which in return leads to higher CEO compensation.

Compensation moreover is conceptualised as total compensation, which includes salary, bonus, stock options, stock grants and all other payments (Balsam, 2002).

To summarise, given the above argumentation, the higher the educational level of the CEO, the greater the compensation the CEO is entitled to. To take an example, a CEO with an MBA should receive higher compensation in comparison to a CEO possessing a bachelor’s degree. Similarly, a CEO with an Ivy League education is also expected to receive a higher compensation, on account of increasing the firm performance.

Thus, a CEO who has received an education from Harvard Business School, for example, should in turn receive higher compensation in comparison to a CEO with an equivalent education level from a non-Ivy League institution. This brings us to the following research question: What

is the relationship between CEO human capital and CEO compensation?

This thesis therefore aims to build upon and contribute to previous literature on the relationship between CEO characteristics and compensation. To address the research question, I

rely on prior literature to establish the research design, data sample and the control variables used in this thesis. For the research design, I use an ordinary least squares (OLS) regression to find a relation between CEO human capital and CEO compensation. My dependent variable is TCOMP, representing total compensation. My main explanatory variables are GRADDEG and IVY, representing CEOs holding a graduate degree and CEOs receiving an Ivy League education respectively. In addition, I control for non-random influences on my dependent variable. Literature has commonly outlined CEO age, gender and years on board as factors that systematically impact CEO compensation (Jalbert et al, 2002). To retrieve the data for my variables, I use COMPUSTAT - EXUCOMP and BoardEx databases. My final sample consists of a 100 S&P 1500 CEOs that have complete observations on the variables of interest.

The results of my OLS regression indicate a positive relationship between CEO compensation and CEOs with graduate degrees. Moreover, there is also a positive relationship between CEOs who have graduated from Ivy League institutions and CEO compensation. These findings are in line with the results of prior academic literature, as discussed in section 2. However, significant results are not found in my analysis, possibly due to a relatively small sample size. As a result, I cannot find conclusive evidence to support the two hypotheses I posit at the end of my literature review. Thus, there is insufficient evidence to state that executives having a graduate degree or receiving an education from an Ivy League institution are entitled to greater CEO compensation.

The remainder of the thesis is structured as follows: Section 2 comprises a literature review, where I summarize relevant prior research conducted on CEO compensation and CEO human capital. The section concludes with the development of two hypotheses. Section 3 entails the methodology, where I describe the data sources, the data sample along with the descriptive statistics, the dependent, independent and control variables. I then elaborate on the research design and give the model specification. Section 4 provides the correlation matrix and the results of the OLS regression conducted in the study. Finally, Section 5 consists of the discussion and limitations of the study along with concluding remarks.

2. Literature review

This research aims to examine the effect of CEO human capital on executive compensation. More specifically, I have chosen to focus on the education characteristics of CEOs, as representation of their human capital. For a better understanding of the aforementioned, I begin by explaining CEO compensation and what it entails. I then go on to explain how CEO compensation has changed over time and the possible reasons behind this change. I end my literature review by explaining the human capital gained by CEOs over the years and how it relates to their compensation, with my main focus being on educational characteristics.

2.1 CEO Compensation

This section discusses CEO compensation extensively. It sheds light on different types of compensation a CEO can get while working in a firm. This can include cash, bonuses, stock options and stock grants (Balsam, 2002).

2.1.1. Components of CEO Pay: Cash vs Stock Based

Balsam (2002) divided CEO compensation packages into two categories. First, cash compensation. This consists of salary and bonus. Second, stock compensation, which consists of stock options and stock grants.

Cash compensation

Salary is a fixed amount which is paid on a monthly basis to executives. Balsam (2002) states that salary can be variable in terms of a raise when an individual performs well, and a cut due to poor performances and/or global crises. To demonstrate the latter, Leelavathi (2020) predicts that salary cuts are highly likely due to the coronavirus pandemic, as it has decreased socio-economic activity around the world.

Bonuses are considered a reward for meeting specific goals which can include firm specific goals, maximizing shareholder value etc. They usually provide the executive with an incentive to outperform. A bonus is usually a result of collaborative performance or goals achieved by a team of people in the firm (Balsam, 2002). Thus, it is not always the case that the individual performance of an executive is the only factor impacting the compensation they receive. Therefore, cash compensation alone might not be related to CEO characteristics. Shaw and Zhang (2010) give

support to the prior statement, as in their paper they state that the performance of a CEO and their cash based compensation are not related. This may be because bonuses and salaries do not make up the entire compensation a CEO receives, but constitute other elements such as stock compensation. It is important therefore to consider both elements of compensation, otherwise significant results might not be found, as in the case of Shaw and Zhang (2010).

