Envy vs. Greed

Envy vs. Greed

Whom should a board pick as their CEO

to enhance shareholder value?

MSc A&C Accountancy 2015 - 2016

Maike Barge S1752103

m.v.barge@student.rug.nl Mobile: 06 50455035 Faculty of Economics and Business

University of Groningen

1

Abstract

Is there a certain degree of chief executive officer (CEO) envy or greed most desirable when aiming to enhance shareholder value? This thesis is the first to investigate this proposed relationship, firmly rooted in agency theory. Two methods for the quantification of the degrees of CEO envy versus greed have been introduced for the purpose of this research paper. Correspondingly, the established degrees of both envy and greed have been compared to changes in shareholder value. The high level of data granularity enabled distinction between four different degrees of each trait, in addition to which an interrelation with changing market circumstances, such as the economic crisis (2007-2011), was found upon analysis of the results. Since the analysis revealed no significant correlation between certain degrees of envy or greed and shareholder value, the proposed relationship cannot be proven and both hypotheses had to be rejected. Subject to certain limitations, including the sample size and geographical scope of this thesis, the generalizability of the results could be significantly influenced. This thesis was an attempt to quantify the CEO character traits envy and greed, to enable data modeling and identify the effect on shareholder value, striving to guide executive boards in the CEO appointment process. Further research, building forward on these concepts, could remarkably affect economic decision-making in the future.

Keywords: (chief) executive officers, envy, greed, shareholder value, agency theory

Supervisor: Shibashish Mukherjee Co-assessor: Vlad Andrei Porumb Word count: 9655

2

Table of Contents

1. Introduction

2. Literature Review

2.1 Prior literature on Envy and Greed 6

2.1.1 Literature on Greed 2.1.2 Literature on Envy 2.1.3 Literature on the chosen industry for envy and greed 2.2 Agency theory with a focus on the concepts on envy and greed 9

2.2.1 The concepts of the Agency theory / Shareholder value

3. Hypothesis development

3.1 Hypothesis development for envy 12

3.2 Hypothesis development for greed 13

4. Research method

4.1 Variables and sample 14

4.1.1 Variables and sample used to quantify greed and envy 4.1.1.1 Validity of the data 4.1.2 Variables and sample used to measure the effect on shareholder value 4.1.2.1 Reliability and validity of the variables 4.2 Methodology 17

4.2.1 Quantifying CEO envy 4.2.2 Quantifying CEO greed 4.2.3 Measuring the effect of CEO envy/greed on shareholder value 4.2.3.1 Regression Model 4.2.3.2 Robustness Model

5. Results

5.1 The quantification of envy 21

5.1.1 Establishment of the percentiles to distinguish in the degrees of envy 5.1.2 The four percentiles of envy

3

5.1.3 Interpretation of the results on the quantification of envy 5.2 The quantification of greed 23

5.2.1 Establishment of the percentiles to distinguish in the degrees of greed 5.2.2 The four percentiles of greed 5.2.3 Interpretations of the results on the quantification of greed 5.3 Statistical tests 24

5.3.1 Descriptive Statistics 5.3.1.1 Interpretations of the t-tests for envy 5.3.1.2 Interpretations of the t-tests for greed 5.3.2 Multicollinearity 5.3.3 Regression results 5.3.3.1 Regression results on envy 5.3.3.2 Interpretations of the regression results concerning envy 5.3.3.3 Regression results on greed 5.3.3.4 Interpretations of the regression results concerning greed 5.3.3.5 Robustness test

6. Discussion and Conclusion

6.1 Discussion and Conclusion 32

6.2 Research limitations and future research 33

7. References

8. Appendices

4

1. Introduction

The desire for power, wealth and food are experienced by every human being. But to what extent? And what impact does a certain degree of desire have on others? This paper investigates the human traits that awaken such feelings. Human traits known as envy and greed. Where envy corresponds to the desire to be in another person’s situation in terms of wealth, food and experience. Smith et al. (1999) similarly portray envy as the “feeling aroused when one person desires another’s advantage”. Greed, on the other hand, refers to an excessive desire for power, wealth and food. In line with this, it is defined as “the desire for a pursuit of extraordinary material wealth” (Haynes et al., 2014).

Considering the accounting atmosphere in which this empirical study is observed, the human traits of Chief Executive Officers (CEOs) are studied. By quantifying envy and greed, the degree of how much each CEO embodies this trait is measured. Special emphasis is put on the “degree” of each trait, since earlier research has shown that CEOs are not uniformly greedy (Haynes et al., 2015), nor is there one degree of envy (Endriss et al., 2003). Controlling for a certain industry, this research will therefore build on earlier academic research by quantifying the degree of a CEOs envy (Bizjak et al., 2011), and the degree of a CEO’s greed (Bebchuk et al., 2011). Intriguing that there is only limited literature on greed and envy within the business sector, this thesis will be the first to quantify both CEO envy and CEO greed. Building upon this, the purpose of distinguishing in the degree of both human traits of CEOs in the accounting atmosphere in which this paper is written, corresponds to:

Investigating the influence a certain degree of CEO envy or greed will have on the value generated for shareholders.

Shareholders are the ones that are referred to as “others” in the opening sentence of this academic research. They are the ones that have their money at stake. They are the actual risk takers, investing their capital into the firm, while the board of directors makes the final decisions.

5 Generally, board members are not made liable or charged with financial penalties for misinterpretations, since firms buy insurances for their board members. Consequently, depending on the compensation structure and the limited risk they bare in their decision making, the board members, especially the CEO, might not be shareholder-oriented under certain circumstances (The Economist, 2002; Walters et al., 2008). Likewise, Jensen and Meckling (1976) state that interest alignment between shareholders and CEOs is not achieved when CEOs ownership of firm shares is negligible. In this setting, CEOs are often expected to grow the firm in order to increase their own salaries (Walters et al., 2008). Interesting to understand, concerning the knowledge regarding shareholder value, is whether these CEO desires of growing the firm to increase their own wealth are stimulated by a high degree of envy or greed. Ultimately, the degree of ownership in the firm granted to the CEOs, referring to stock options, will also play an important role in setting his1 _{bargaining power and affecting the alignment of interests with shareholders. These}

diverging motivations, and ultimately moral hazard problems as described by the agency theory, will be investigated in this paper. (Jensen and Meckling, 1976; Fama, 1980; Fama and Jensen, 1983).

Gaining knowledge on the extent to which the degree of CEO envy or greed effect shareholder value, taking account of the compensation structure, would be remarkably significant to the business sector. The board of directors could take account of this newly gathered information by making more strategic decisions in hiring a CEO, and for the shareholders who put their money at stake, it would be an interesting new source of information to base their choice of investments on.