Stock compensation

Stock Options authorize the CEO to secure the stock at a fixed amount over a fixed period of time. It gives an advantage to the CEO when the share price of the firm is above the fixed price of the option and is not worth anything when the opposite holds true. How much the option is worth depends on how the firm performs. The difference between a stock grant and a stock option is that the CEO does not have to pay for a stock grant, making it valuable to the CEO in all cases except for when the price is 0.

The aforementioned characteristics of equity based compensation show that both the performance of the firm and the CEO is necessary for the executive to gain from this type of reward. According to Leone et al. (2006), equity based compensation motivates the executive to enhance their performance in the coming periods.

2.1.2. CEO Compensation and its Dynamics Over Time

As mentioned previously, post-World War II trends in compensation have been a topic of concern and discussion in academia. Frydman and Jenter (2010) extensively researched the trends showing an increase in executive compensation between 1936 – 1970, where they found no major changes. Since the 1970s, however, the compensation of executives rose dramatically. I now turn to explore theoretically why this may be the case.

A possible explanation of the increase in CEO compensation is the agency theory. The theory posits a separation between ownership and control (Berk and DeMarzo, 2019). This distinction is between managers (CEOs) and shareholders of the firm, and the conflict of interest between the two is noted. This conflict of interest originates due to a misalignment of the incentives of both parties (Berk and DeMarzo, 2019). According to Jensen and Meckling (1976), owners in the agency theory are defined as principals, who delegate tasks to the managers, also known as the

agents. These agents perform the assigned tasks in return for compensation and the principals are reliant on them to make decisions for the firm. Bloom et al. (1998) state that shareholders create compensation packages for the managers to motivate them to work towards the maximization of shareholder value. Studies show that board of directors are responsible for setting the compensation of the CEOs. Core (2000) and Lambert et al. (1993) find that for Canadian and American CEOs respectively, that CEOs receive a higher compensation when they have appointed most of the members of the board. Research by Boyd (1994) also shows that CEO pay is higher in firms where board control is low. There is also a positive relationship between insiders in a firm and CEO salaries, which is in line with the aforementioned statement that the board is responsible for setting CEO compensation. In contradiction to Core (1997), Bhagat and Black (1999) state that after the 1960’s, the majority of the board consists of independent directors. Thus, this suggests that agency theory is unable to conclusively explain the increase in CEO compensation.

Similar to the agency theory, the fat cat theory can also be a factor which can explain the increased executive compensation over the years. According to Lin et al. (2018), a fat cat is someone who takes control of the company and has the discretion to increase their salary. Murphy and Záboník (2004), on the basis of the fat cat theory, argue that CEOs who are in the firm for a longer period of time build a close relationship with the board. This in return allows them to increase their compensation, which may be at the expense of shareholders. However, this theory is not in line with the research done by Claudio et al. (2013), who state that external hires are paid significantly more than internal hires due to their generalist skills. Therefore, the explanatory power of the fat-cat theory, in regards to explaining the increase in CEO compensation, is contested.

According to the two aforementioned theories it can be argued that research regarding increased CEO compensation has produced varied results, which may lead to differing conclusions. Although agency theory and the fat cat theory give some perspective of what might affect the CEO compensation, the emphasis of this thesis is to see the increasing trend in CEO compensation over the years and relate it to CEO human capital, represented by educational characteristics. According to the market-based theory, skills and human capital, accumulated by CEOs over the years, has increased their demand. This excess demand has in turn has led to an increase in their compensation. Murphy and Zábojník (2004) and Frydman (2019) explain that CEO compensation has increased over the years due to a change in human capital a firm demands

from CEOs, ranging from firm specific skills to general managerial skills. They go on to explain that this shift in skills increased the competition between prospective CEOs and gives an external hiring option to the firm, as external CEOs might possess general managerial skills that are required by the firm. Frydman and Jenter (2010) add to this by stating that the market based theory forecasts an increase in CEO compensation along with a rise in external hiring.

Cluadio et al. (2013) further show that the experience gained by CEOs in their previous career acts in their favor in terms of receiving higher compensation. Their study indicates that current CEOs possess more general managerial skills than they did previously and possessing these skills increases their compensation significantly.

While there are several market based characteristics and skills a CEO can possess, such as career background, tenure, and education, this thesis is going to focus exclusively on the latter. As such, education characteristics, as a measure for human capital gained over the years, will be explored in relation to CEO compensation.

2.2 CEO Characteristics and Skills and Their Relation to CEO Pay

This section focuses on the qualifications, education level, and the skills gained by CEOs over the years, and how these attributes impact the compensation they earn. It starts off by explaining the knowledge, experience and personality attributes gained by CEOs over time and then goes on to talk about why firms prefer candidates with a graduate degree. It also goes on to create a relation between CEOs who have a graduate degree and generalist CEOs who are preferred by firms, as mentioned by Claudio et al. (2013) in their research.

2.2.1 Knowledge, Experience and Attributes Gained (Human Capital) by CEOs

Combs and Skill (2003) state that every CEO brings a different set of knowledge, experience and capabilities with them which can increase the firm value of the organisation they manage. They can leverage the set of skills they bring to the firm to manipulate their compensation as they are aware of the fact that they can add a significant amount of value to the firm (Pandher and Currie, 2013). Consistent with the previous statement, Hitt et al. (2001) in their study claim that the performance of firms, when compared, differs due to the difference in human capital of CEOs and that some firms gain higher profitability because their CEOs possess greater human

capital. Becker (1972) states that CEO education and previous experience in management are the most common traits used when measuring human capital. Therefore in this thesis I aim to focus on the educational background of CEOs and how it affects their compensation.

2.2.2 Education as a Measure of Human Capital

Previous research suggests that educational background and an individuals’ characteristics do have a relation (Jalbert et al., 2011). According to the screening theory, a person can only become aware of the value of the skills they have acquired when it is complemented with a recognized qualification i.e. a degree (Jalbert et al., 2002). Dollinger (1984) states that educated individuals are more tolerant to vague and ambiguous situations. For instance, if a situation arises that is not inline with the CEOs expertise, a degree level educated CEO would be able to handle that situation in a better way as compared to a CEO who does not hold a degree. Confirming the previous statement, Thomas et al. (1991) state that CEOs with graduate degrees are better at processing information and in return are more responsive to changes, which implies that CEOs with a graduate degree and firm performance have a positive relation. Also, as previously stated by Balsam (2002), an increase in firm performance leads to an increase in equity based compensation of the CEO, which increases their total annual compensation. Bantel and Jackson (1989), in their study, found that there is a positive relation between the education level of a manager and corporate innovation. Jalbert et al. (2004) conducted a research on degrees earned by CEOs and found that a large fraction of companies are managed by CEOs who have earned an MBA. They also found that those executives who do not have an MBA have been with the firm for a greater period of time and have been promoted to CEO level i.e. internal hiring. Adding to this, Graham et al., (2005) found that CEOs with an MBA used methods such as the net present value (NPV) for capital budgeting and the CAPM in cost of capital computations. Hence, these CEOs will make better decisions and predictions with regard to future investments and budgeting that in return may increase the firm value.

Another interesting finding with regard to CEO education states that CEOs from more selective schools have relationships with government representatives can increase firm performance, for example by receiving beneficial tax treatments (Burt, 1992). A study by Perez-Gonzalez (2006) states that a firm performs worse if it is managed by a non-Ivy League undergraduate. Hence, as mentioned previously, the CEO’s equity compensation depends on the

performance of a firm; if the firm performs well, the compensation of the CEO increases. Park (1999) found that individuals with an undergraduate degree experience an earning increase of 21% which supports that education and compensation do have a positive relation. Gottseman and Morrey (2006) conducted research to examine the effect of CEO education on firm performance and their compensation. The results concluded that even though there is not a concrete relationship between CEOs who attended prestigious schools, for example Princeton, and an increase in firm performance, CEOs who attended these schools did have higher compensation packages. Similarly, Pascerella and Smart (1990) find that earning a degree and the institution that degree was earned from, is an important determinant of the CEOs salary.

The arguments mentioned above give an indication that education level and the educational institute are important determinants of CEO compensation. In line with the literature review, I expect that my research will demonstrate that CEOs that have a graduate degree as compared to an undergraduate degree, will get paid a higher compensation package. This leads me to my first hypothesis statement:

H1: CEOs in possession of a graduate degree will earn higher compensation

Secondly, as mentioned earlier, prior research suggests that CEOs who graduated from top tier schools, for example Ivy League schools, will get paid more than CEOs who have not graduated from a top tier school. Hence my second hypothesis is as follows:

H2: CEOs who have graduated from an Ivy league school will earn higher compensation

3. Methodology

This section discusses the method I use to test the hypotheses posited in the previous section. I conduct an empirical investigation using the OLS method to evaluate the relation between my dependent variable and the explanatory variables mentioned in this section. The list of variables, along with their explanation, is shown below in table A1.

3.1 Independent Variable

In my analysis, I use two main independent variables; GRADDEG and IVY. I use the university degree obtained by CEOs, represented by the variable GRADDEG, which is either an undergraduate degree or a graduate degree. I also use the institution the CEO graduated from and assess whether it is an Ivy league school or a non-ivy league school, represented by the variable

IVY. The combination of this information will help me assess the general human capital of a CEO and their management skills. As mentioned in the literature review, previous studies show that CEOs with graduate degrees are better managers, they contribute positively to firm performance, and in return receive higher compensation. According to Gottesman et al. (2006), using these variables may have some negative effects on my research. They state that the education of CEOs does not necessarily always represent their human capital. Hence to overcome this weakness, I also include the number of boards the CEO is or has been a part of. Adding this variable would highlight the career experience of the CEO, hence taking both education and career experience of the to measure their human capital. The aforementioned variables were extracted from BoardEx.