This paper continues by introducing the relevant prior literature to this research. By dividing this section into two sub-sections, a clear distinction can be made between the two steps that are essential before answering the main research question. Next, the data sources and methodological approach will be elaborated. Proceeding, the empirical results will be discussed, and some concluding remarks and a brief description of the research limitations will finish off this paper.

6

2. Literature review

An extensive amount of literature can be found on CEO compensation, clarified by a range of underlying reasoning and approaches. Notable is the limited amount of literature on the concepts of greed and envy related to CEO compensation. Therefore, this thesis tries to shed light on the concepts of envy and greed, capturing differences in compensation within section 2.1. The established degrees of CEO envy and greed will then, in section 2.2, be applied to study the effect a certain degree of CEO envy or greed might have on shareholder value.

2.1 Prior literature on Envy and Greed

Conform to the definitions revealed in the introduction, Seuntjens et al. (2014), summarize envy as being driven by external factors such as wanting what others have, while greed as being driven by internal motivations such as wanting more.

2.1.1. Literature on Envy

Envy, the desire for another’s well-being, in the business sector can best be described as aspiring

to have what your closest competition has. This involves continuous benchmarking, resulting in personal satisfaction when receiving equal or higher compensation, compared to the closest competition. Earlier academic research (Albuquerque et al., 2012; Bizjak et al., 2008 & 2011; Cummenerl et al., 2015; and Faulkender and Yang , 2010) supports that benchmarking CEO compensation to comparable peer groups is an accepted accounting proxy for envy. More specifically, similar to the ideas of Cummenerl et al. (2015) who use the difference between the pay of two competing managers as a proxy for envy, this paper will quantify the degree of envy by comparing the compensation of a single CEO to their prior years peer group mean.

Shortly addressed in the introduction, there is no uniform degree of envy (Endriss et al., 2003). When one person desires another’s advantages, Smith et al. (1999) in their research speak of envy. Based on Cummenerl et al. (2015), some CEOs have been found to be more susceptible to another’s well-being than others. Therefore, in order to distinguish in the degree of envy, a CEO not showing any interest in a comparison in salary to their industry peers, earning less than

7 their industry peers, will be distinguished to have a low level of envy. Vice versa, a person reflecting and adapting his own salary to the mean salary of his peers, and earning similar amounts of salary or higher, will be categorized to have a high level of envy.

The important economic role of selecting suited compensation peer groups in the pay-setting process is highlighted by Bizjak et al. (2011). Properly structured, peer groups can be valuable sources for determining relevant market wages for CEOs, and can provide information regarding suitable compensation-levels necessary to both retain and motivate top executives. Criticism on benchmarking and peer group choices reveal skepticism regarding CEOs and co-opted boards opportunistically choosing peer firms in a way that will inflate CEO pay (Bebchuk and Fried, 2004). The results of Bizjak et al. (2011), regarding peer group choices, also support this trend towards an opportunistic behavior in the selection of peers, favoring larger and better performing firms.

In pursuit of restricting this opportunistic behavior, the peer group constellation of a certain industry will be selected (Bizjak et al., 2011; and Crummenerl et al., 2015). A crucial choice, since compensation programs are often being designed to be competitive within identical industries sharing common labor market characteristics.

2.1.2 Literature on Greed

Greed, the desire for an active pursuit of excessive material wealth (Haynes et al., 2014), in the business sector can best be described as aspiring to be wealthier than the rest of your proximate colleagues. Greed is broader than just the desire for material possession (Tickle, 2004). Although it can be strongly felt for food and power, it can just as well be felt for things having no connection with success or status (Richins, 2004). Due to the accountancy focus of this research, other psychological explanatory factors will be disregarded and the statistical research will purely distinguish in quantifiably different levels of greed. The same holds true for how this paper examines envy.

Converting Bebchuk et al. (2011) ideas on CEO pay slice differences into a measurement for CEO greed, a quantitative measure can be derived from a comparison between the compensation of the CEO and the related top executives. By selecting the CEO and executives from the same firm, any firm specific characteristics that effects the average level of compensation are

8 controlled for. The resulting difference in compensation per firm will then act as a proxy for the degree of greed: a high compensation difference corresponding to a high degree of greed and vice versa. In line with the results of Haynes et al. (2015), the comparison to top executives is based on the idea that this group of employees is in the actual position to pursue extraordinary wealth. In other words, while employees on all organizational levels might have the desire for extraordinary wealth, only a few are in the position to actually realize it.

Similar to envy, there is no uniform degree of greed. When self-interest reaches a tipping point and becomes too large, Haynes at al. (2015) speak of greed. Some CEOs have been found to direct more of the firm’s resources towards themselves. Therefore, a high degree of greed, in this research, will be measured by a high difference in compensation and vice versa. (Haynes et al., 2015).

Lastly, selecting a certain industry will allow for a more reliable quantification of the degree of greed (Haynes et al., 2015; Miller & Dröge, 1986). This is in line with greed being more noticeable in firms with short-term time horizon aiming to become public firms in a few years, and/or seeking rapid growth, compared to family firms. Since family firms have a multi-generational time horizon, greed is less pronounced compared to rapidly growing public firms. Ultimately, selecting public firms within a fast growing and volatile industry, CEO greed has the highest probability to be detected (Haynes et al., 2015).

2.1.3. Literature on the chosen industry for envy and greed

The industry selected in this research is the mining industry. In the United States, the mining of

minerals or other geological materials from the earth, has always played an important part in the well-being of the citizens. Measured in 2011, the mining sector added a value of nearly $290 billion and employed around 727 thousands of people (Mining Global magazines, 2016; Statista, 2016₎. Next to this, the gross output2 _{of the mining industry in the United States has increased} over the years (1998 – 2014). When concentrating on the time-scheme of this research (2004 - , the mining industry has shown to grow rapidly until the year 2008, when it experienced a 2014)

2 _{The gross output has been chosen to evaluate the growth of the industry, since it measures “the total economic} activity in the production of new goods and services in an accounting period” (Statistics, 2016).This means that this measure of growth is a much broader measure of the economy compared to GDP which focuses mainly on the final outcome in goods and services (Statistics, 2016).

9 significant drop in its growth. Illustration 2 (see Appendix 8.1) demonstrates this growth over the years (Statista, 2016). The sudden growth drop in 2008 can most likely be explained by the economic crisis (2007-2011), which has essentially negatively influenced every industry from 2007 until 2011 (Duncan, J., 2009). Weakening the global economy, it negatively influenced the raw material prices of the mining industry, where resources are driven by industrial demand (e.g. oil, gas, base metals and precious metals).