3.2 Dependent Variable

My dependent variable in this study is the annual compensation of CEOs, represented by the variable TCOMP. The CEO compensation is extracted from COMPUSTAT – EXUCOMP, which provides the total annual salary of the CEO, inclusive of both cash and stock based compensation. In the literature review, I have outlined the difference between cash and stock based compensation.

3.3 Control Variables

Since CEO compensation depends on several other variables that may have a systematic influence on it, I control for these variables.

The first variable I control for is the number of years on boards, represented by the variable BOARDYEARS. Since the number of years on boards relates to the career experience gained by the CEO, this will have a positive relation with the CEO compensation (Jalbert et al., 2002). The second variable I hold constant is the gender dummy variable, MALE. This is because Bertrand and Hallock (2001), in their study, find that male executives are paid significantly more than female executives. Also, Khan and Vieito (2013) find that since female CEOs are risk averse, the performance of the firm under their management differs from when the firm is managed by a male CEO. As mentioned in the previous section, since firm performance and the equity based compensation of a CEO have a positive relation, I will hold the gender dummy MALE constant by controlling for its effect.

The third and final variable I control for in my analysis is the AGE of the CEO. In the study by Jalbert et al. (2002), the age of the CEO has a positive relationship with the compensation and has a negative relationship with performance of the firm. Hence, to avoid any discrepancies in my results, I control for the CEOs age.

3.4 Data Sample

The sample for this thesis contains 100 CEOs who have been in this position between the years 2014 until 2020. These individuals have held the CEO position in S&P 1500 firms. My data is extracted from Wharton Research Data Services (WRDS), which consists of the databases COMPUSTAT – EXUCOMP. The compensation of the 100 randomly selected CEOs is hence obtained from EXUCOMP. To get the educational data of the CEOs and other characteristics such as age, years on board and gender, I use BoardEx. I then manually match the compensation of the 100 selected CEOs to their aforementioned characteristics from BoardEx. In the COMPUSTAT – EXUCOMP, I search the entire database which gives me approximately 10000 observations, from which I select observations that are complete and not duplicated, on all variables of interest. This significantly reduces my data sample to 100 CEOs which I will use in my research.

The studies I evaluate in my literature review use a much larger sample than the one in using for example Jalbert et al., (2002) in their study, analyzed 800 CEOs. Even though a larger sample could make my results more generalisable and unbiased, I think using imputation techniques can have result in inaccurate estimates. Therefore, I only use the observations that are complete, thus I examine only 100 CEOs.

3.5 Descriptive Statistics

Table 1: Descriptive Statistics

Variable Obs Mean Std. Dev. Min Max

TCOMP 100 9691.37 11797.2 109.54 108282.3 GRADDEG 100 0.61 0.490207 0 1 IVY 100 0.29 0.456048 0 1 AGE 100 57.51 9.249073 37 89 MALE 100 0.96 0.196946 0 1 BOARDYEARS 100 11.6 11.31914 0 52.4

Table 1 entails the descriptive statistics of the variables of interest. The sample which consists of 100 observations, shows the mean annual total CEO compensation, which consists of both stock and cash based compensation is 9691.37 thousand dollars. The average age of CEOs is 57.51 years with an average of 11.6 years on boards. The summary statistics also show that the majority of CEOs on S&P 1500 companies are male, with females being only 4% of the sample and males accounting for 96% of the CEO positions. This observation is inline with the research conducted by Peltomäki et al., (2015), which shows that only 3% of CEO positions in S&P 1500 firms are held by women.

Looking at the education of these 100 observed CEOs, 29% of the CEOs have a degree from an Ivy league institution. This degree can be either a graduate degree or an undergraduate degree, there is no distinction made in the IVY factor whether the degree is a graduate degree or

an undergraduate degree. Lastly, the statistics show that 61% of CEOs do hold a graduate degree, which can be either an MBA, MS or a Law certification.