These fluctuations in demand, facilitating large movements in industry profits, have caused changes to the compensation of CEOs (Bertrand, M., & Mullainathan S., 2011; Mining Global Magazine, 2016). In nearly all of the years in which this research was conducted, compensation showed to be correlated with oil prices, either moving up or down. Only in a few value decreasing cases did this correlation not meet the expectations, and it was concluded that “while CEOs are always rewarded for good luck, they may not always be punished for bad luck” (Bertrand, M., & Mullainathan S., 2011). Most likely, this will be the case for small decreases in commodity prices, but a correlation will hold true for larger circumstances such as the impact of the economic crisis. All combined, the mining industry remains a lucrative industry, not only for its economic character, but also for its leaders. Limited guidance on director oversight (Kellerman, J., & Engelstad. N., 2008), allows CEOs to make their own strategic decisions.

2.2 Agency theory with a focus on the concepts of envy and greed

The second part of this literature study will focus on the agency theory (Jensen and Meckling, 1976; Fama, 1980; Fama and Jensen, 1983). Shedding light on the moral hazard problems that may arise when aligning the interests of CEOs with the ones of the shareholders, the agency theory provides valuable concepts with regard to CEO envy and greed, and the corresponding consequences on shareholder value.

2.2.1 The concepts of the Agency theory / Shareholder value

The Agency theory is concerned with the relationship between two parties: the principal and the

agent. The agency relationship considered in this research is the one between the agent, relating to the CEO of a firm, and the principals, corresponding to the shareholders of the firm. By delegating the authority to make decisions on behalf of the shareholders to the CEOs, agency

10 problems can arise due to incomplete information and inefficiencies (Jensen and Meckling, 1976; Fama, 1980; Fama and Jensen, 1983).

Shareholder value has been defined as the present value of expected future cash flows , which

have been discounted at a rate that appropriately reflect the relevant risk (McCarthy et al., 2004; Investopedia, 2016). In the setting of this research, it is the value delivered to shareholders because of the CEO's ability to grow earnings, dividends and share price. In other words, shareholder value is the sum of all generated strategic decisions that have an effect on the firm’s ability to increase free cash flow over time.

How can CEOs create shareholder value? Shareholder value can be created when decisions are made that increase return from existing assets, make new investments with rates of return above the investment’s cost of capital, sell assets that are not profitable for the firm, and return cash to investors by means of dividends when profitable investments are not available (McCarthy et al., 2004; Rappaport A., 2016). Understanding where value can be created or destroyed can allow for more efficient decision making and allocation of assets. A CEO capable of identifying these future value generating opportunities and investing firm capital wisely and timely will generate shareholder value (McCarthy et al., 2004).

In order to make sure that CEOs are making decision that are in the best interest of the firm, shareholders design the compensation package of the CEOs in a way that increases the CEOs incentives to maximize the firm value and resulting shareholder value (Bertrand & Mullainathan, 2001). Results by Edmans and Gabaiz (2009) on optimal contracting theories present that performance related pay is best used to align managers’ incentives with those of shareholders. The question remaining is whether performance-related pay works in aligning incentives when managers are envious or greedy? More specifically, do envious or greedy CEOs, well paid for their performance, really act in the best interest of their shareholders?

The key dilemma in this investigation is known as ‘moral hazard’ problem within the agency theory, implying that the interest of the agent and principals are not always aligned. Therefore, the efficient alignment of interests and the resulting CEO compensation design are important

11 subjects when studying the influence of CEO envy/greed on shareholder value. (Fama, 1980; Fama and Jensen, 1983; Frydman and Jenter, 2010; Jensen and Meckling, 1976).

12

3. Hypothesis development

3.1 Hypothesis development for Envy

According to the agency theory, the incentives of the CEO must be aligned with that of his shareholders by means of a suitable compensation policy. But what if the alignment of incentives through compensation contracts is less effective when CEOs are more prickled by external motivations, wanting what their market peers have? Based on earlier research by Crummenerl et al (2015), a high degree of envy is expected to result in a divergence between the incentives of shareholders and CEOs. The authors reported a more aggressive attitude in line with an increasing degree of CEO envy. This was determined by means of a higher tendency towards overinvestments and seeking economically unreasonable mergers (Goel and Thakor, 2005 & 2010). As a result, shareholder investments become more risky and speculative, decreasing shareholder wealth when CEO envy is high.

Besides, CEOs with higher levels of envy have been found to damage the competition when increasing their own effort levels. This phenomena is reflected by envious CEOs intentionally increasing their production levels to drive down market prices (Crummenerl et al., 2015). While this might be beneficial for the generation of earnings in the short run, reflecting a possible performance indicator for the compensation policy of the CEOs (Jensen and Murphy, 1990), this behavior will be value decreasing in the long run. Decreased market prices, next to a more fragile industry, will lower the share valuations of the industry, causing shareholder investments to decline (Nguyen, J., 2016). CEOs being incentivized by wanting what their competition has or more, by even making their competing peers worse off on their way to accomplishing their goal, will lose sight of the incentives of their shareholders, decreasing their wealth.

Therefore, it is predicted that:

H01: Envious CEOs do not generate less shareholder value than less envious CEOs.

13

3.2 Hypothesis development for Greed

According to the moral hazard within the agency theory, CEOs must be motivated through suitable compensation policies to align the CEO’s interest with the ones of the shareholders

(Jensen and Meckling, 1976; Fama, 1980; Fama and Jensen, 1983). Since greedy CEOs are found to be driven by internal motivations (Seuntjens et al., 2014), the alignment of the financial interest of the CEOs with the owners should be accomplished when efficient compensation and reward systems are established (Haynes et al., 2015). Therefore, whenever a CEO is greedy, i.e. motivated to increase his own welfare, the CEO will be forced to take decisions that will also maximize shareholder value. Yet, when incentives are not aligned, monitoring costs will offset the positive relationship between CEO wealth and shareholder value. Therefore, to achieve the optimal alignment of incentives, the compensation design of the CEO is an important subject within managerial and financial research (Frydman and Jenter, 2010; Hill and Phan, 1991; Agrawal and Knoeber, 1996).

Furthermore, the competition within the managerial labour market next to the structure of the CEOs compensation contracts have been found to be another explanatory variable when aligning the incentives of CEOs and their shareholders (Sridharan, 1996).

Concluding, a more greedy CEO is predicted to be value enhancing for shareholders since his actions to maximize his own wealth, will enhance shareholder value when incentives are aligned effectively. The corresponding hypothesis will be:

H02: Greedy CEOs do not generate more shareholder value than less greedy CEOs.