3.6 Research Design

To test my hypotheses, I use the OLS regression model, which is a statistical method used to analyse the linear relationship between the dependent and the independent variables. It does this by minimizing the sum of squares between the predicted and observed values of the dependent variable (Stock and Watson, 2014). Hence, I use this model to find a relation between the education of CEOs and their total compensation. In the equation constructed below, TCOMP is the dependent variable which illustrates the total yearly compensation of the CEO for each firm 𝑖 at time 𝑡. GRADDEG and IVY are the main independent variables which represent the education level and the educational institute of the CEO of each firm 𝑖 at time 𝑡, respectively. AGE, GENDER and BOARDYEARS are the control variables, as literature shows they have a non-random influence on compensation, while 𝜀_𝑖𝑡 represents the error term in the equation. My first hypothesis, H1, expects to show a positive relation between the total compensation of CEOs and CEOs with graduate degrees. Hence, to verify this hypothesis, the regression should show a significant and positive GRADDEG coefficient. To verify the second hypothesis statement, H2, which predicts to show a positive relation between the CEOs total annual compensation and CEOs who graduated from Ivy league schools, the independent variable IVY should have a significant and positive coefficient.

𝑻𝑪𝑶𝑴𝑷 = 𝜶 + 𝜷_𝟏𝑮𝑹𝑨𝑫𝑫𝑬𝑮_𝒊𝒕 + 𝜷_𝟐𝑰𝑽𝒀_𝒊𝒕 + 𝜷_𝟑𝑨𝑮𝑬_𝒊𝒕 + 𝜷_𝟒𝑴𝑨𝑳𝑬_𝒊𝒕 + 𝜷𝟓𝑩𝑶𝑨𝑹𝑫𝒀𝑬𝑨𝑹𝑺𝒊𝒕 + 𝜺𝒊𝒕

4. Results

This section describes the empirical results of the research, which I obtained from the statistical analysis described above. I first discuss table 2, which is the correlation matrix that entails correlations amongst the variables in the study. It shows how one variable changes in response to another. Next, I illustrate table 3, which shows the result I obtain by regressing TCOMP on my independent and control variables.

4.1 Correlation Matrix

Table 2 contains the correlations between all the variables included in my model. A perfect correlation has an absolute value of 1 and a correlation with no relationship has a value of 0. I will interpret the results in table 2 using this logic. This table illustrates that there is a positive correlation between my dependent variable - total compensation (TCOMP) – and my first independent variable that is if the CEO holds a graduate degree (GRADDEG). This result is in line with the literature review and first hypothesis, both of which state that a CEO with a graduate degree will earn a higher compensation. The correlation between TCOMP and my second independent variable IVY is also positive. This positive relation also confirms the statements in my literature review and second hypothesis, that an ivy league graduate will earn a higher compensation. The significance of these results will be discussed in the next subsection which talks about the OLS regression results. The negative correlations of AGE and the gender dummy MALE are negative, which are contrary to expectations and shown in the literature review. Finally, BOARDYEARS is positively correlated with CEO compensation. This is expected as more experience enhances skills and human capital, resulting in higher compensation.

Table 2: Correlations Matrix

TCOMP GRADDEG IVY AGE MALE BOARDYEARS

TCOMP 1 GRADDEG 0.117 1 IVY 0.069 0.2399 1 Age -0.0653 0.022 -0.0115 1 MALE -0.4474 -0.1632 0.018 0.1 1 BOARDYEARS 0.0291 0.0024 0.0057 0.5427 0.0362 1

4.2 OLS Regression Results

To test my hypotheses, I run an OLS regression of total annual CEO compensation (TCOMP) on GRADDEG, IVY, AGE, MALE, and BOARD YEARS. Table 3 shows that regressing TCOMP on my independent and control variables, provides me with a significant regression of (F(5, 94) = 5.05, R-squared = 0.2117). This suggests that the model is able to explain 20.17% of the variance. The constant of the model is 37800.84, which is positive and statistically significant at 1%. This shows that when all independent variables are equal to zero, the total compensation is 37800.84.

The main explanatory variable GRADDEG has a positive coefficient of 720.355. This suggests that holding a graduate degree would result in a 720 ($) increase in total compensation, with all other independent variables held constant. This result is in line with what was hypothesized earlier, that a graduate degree holder is expected to have a higher salary. However, these results are not significant, possibly due the small sample size. The second explanatory variable IVY also results in a positive coefficient of 1774.5. Suggesting that the compensation of a CEO who

graduated from an Ivy league institution would be 1774.5 ($) higher, holding all other independent variables constant. This result is consistent with my second hypothesis, that an ivy league graduate will earn a higher salary. Again, due to a small sample size, the results are not significant.

The control variables AGE and MALE, both have a negative coefficient, which is suggestive that an older CEO and a male CEO is likely to have a lower total compensation. However, only the coefficient of the MALE control is significant at 1%, while the coefficient of AGE is insignificant. As discussed in the literature review, this finding is not in line with the previous research. When Mohan and Ruggiero (2003) compare compensation paid to male and female CEOs, they find that female CEOs get paid a lower compensation package. My results may be unreliable due to the fact that in a sample of 100 CEOs, female CEOs only represent 4% of the data. The last control variable, BOARDYEARS, has a positive coefficient of 82.42, which implies that a CEO with more experience, As suggested by the coefficient of number of BOARDYEARS, a CEO gets paid 82.42 ($) more for every additional year of board experience. Although the direction of this relationship is in line with expectations, the coefficient is not significant.

The results of my OLS regression suggest that CEOs with graduate degrees and CEOs who graduated from ivy league schools do not get paid a higher total compensation package. Hence I reject both my hypotheses, according to which graduate and ivy league CEOs get paid a higher salary.

Table 3: OLS Results

Dependent Variable TCOMP

GRADDEG 720.4 (0.31) IVY 1774.5 (0.73) AGE -81.91 (-0.59) MALE -26368.3*** (-4.71) BOARDYEARS 82.84 (0.73) _cons 37800.8*** (4.23) N 100 r2 0.212 t statistics in parentheses * p<0.05, ** p<0.01, *** p<0.001

To examine the robustness of my regressions, first, I test for multicollinearity, as it is an important assumption underlying the multiple linear regression analysis. Multicollinearity exists when the independent variables are highly correlated. The VIF (variance inflation factor) can be used to determine multicollinearity. According to Paul (2006), VIF scores below 5 are usually not problematic. Table A2 shows that the VIF scores for my regression model. All individual VIF scores are well below 5, and the mean VIF score is 1.24. Therefore, my model does not suffer from the problem of multicollinearity.

Next, I examine the distribution of my variables as it is also an important assumption. Most importantly, I am interested in the distribution of my dependent variable, TCOMP, since most of the other variables are dummies. The histogram of TCOMP can be seen in figure A1, which shows that the variable is positively skewed, and therefore is not normally distributed. Therefore, I use a log transformation which significantly improves its distribution. I then repeat the regression analysis, with the logarithm of TCOMP. The results are reported in table A3. The coefficients are still statistically insignificant, in fact now the MALE variable is also insignificant. The only major change is that IVY also now has a negative coefficient, but not significant. Therefore, the results do not significantly change in comparison to the original analysis.

5. Discussion and Conclusion

Does CEO human capital matter? Are executives who hold a graduate degree or a degree from an Ivy league school more attractive for a firm? And does this affect the compensation package they are paid? These are the main questions this thesis attempts to answer. This thesis has investigated the relationship between CEO human capital as a CEO characteristic and its effect on total compensation. Human capital has been separated into two components: i) graduate degree holder CEOs and ii) CEOs who graduated from an Ivy League institution. According to my analysis of the topic, I was expecting to find a positive relationship between both elements of human capital and the total compensation of the CEO. As mentioned earlier, my hypotheses stated that i) a CEO with a graduate degree would earn a higher compensation and ii) a CEO who graduated from an Ivy league institution would also earn higher compensation.

Prior studies conducted, as mentioned in the literature review, show that there is indeed a positive relation between CEO skills and the compensation they are awarded. This suggests that several factors pertaining to human capital do have an impact on CEO compensation. After analyzing studies by Combs and Skill (2003) and Pander and Currie (2013), who conclude that skillful CEOs are compensated based on their human capital, this thesis aimed to take a different approach on human capital by examining the education level of CEOs and its effect on the compensation. The results of this research, showing a positive relation between Ivy League graduates and total compensation, are in line with the literature review. For example, Grottseman and Morrey (2006) also concluded in their paper that graduates from Ivy league schools do earn a higher compensation. However, my results in the thesis, although indicating a positive relationship, show high levels of insignificance. Similarly, according to Thomas et at. (1991), firms managed by graduate degree holders perform significantly better, which in turn results in a higher equity based compensation, thus, greater overall compensation (Balsam, 2002). My results also show a positive relation between CEOs with graduate degrees and total compensation, but this result is also not significant.

Although the main independent variables, after running the appropriate regressions, have the expected direction of the coefficients i.e. both GRADDEG and IVY are positive, the results are not significant. This may be due to several reasons, mostly associated with the data. First, the sample size is relatively small for reliable estimates. 100 CEOs make it to the final sample, which

is a small number for a quantitative study. Moreover, the coefficient of MALE is significant and negative. This is contradictory to what the literature suggests, and also to what is observed in the corporate sector. A possible explanation for this unexpected result is the relative proportion of the female CEOs in the sample. There are 4 female CEOs, in comparison to 96 male CEOs, thus making the comparison inaccurate. As such, the results cannot be considered reliable.