14

4. Research method

This section introduces the variables together with the sample, and clarifies how this research is conducted. Besides, the regression model and the reliability and validity issues with regard to a robustness test are elaborated.

4.1 Variables and sample

4.1.1 Variables and sample used to quantify greed and envy

Data used in this investigation has been obtained fromWharton Research Data Services (WRDS). To quantify the degree of envy and greed (Section 4.2), annual compensation data obtained via Capital IQ - Compustat executive compensation from the North American mining industry was collected using the Standard Industrial Classification Code (SIC code: 1000-1499). Next to this, data on the salary, the fiscal year, the CEO description and the company ID were selected (summarized in table 1).

The total sample for the quantification of envy and greed (Section 4.2) is based on 113 firms from the North American mining industry, resulting in 758 CEO observations spread over the fiscal years 2004-2014 for envy and the fiscal years 2003-2014 for greed. Data for the fiscal years 2015 and 2016 was either incomplete or not available.

4.1.1.1 Validity of the data

A validity check was executed on behalf of the data retrieved from WRDS. This was done by comparing the SIC code with the SIC description, confirming the collection of solely mining industry data, and by verifying that the executives were correctly stated as CEO in the sample by comparison to the executive description.

15

Table 1: Variable overview

WRDS variable name Variable description

Salary The dollar value of the base salary (cash and non-cash) earned by the named executive officer during the fiscal year (value given in thousands of dollars). SIC code Standard Industrial Classification Code. In this research, SIC code 1000-1499 is

applied, which corresponds to the mining industry SIC description A description of the kind of mining industry

Current CEO This is a variable which points out all the executives that were titled CEO in the corresponding fiscal year. Using this variable, CEOs and non-CEOs could be codified (dummy 1: CEO, dummy 0: non-CEO)

Most recent title Current description of the function of the executive

YEAR-fiscal year The fiscal year

Company ID or Gvkey The company code

4.1.2 Variables and sample used to measure the effect on shareholder value

To measure the effect a certain degree of envy or greed has on the change in shareholder value, data on annual compensation and annual fundamentals was collected from WRDS – Capital IQ - Compustat. All variables and subsequent calculations are presented in table 2.

Due to missing values in the control variables, the total sample remaining for analysis (section 4.2.3) decreased to 60 firms . This resulted in 184 CEO observations spread over the fiscal years 2003-2014 .

4.1.2.1 Reliability and validity of the variables

Next to the verification of the base variables (SIC code and current CEO) , firm leverage was established by two different formulas to check for the consistency in the results. Please refer to section 4.2.3.2 regarding the regression models, and the robustness test in section 5.3.3.5.

16

Table 2: Overview of the variables

Variable type WRDS variable Variable description

Dependent CEQ The total common/ordinary equity. This variable measures the

total common shareholder interest in the company.

Independent Salary The dollar value of the base salary (cash and non-cash) earned by the named executive officer during the fiscal year (value given in thousands of dollars).

Base SIC code Standard Industrial Classification Code. In this research, SIC

code 1000-1499 is applied, which corresponds to the mining industry

Base YEAR-fiscal The fiscal year

Base Current CEO This is a variable which points out all the executives that were titled CEO in the corresponding fiscal year. (Codified into dummy

variables: dummy 1: CEO and dummy 0: non-CEO)

Base Company ID The company code

Control Bonus $ The bonus received by the executives

Control Option awards num The options granted to the executives

Control AT Total assets. Measures the size of the company

DLTT Total long term debt. Represents the debt obligations which are due more than one year from the company's balance sheet date

Control (CEQ+DLTT)/AT Firm performance (computed from WRDS data)

Control DLTT/CEQ Firm leverage 1 (computed from WRDS data)

Control DLTT/AT Firm leverage 2 (computed from WRDS data)

4.1.3 Information on the chosen time span and continent of interest

All data obtained from WRDS is based on publicly listed North American firms. The choice for the United states is clarified by the extensive information disclosure practices, mandatory and voluntary. Besides, North America has a large variety of regions offering different landscapes, possessing a vast diversity of raw resources, subtracted by means of mining. Therefore, the richness of data available on WRDS allows for an extensive database, spread over a period of eleven years (2004-2014) . This chosen period is especially interesting since it includes the

17 economic financial crisis which took place from 2007 until 2011. Being in the possession of a complete dataset of such a long time span allows for a comprehensive study on envy and greed, and the effect the different degrees of each CEO characteristic have on shareholder value before, during and after the economic crisis.

4.2 Methodology

4.2.1 Quantifying CEO envy

Envy will be quantified by comparing the salary received by each individual CEO, per fiscal year , with the mean salary received by the CEO peers in the prior fiscal year . The mean peer

(t) (t-1)

compensation from the prior year (t-1) has been chosen for this research method, since CEOs have been found to compare themselves to the prior results of their peers (Bizjak et al., 2011). In line with research by Endriss et al., (2003), four different percentiles are established to distinguish in the degree of envy. The establishment of the borderlines within which these percentiles fall have been inspired through research by Albuquerque et al (2013); Franco et al., (2011); Cadmen & Carter (2009). According to their findings, CEOs are opportunistically selecting firms in the peer group in order to inflate their pay, and most of the time when CEOs compare themselves to their peers, the composition of the peer group will deviate from the peer group that would have been chosen purely based on the economic criteria. Thus, when applying this theory to this research, the CEOs that generate a salary which is extremely below the mean salary is coded one. A higher level of envy will be reached when a CEO receives an excessively higher compensation than would be reasonable compared to his peers. This CEO will be categorized as falling within code four. Descriptions of the quartiles for envy have been summarized in table 3. All quantiles have been applied numerically to the CEOs in this research, since the degree of envy, or greed, a person possesses may be a sensitive subject. Additionally, no names have been revealed in relation to the results.

18

Table 3: Numerical code descriptions for envy

Numerical code* Degree of envy Description

1 Extremely limited CEO salary extremely below the mean salary of CEO peers

2 Limited CEO salary slightly beneath or equal to the mean

salary of CEO peers

3 High CEO salary equal or slightly above the mean

salary of CEO peers

4 Extremely high CEO salary extremely above the mean salary of

CEO peers

* The numerical codes have been distributed according to percentile divisions (demonstrated in section 5.1)

4.2.2 Quantifying CEO greed

Quantifying greed will be realized by comparing the salary received by the CEO of each chosen company with the mean salary received by the remaining executives of that firm. This unique method of quantifying greed has partly been derived from the inspiring ideas of earlier research by Bebchuk et al., (2011), who investigated CEO pay slice in the context of CEO wealth maximization. By comparing the CEOs salary with the average salary received by his closest executives, a categorization in the degree of CEO greed is established.