Limitations and Recommendations for Future Research

This thesis has various limitations which account for its insignificant results. These limitations can be worked on to improve future research. Firstly, as mentioned previously as well, the sample size of this thesis is small for a quantitative study. Prior research conducted on the topic uses a much larger sample size, for example more than 4000 observations, to get results which are significant. Secondly, I only examined CEOs of listed companies, specifically of S&P 1500 firms. I might have gotten different results if I would have included privately held firms along with different databases such as the Russell 3000 and FTSE firms. The reason for using listed companies was that data for privately held firms was limited and difficult to obtain. Lastly, this research was conducted using firms which are traded on US stock exchanges i.e. US based firms. The results might be different and generalizable if firms from different countries were examined. However, due to different economy structures, pay scales and exchange rate differences, this method might invalidate the results of the study.

Conclusion

I investigate the relationship between total compensation of CEOs and the education level of CEOs. My results show that human capital, represented by educational characteristics of CEOs, are positively related to total compensation. However, the results are not significant. This poses many questions, one of which being, if the education level of CEO is not of significance, according to previous literature, why do firms hire CEOs with graduate degrees and an Ivy league education and pay them higher executive compensation? This can be because, when hiring, there is limited information available to the hiring committee and a graduate degree or an education from an Ivy league institution acts as a signal of quality. A higher total compensation can also be due to other attributes of human capital that the CEO may possess, for example, proven career success, years of experience and whether or not the CEO has been in an executive position previously. In

conclusion, the level of education is not the only factor that determines the high annual total compensation of a CEO.

6. References

Balsam, S. (2002). An introduction to executive compensation. Academic Press.

Bantel, K. A., & Jackson, S. E. (1989). Top management and innovations in banking: Does the composition of the top team make a difference?. Strategic management journal, 10(S1), 107-124.

Becker, G. S. (2009). Human capital: A theoretical and empirical analysis, with special

reference to education. University of Chicago press.

Bertrand, M., & Hallock, K. F. (2001). The gender gap in top corporate jobs. ILR Review, 55(1), 3-21.

Bhagat, S., & Black, B. (1999). The uncertain relationship between board composition and firm performance. The Business Lawyer, 921-963.

Bloom, M., & Milkovich, G. T. (1998). Relationships among risk, incentive pay, and organizational performance. Academy of Management journal, 41(3), 283-297. Burt, R. S. (1992). Structural holes: The social structure of competition. Harvard university

press.

Combs, J. G., & Skill, M. S. (2003). Managerialist and human capital explanations for key executive pay premiums: A contingency perspective. Academy of Management

Journal, 46(1), 63-73.

Core, J. E. (2000). The directors' and officers' insurance premium: An outside assessment of the quality of corporate governance. Journal of Law, Economics, and Organization, 16(2), 449-477.

Custódio, C., Ferreira, M. A., & Matos, P. (2013). Generalists versus specialists: Lifetime work experience and chief executive officer pay. Journal of Financial Economics, 108(2), 471-492.

Dollinger, M. J. (1984). Environmental boundary spanning and information processing effects on organizational performance. Academy of Management Journal, 27(2), 351-368.

Frydman, C. (2019). Rising through the ranks: the evolution of the market for corporate executives, 1936–2003. Management Science, 65(11), 4951-4979

Frydman, C., & Jenter, D. (2010). CEO compensation. Annu. Rev. Financ. Econ., 2(1), 75-102. Gaines-Ross, L. (2000). CEO reputation: A key factor in shareholder value. Corporate

Reputation Review, 3(4), 366-370.

Gottesman, A. A., & Morey, M. R. (2006). Does a better education make for better managers? An empirical examination of CEO educational quality and firm performance. An

Empirical Examination of CEO Educational Quality and Firm Performance (April 21, 2006). Pace University Finance Research Paper, (2004/03).

Graham, J. R., Harvey, C. R., & Rajgopal, S. (2005). The economic implications of corporate financial reporting. Journal of accounting and economics, 40(1-3), 3-73.

Hitt, M. A., Bierman, L., Shimizu, K., & Kochhar, R. (2001). Direct and moderating effects of human capital on strategy and performance in professional service firms: A resource-based perspective. Academy of Management journal, 44(1), 13-28.

Jalbert, T., Rao, R. P., & Jalbert, M. (2002). Does school matter? An empirical analysis of CEO education, compensation, and firm performance. International Business and Economics

Research Journal, 1(1), 83-98.

Jalbert, T., Jalbert, M., & Perrina, G. (2004). Does degree matter? An empirical analysis of CEO educational paths. Journal of College Teaching and Learning, 5(1), 65-73.

Jalbert, T., Furumo, K., & Jalbert, M. (2010). Does educational background affect CEO

compensation and firm performance?. The Journal of applied business research, 27(1), 15-40.

Jensen, M. C., & Meckling, W. H. (1979). Theory of the firm: Managerial behavior, agency costs, and ownership structure. In Economics social institutions (pp. 163-231). Springer, Dordrecht.

Khan, W. A., & Vieito, J. P. (2013). CEO gender and firm performance. Journal of Economics

and Business, 67, 55-66.

Lambert, R. A., Larcker, D. F., & Weigelt, K. (1993). The structure of organizational incentives. Administrative Science Quarterly, 438-461.

Leelavathi, R. (2020). ECONOMIC IMPACT OF CORONAVIRUS PANDEMIC IN INDIA. Tathapi with ISSN 2320-0693 is an UGC CARE Journal, 19(20), 88-102. Leone, A. J., Wu, J. S., & Zimmerman, J. L. (2006). Asymmetric sensitivity of CEO cash

compensation to stock returns. Journal of accounting and economics, 42(1-2), 167-192. Lin, F., Lin, L. J., Yeh, C. C., & Wang, T. S. (2018). Does the board of directors as Fat Cats

exert more earnings management? Evidence from Benford’s law. The Quarterly Review

of Economics and Finance, 68, 158-170.

Miller, D., Xu, X., & Mehrotra, V. (2015). When is human capital a valuable resource? The performance effects of Ivy League selection among celebrated CEOs. Strategic

Management Journal, 36(6), 930-944.

Mohan, N., & Ruggiero, J. (2003). Compensation differences between male and female CEOs for publicly traded firms: a nonparametric analysis. Journal of the Operational Research

Society, 54(12), 1242-1248.

Murphy, K. J. (1999). Executive compensation. Handbook of labor economics, 3, 2485-2563. Murphy, K. J., & Zabojnik, J. (2004). CEO pay and appointments: A market-based explanation

for recent trends. American economic review, 94(2), 192-196.

Pandher, G., & Currie, R. (2013). CEO compensation: A resource advantage and stakeholder‐ bargaining perspective. Strategic Management Journal, 34(1), 22-41.

Park, J. H. (1999). Estimation of sheepskin effects using the old and the new measures of educational attainment in the Current Population Survey. Economics Letters, 62(2), 237-240.

Pascarella, E. T., & Smart, J. C. (1990). Is the effect of grades on early career income general or conditional?. The Review of Higher Education, 14(1), 83-99.

Paul, R. K. (2006). Multicollinearity: Causes, effects and remedies. IASRI, New Delhi, 1(1), 58-65.

Peltomäki, J., Swidler, S., & Vähämaa, S. (2015). Age, gender, and risk-taking: Evidence from the S&P 1500 executives and firm riskiness. Social Science Electronic Publishing. Pérez-González, F. (2006). Inherited control and firm performance. American Economic

Schweinberger, A. G. (1979). The theory of factor price differentials: The case of constant absolute differentials. Journal of International Economics, 9(1), 95-115.

Shaw, K. W., & Zhang, M. H. (2010). Is CEO cash compensation punished for poor firm performance?. The Accounting Review, 85(3), 1065-1093.

Stock, J.H. and Watson, M.W.. 2014 Introduction to Econometrics. Pearson Education, fourth edition.

Thomas, A. S., Litschert, R. J., & Ramaswamy, K. (1991). The performance impact of strategy‐ manager coalignment: An empirical examination. Strategic management journal, 12(7), 509-522.

7. Appendix

Table A1: Variable Description

Variable Description

GRADDEG This independent variable takes the value of 1 if the CEO has a graduate degree for e.g. Masters, MBA or Law and 0 if the CEO has only an undergraduate degree. (BoardEx)

IVY This independent variable takes the value of 1if

the CEO has graduated from an Ivy League school and 0 if he/she graduated from a non-ivy league school (BoardEx)

TCOMP This is our dependent variable represented in

thousands of dollars. It includes the total compensation i.e. cash based and equity-based compensation provided to the CEO.

(EXUCOMP)

AGE This control variable indicated the age of the

CEO today. (BoardEx)

MALE This control variable (dummy) has the value of

1 when the CEO is male and 0 when the CEO is female. (BoardEx)

BOARDYEARS This control variable represents the number of years the CEO has been part of a board. (BoardEx)

Table A2: VIF scores to check for multicollinearity

Variable VIF 1/VIF

AGE 1.43 0.697204 BOARDYEARS 1.42 0.704741 GRADDEG 1.1 0.91227 IVY 1.07 0.938217 MALE 1.04 0.958338 Mean VIF 1.21

Table 3: OLS Results

Dependent Variable log_COMP

GRADDEG 0.537 (0.24) IVY -0.291 (-1.16) AGE -0.0227 (-1.57) MALE -1.088 (-1.89) BOARDYEARS -0.0024 (0.2) _cons 11.07*** (12.02) N 100 r2 0.0875 t statistics in parentheses * p<0.05, ** p<0.01, *** p<0.001