When distinguishing in the degrees of greed by creating percentiles, it is assumed that CEOs that receive remarkably higher salaries compared to their executives are more greedy than CEOs that differ in pay slice only little or not at all from their executives. Since earlier research confirms that CEOs are not uniformly greedy (Haynes et al, 2015), the four percentiles of greed have been coded from having an extremely limited degree of greed (code 1) until having an extremely high degree of greed (code 4). The four different percentiles have been displayed in table 4.

19

Table 4: Numerical code descriptions for greed Numerical

code

Degree of greed Description

1 Extremely limited CEO salary extremely below the mean salary of executives 2 Limited CEO salary slightly beneath or equal to the mean salary of

executives

3 High CEO salary equal or slightly above the mean salary of

executives

4 Extremely high CEO salary extremely above the mean salary of executives * The numerical codes have been distributed according to percentile divisions (demonstrated in section 5.2)

4.2.3 Measuring the effect of CEO envy/greed on shareholder value.

The linear regression, performed to measure the effect CEO greed/envy has on shareholder value, relates to the dependent variable, shareholder value, and to the independent variable, salary. Following earlier literature in regard to compensation policies within the agency theory, appropriate control variables have been added (Jensen, M.C. & Murphy, K.J., 1990).

4.2.3.1 Regression model

𝑪𝑬𝑸_𝒊= 𝜶 + 𝜷𝟏(𝑺𝒂𝒍𝒂𝒓𝒚_𝒊) + 𝜷𝟐(𝑪𝒐𝒏𝒕𝒓𝒐𝒍 𝒔𝒆𝒕 𝟏_𝒊) + 𝜺_𝒊

where subscript i equals to the index firm, and

CEQ = the change in total common shareholder interest in the company calculated over one year

Salary = the change in salary received by a certain CEO, the difference in salary in relation to his industry peers or firm executives, which is calculated over one year. The difference in salary of a CEO is adopted as a measure to identify a CEO to fall within a certain group of greed/envy

Control set 1 = the control variables have been bundled in the model, corresponding to: Bonus = the change in bonus received by the CEO measured over one year. Options = the change in options awarded to the CEOs measured over one year Size = the logarithm of total assets – measuring the change in firm size Performance = the change in firm performance measured over one year

((CEQ+DLTT)/AT)

20

4.2.3.2 Robustness model

For the robustness test, the dependent and independent variable within the regression model remain unchanged. The change applied to the regression model relates to the control variable; firm leverage (leverage 1: DLTT/CEQ). To test for the robustness of the variables within the regression model, firm leverage has been computed by a different formula (leverage 2: DLTT/AT). This reliability test has only been executed for one of the variables, regarded sufficient, since most of the control variables have been taken directly from the database WRDS. The regression model follows:

𝑪𝑬𝑸_𝒊= 𝜶 + 𝜷𝟏(𝑺𝒂𝒍𝒂𝒓𝒚_𝒊) + 𝜷𝟐(𝑪𝒐𝒏𝒕𝒓𝒐𝒍 𝒔𝒆𝒕 𝟐_𝒊) + 𝜺_𝒊

where subscript i equals to the index firm, and

CEQ = the change in total common shareholder interest in the company calculated over one year.

Salary = the change in salary received by a certain CEO, the difference in salary in relation to his industry peers or firm executives, which is calculated over one year. The difference in salary of a CEO is adopted as a measure to identify a CEO to fall within a certain group of greed/envy

Control set 2 = the control variables have been bundled in the model, corresponding to: Bonus = the change in bonus received by the CEO measured over one year. Options = the change in options awarded to the CEOs measured over one year Size = the logarithm of total assets – measuring the change in firm size Performance = the change in firm performance measured over one year

((CEQ+DLTT)/AT)

21

5. Results

This section reveals the results of the quantification of envy and greed, next to the results of the descriptive statistics, multicollinearity and regression analysis. Intermediate interpretations of the results have been added to each section to analyze the findings. The final discussion of this thesis entails a summation of these more detailed, separate interpretations.

5.1. The quantification of Envy

5.1.1. Establishment of the percentiles to distinguish in the degrees of envy

All CEOs have been categorized to fall within one of the four groups, established according to the percentile divisions:

Code 1: CEOs that earn less than 20% of last year’s mean peer compensation. Code 2: CEOs that earn up to 38% of last year’s mean peer compensation. Code 3: CEOs that earn up to 80% of last year’s mean peer compensation. Code 4: CEOs that earn above 80% of last year’s mean peer compensation.

Changing market circumstances are also reflected in the differences in absolute values ascribed to each group (1-4) demonstrated in table 5. The applied percentile division has been selected carefully to account for these changes over time.

Table 5: Absolute values ascribed to each code (2004-2014)

Code 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014

1 -92.96 -63.13 -147.73 -206.62 -145.83 -181.82 -184.40 -136.24 -170.88 -183.36 -188.24

2 - - - -4.07 - -7.76 - - - -33.36 -

3 156.25 26.66 85.03 19.29 126.04 128.11 202.39 150.23 258.02 261.71 227.03

4

22

5.1.2 The four percentiles of envy

To assist in the interpretation of the results, the number of CEOs assigned to each percentile (code 1-4) are displayed as a percentage of the total sample in each year. For the numerical distribution across the sample, please refer to Illustration 3 (Appendix 8.1).

Graph 1: The CEO sample distribution across the four percentiles over time.

5.1.3 Interpretation of the results on the quantification of envy

The CEO sample is found to be distributed almost evenly among less envious (code 1+2) and envious CEOs (code 3+4), at approximately 55% to 45% respectively, and fluctuate only slightly over the years. A minor increase in the number of envious CEOs, and a subsequent decrease in the number of less envious CEOs, can be observed during the years of the economic crisis (graph 1). This slight increase in the number of CEOs that fall within the envy group might be explained by CEOs feeling a stronger drive to compare their compensation to their more wealthy market peers in more turbulent years and when exposed to more risk, compared to more stable years.

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 1 2 3 4

23

5.2 The quantification of Greed

5.2.1 Establishment of the percentiles to distinguish in the degrees of greed

All CEOs have been categorized to fall within one of the four groups, established according to the following percentile divisions:

Code 1: CEOs that earn less than 5% of the mean compensation difference. Code 2: CEOs that earn up to 50% of the mean compensation difference. Code 3: CEOs that earn up to 80% of the mean compensation difference. Code 4: CEOs that earn above 80% of the mean compensation difference.

Table 6 shows the absolute values as a result of the chosen percentile division, where changes over time reflect changing market circumstances. It can be observed that the differences in compensation of the CEOs compared to the executives have become larger, causing greater absolute and relative differences.

Table 6: Absolute values ascribed to each code (2003-2014)

Code 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014

1 -106.92 -105.75 -63.64 -55.19 -104.89 -59.58 -73.49 -42.97 -9.67 -26.00 -39.13 -

2 - - - 7.50 25.69 1.45 61.73 227.50 277.50

3 176,02 191,34 170.14 215.32 230.60 249.67 263.85 306.24 346.16 403.52 462.48 497.32

4

* All values are in thousands of dollars ($)

5.2.2 The four percentiles of greed

To aid in the interpretation of the results, the number of CEOs assigned to each percentile (code 1-4) have been displayed as a percentage of the total sample in each year. For the numerical distribution across the sample, please refer to Illustration 4 (Appendix 8.1).

24

Graph 2: The CEO sample distribution across the four percentiles over time.

5.2.3 Interpretation of the results on the quantification of greed

The CEO sample is found to be distributed unevenly among the less greedy (code 1+2) and greedy CEOs (code 3+4), at approximately 30% to 70% respectively. A rapid increase in the share of greedy CEOs, and a subsequent decrease in the number of less greedy CEOs, can be observed during the years of the economic crisis (graph 2). This might be explained by the economic volatility increasing the CEOs tendency to secure their own wealth, irrespective of the wealth of the firm. Equivalently, this increase in compensation might also be explained by increased effort levels of the CEO trying to pull the company out of the distressed situation and therefore requesting a higher compensation.

5.3 Statistical tests

5.3.1 Descriptive Statistics

This section reveals the results on the descriptive statistics, including data on the total amount of observations, the mean, the standard deviation and corresponding t-tests. As a result of the inclusion of control variables (summarized in table 2), the complete dataset had to be reduced significantly, due to missing values. Additionally, all of the data included in the dataset is in the

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 1 2 3 4

25 format of a percentage change to control for comparability next to firm specific characteristics. Furthermore, the effect of possible outliers has been eliminated by winsorization, and the dataset has been prepared into nine different envy/greed sub-groups. The SPSS output on all of the descriptive statistics, for all the sub-groups, has been included in Appendix 8.2.

The descriptive statistics of the two most extreme groups (code 1 vs. 4) of envy have been summarized in the table 7. Changes in the two means have been compared to each other by the application of a t-test (Lane, D.). The same procedure has been repeated for the comparison of the most extreme groups for greed (table 8). An example calculation has been enclosed in the Appendix 8.1 (Illustration 5).

Table 7: Descriptive statistics on code 1 vs. 4 in regard to envy

Variables Group 1 Group 4 T-test

N Mean S.d. N Mean S.d. T P ΔCEQ 64 0.207 0.333 65 0.232 0.327 0.254 0.800 ΔSalary 64 0.144 0.210 65 0.197 0.451 0.525 0.601 ΔBonus 64 0.305 1.246 65 0.158 1.203 0.751 0.454 ΔOptions 64 0.336 1.578 65 0.121 1.279 1.018 0.311 ΔLeverage 64 3.253 25.178 65 6.206 50.558 2.725 0.007** ΔFirm size 64 7.461 0.851 65 9.259 0.864 11.03 0.000** ΔPerformance 64 0.006 0.079 65 0.003 0.074 0.046 0.963

Δ or ln variables: values are illustrated as a change over time/1year P-Probability: two tailed, and a degree of freedom of 127

** Significant at a 1% level

Table 7 shows the results of the comparison between the two means of the first group (code 1) and the fourth group (code 4) in regard to envy. According to the t-test results, the t-values for the change in firm leverage and firm size are significant at a 1% level (Lane., D.; Field, A.,

26 2009). Both control variables show a significant difference in the mean between the two groups (code 1 vs 4). A comparison of the remaining variables does not deliver significant results.

5.3.1.1 Interpretations of the t-tests for envy

A higher mean firm leverage, together with a larger mean firm size, have been identified for CEOs positioned within code 4 compared to code 1. In line with empirical research, the data confirms that firm leverage is positively related to firm size (Jensen & Meckling, 1976; Kurshev & Strebulaev, 2005). On the contrary, based on prior literature, envious CEOs are expected to be found in smaller and more influential firms (Bizjak et al., 2011; Crummenerl et al., 2015). These results are however not confirmed by the data (table 7), which shows that envious CEOs appear to be positioned within larger firms, on average.

Table 8: Descriptive statistics on group 1 and 4 in regard to greed

Variables Group 1 Group 4 T-test

N Mean S.d. N Mean S.d. T P ΔCEQ 13 0.354 0.555 79 0.225 0.322 0.649 0.518 ΔSalary 13 0.162 0.208 79 0.186 0.350 0.149 0.882 ΔBonus 13 0.254 1.723 79 0.207 1.183 0.129 0.898 ΔOptions 13 0.756 2.481 79 0.074 1.232 1.671 0.868 ΔLeverage 13 0.258 0.869 79 5.105 45.859 3.350 0.001** ΔFirm size 13 7.930 1.189 79 8.709 1.035 2.468 0.016* ΔPerformance 13 0.242 0.080 79 -0.002 0.076 2.914 0.005**

Δ or ln variables: values are illustrated as a change over time/1year P - Probability: two tailed, and a degree of freedom of 90

* Significant at a 5% level ** Significant at a 1% level

Table 8 shows the results of the comparison between the two means of the first group (code 1) and the fourth group (code 4) in regard to greed. According to the t-test results, the t values for the change in firm leverage and firm performance are significant at a 1% level, and the logarithm

27 of firm size is found to be significant at a 5% level (Lane., D.; Field, A., 2009). All three control variables show a significant difference in the mean between the two groups (code 1 vs 4). A comparison of the remaining variables does not deliver significant results.

5.3.1.2 Interpretations of the t-tests for greed

A lower mean firm performance, together with a larger mean for firm size and firm leverage, has been recognized for CEOs positioned within code 4 compared to code 1. In line with the interpretation of the t-tests on envy, the data on greed also shows that firm leverage and firm size are positively related (Jensen & Meckling, 1976; Kurshev & Strebulaev, 2005). On the contrary, the data (table 8) reveals that greedy CEOs are found in larger mean sized firms. This is not in line with prior research that found the effect of greed to be more immediate in smaller firms (Haynes et al., 2015; Miller & Dröge, 1986). Next to this, the decrease in the mean firm performance is not in line with the expectancies developed according to the agency theory (section 3.2). This might be explained by a more excessive focus of the CEO on immediate benefits (Haynes et al., 2014). Neglecting long term interest and the consequences this might have on the performance of the firm, greedy CEOs might be more tempted to extract excess rents from the firm for their own material gain (Seuntjens et al., 2015).

5.3.2 Multicollinearity

To check for the multicollinearity between the variables, a Pearson correlation matrix has been generated (table 9).

Table 9: Pearson Correlation matrix of the entire dataset

Variables Δ CEQ Δ Salary Δ Bonus Δ Options Δ Leverage Δ Size Δ Performance

Δ CEQ 1 .141 .061 -.136 -.080 .021 .180* Δ Salary .141 1 -.019 -.061 -.016 .147* .015 Δ Bonus .061 -.019 1 .046 .008 -.081 .021 Δ Options -.136 -.061 .046 1 -.007 -.011 -.013 Δ Leverage -.080 -.016 .008 -.007 1 -.040 -.080 Δ Size .021 .147* -.081 -.011 -.040 1 -.012 Δ Performance .180* .015 .021 -.013 -.080 -.012 1

* Correlation is significant at a 5% level (2-tailed) N: 183 observations

28 The table includes SPSS output of the correlation coefficients for all seven variables (summarized in table 2). Multicollinearity can cause large concerns when the correlation matrixes include Pearson correlation values that are higher than 0,70 and significant at a 5% level (Cohen, J., 1988; Field, A., 2009). No significant correlations have been found between the different variables. This means no interrelations exist that could cause any distortion to the regression analysis.

5.3.3 Regression results

Regression analyses have been performed to test the H1 and H2. The regression results for the two most extreme groups of both envy and greed have been displayed below (table 10 and table 11 respectively). The complete SPSS data with regard to the regression models can be found in

Appendices 8.2.

Furthermore, Z-scores have been calculated from the data to compare the difference between two regression coefficients across the independent samples (group 1 and group 4). This statistical test, for the equality of regression coefficients, has been established according to the ideas of Paternoster et al., (1998) & Clogg et al., (1995). The Z-score formula has been included in the Appendix 8.1 (Illustration 6).

5.3.3.1 Regression results on envy

Table 10: Regression model on the two most extreme groups (group one and four) on Envy

Variables Group 1 Group 4 1 vs. 4

B Std. Error t P B Std. error t P Z (Constant) .484 .363 1.335 .187 -.109 .465 -.234 .816 1.005 Δ Salary .034 .196 .174 .862 .133 .093 1.438 .156 -.456 Δ Bonus .043 .034 1.294 .201 -.018 .039 -.470 .640 1.179 Δ Options -.022 .027 -.839 .405 -.007 .035 -.197 .845 -.339 Δ Leverage -.00001 .002 -.006 .995 -.001 .001 -.916 .364 .044 Δ Firm size -.040 .049 -.814 .419 .035 .050 .700 .487 -1.071 Δ Performance 1.330 .521 2.554 .013** .262 .575 .456 .650 1.376 ** Significant at a 5% level. a. Dependent variable: CEQ

29 The regression results on envy (table 10) show that the effect of salary, for both groups, on CEQ is not significant, reporting a p-value of 0.862 and 0.156 respectively. This means that the H1 has to be rejected. Consequently, the H01 is accepted, since there is no significant relationship that would justify envious CEOs to generate less shareholder value than less envious CEOs. Contrarily, the control variable firm performance is found to have a significant effect on CEQ within group one, a p-value of 0.013.

5.3.3.2 Interpretations of the regression results concerning envy

Assuming that the influence of the independent variable (Δ Salary) is significant, CEOs within group four would have had a larger influence on the dependent variable (Δ CEQ) compared to group one. This would mean that envious CEOs would generate more shareholder value compared to less envious CEOs. Although not significant, these are interesting results, which could have provided proof for the agency theory to work efficiently by means of incentive alignment, even when CEOs are more externally focused in their wealth increasing practices. Also remarkable in regard to the regression results is that the influence of firm performance on the shareholder value is significant for group one. This might also be explained by the agency theory, given that CEOs may compare their own salary to the performance of the firm, and not to the compensation received by other peers.

5.3.3.3 Regression results on greed

Table 11: Regression model on the two most extreme groups (group one and four) on Greed

Variables Group 1 Group 4 1 vs. 4

B Std. Error t P B Std. Error t P Z (Constant) .045 .963 .047 .964 .494 .319 1.548 .126 -.443 ΔSalary 1.709 .846 2.020 .090* .131 .105 1.247 .216 1.851 ΔBonus .020 .070 .281 .788 -.007 .033 -.220 .827 .349 ΔOptions -.056 .043 -1.314 .237 -.021 .031 -.659 .512 -.660 ΔLeverage -.212 .133 -1.594 .162 -.001 .001 -1.016 .313 -1.602 ΔFirm size .012 .135 .087 .933 -.033 .036 -.897 .373 .322 ΔPerformance 1.198 1.390 .862 .422 .763 .484 1.577 .119 .296 * Significant at a 10% level. a. Dependent variable: CEQ

30 The regression results on greed show that the coefficient salary has a significant effect on shareholder value (1.709, p = 0,090) . This relation is observed within group 1. The additional Z-test also confirms this relation with a positive z-score (1.851). Yet, the coefficient salary within group four (0.131, p = 0.216) is not significant and is much smaller compared to the coefficient of group one. As a consequence, the H2 has to be rejected. This means, the H02 is accepted, since there is no significant relationship that would justify greedy CEOs to generate more shareholder value than less greedy CEOs, nor does the data support the opposite relationship.

5.3.3.4 Interpretations of the regression results concerning greed

Based on the assumption that both salary coefficients have a significant effect on shareholder value, CEOs within group one would have a larger influence on shareholder value compared to group four CEOs. These observations are extremely interesting, since this would mean that less greedy CEOs generate more shareholder value than greedy CEOs. Regression results which would not be in line with what has been hypothesized according to the agency theory. This might be explained by the alignment of incentives between CEOs and shareholders, to become less in line with each other as the degree of greed increases. Considering the entire regression output on greed (Appendix 8.2), it can be observed that only the lower degrees of greed (code 1+2) show a significant relation of the change in salary to the change in CEQ, while the two higher degrees of greed (code 3+4) do not present significant results. Is it possible that CEO effort becomes more distressed at a certain degree of greed? A degree at which CEOs are already being paid for the decisions they make. This would mean that increased effort levels of the CEO would no longer be reflected in his salary, and consequently will not affect the shareholder value. Since no significant data has been found, these questions remain unanswered, and the agency theory remains reinforced.

5.3.3.5 Robustness test

Firm leverage was calculated in two different manners in order to control for the robustness of the establishment of the variable and the influence on the regression model. This was done in order to verify the correctness and appropriateness of the measurement of firm leverage in the standard model. The repeated statistical observations, exchanging leverage 1 (LTD/CEQ) with leverage 2

31 (LTD/TA) resulted in no significant differences on the model. Therefore, the robustness of the regression model including firm leverage 1 is assumed. The results are unreported and are available on request.

32

6. Discussion and Conclusion

6.1 Discussion and Conclusion

The results on the first part of this research, the quantification of envy and greed, showed that the degree of both characteristics are correlated to the market circumstances in which the CEOs in question are operating. As such, it has been found that the degree of greed, in the North American mining industry, has increased during times of recession. This has been interpreted as the CEOs tendency to secure their own wealth, irrespective of the wealth of the firm, or as increasing his effort levels in his attempt to pull the company out of the distressed situation, requesting higher compensation levels. For envy, this correlation was less visible, and the small increase may be interpreted as a stronger CEO drive to compare their compensation to their market peers in more turbulent years and when exposed to more risk.

The second part of this research examined the research question of this thesis: Envy vs greed – whom should a board pick as their CEO to enhance shareholder value. Even though the H1 and H2 had to be rejected, the interpretation of the results allow for interesting observations. The results on envy, although not significant, showed that more envious CEOs generate more shareholder value compared to less envious CEOs. A more extensive data collection might deliver proof on the agency theory working efficiently by means of incentive alignment, even when CEOs are more externally focused on increasing their wealth. The observed results might also be industry dependent. The mining industry is not a very volatile industry, which can be concluded from the continuous demand of raw materials from the economy (Statista, 2016). This stable demand is reflected in the active recovery of the mining industry after the short recession caused by the economic crisis (Mining global magazines, 2016). CEOs need to be adaptive and economically driven, since limited guidance on director oversight within the mining industry (Kellerman, J., & Engelstad. N., 2008) requires more independent strategic decision making. After all, when successfully aligning their strategic incentives with the incentives of their shareholders, it can be concluded that a larger external focus of CEOs does not necessarily need

33 to influence shareholder value negatively, as long as the alignment of incentives takes place effectively. The agency theory remains reinforced.

With regard to greed, the regression results of salary on CEQ are significant for CEOs with low degrees of greed. This correlation was found to be insignificant and much smaller for greedy CEOs. Tremendously interesting, since these findings challenge the concept of the agency theory at a high level of greed. While the results significantly show that the agency theory is reinforced within companies managed by less greedy CEOs, a high degree of greed causes this relationship to lose value. For greedy CEOs, compensation is no longer in line with their efforts, reflected by a lower and insignificant effect on shareholder value. According to the literature on the mining industry, the limited degree of oversight within the industry allows for CEOs to make their own strategic decisions. Interested in their own wealth maximization, the incentives of greedy CEOs within the mining industry might be more problematic to bring in line with those of the shareholders, challenging the agency theory.

Summarized, within the mining industry, less greedy CEOs appear to be more value enhancing for shareholders compared to greedy CEOs. For envious CEOs, these preferences cannot be outspoken due to the statistical insignificance of the results.

6.2 Research limitations and future research

Though the newly devised methodology for this research paper has been firmly based on existing literature and previous research, the proposed relationship between CEO envy or greed and shareholder value had not been investigated before. Given the elaborate theoretical framework supporting the methodology of this paper, the rejection of the main hypotheses and lack of statistical significance is more likely to be the result of the data sample scope and size.

The selection of the North American mining industry was a strategic decision, which allowed for the establishment of the two methods for the quantification of CEO envy and greed. While allowing for the comparison of results and clarity of the market, the selection of this particular industry limited the generalizability of the results. Accordingly, now that the basic principles for this type of study have been established, the methods could be applied to a wider geographical scope, and need to be tested on different industries. Additionally, the choice of multiple industries and more geographically dispersed data will also control for insufficiencies in the sample size.

34 For the purpose of this specific study, the sample size used for the second part of the research was extensively reduced to include the various control variables, which had many missing values.

As a final thought, the findings presented in this paper contain valuable suggestions for future research, which include but are not limited to an investigation of how the varying degrees of CEO greed and envy impact the relationship between these personality traits and shareholder value across industries and geographies. It would be interesting to investigate if certain thresholds exist in the degree to which CEOs can be characterized as greedy or envious, above or below which the relationship between these traits and shareholder value changes.

35

7 References

Agrawal, A., & Knoeber, C. (1996). Firm performance and mechanisms to control agency theory problems between managers and shareholders. The Journal of Financial and Quantitative Analysis. Vol. 31, No. 3, pp. 377-397.

Albuquerque, A. M., Franco, G.D., & Verdi, R. S. (2013). Peer choice in CEO compensation. Journal of Financial Economics. 108 (2013) 160–181.

Bebchuk, L.A., Cremers, M.K., & Peyer, U.C. (2011). The CEO pay slice. Journal of Financial Economics, 102(1), 199-221.

Bebchuk, L.A., & Fried. (2004). Executive Compensation as an Agency Problem. Journal of Economic Perspectives. Vol. 17. Nr.3. 71-92.

Bertrand, M., & Mullainathan S. (2011). Are CEOs rewarded for luck? The ones without principals are. The quarterly journal of Economics. Vol. 116 (3), pp. 901-932.

Bizjak, J.M., Lemmon, M.L., & Naveen L. (2008). Does the use of peer groups contribute to higher pay and less efficient compensation? Journal of Financial Economics. 90, 152–168.

Bizjak, J.M., Lemmon, M.L., & Nguye T. (2011). Are all CEOs above average? An empirical analysis of compensation peer groups and pay design. Journal of Financia Economics. 100. 538–555.

Cadmen, B.D., Carter, M.E., & Hillegeist S.A. (2009). The Incentives of Compensation Consultants and CEO pay. Journal of Accounting and Economics. Vol. 57 Iss 4.

36 Clogg, C., Petkova, E., and Haritou. A. (1995). Statistical methods for comparing regression coefficients between models. American Journal of Sociology 100:1261-1293.

Cohen, J. (1988). Statistical Power Analysis for the Behavioral Sciences, 2nd Edition. Hillsdale: Lawrence Erlbaum.

Crummenerl, M., Doll, T., & Koziol, C. (2015). How to Pay Envious Managers – a Theoretical Analysis. European Financial management. 21( 4), 811–832.

Duncan, J. (2009). The Global Economic Crisis – Worldwide Impact on Mining Companies. KPMG LLP. Available at: http://www.mineafrica.com/documents/macleod%20dixon % 20article.pdf [accessed: 28-05-2016].

The Economist. (2002). “Is greed good?”. The Economist, Special report International Finance.

Edmans. A., & Gabaiz. X. (2009). A Multiplicative Model of Optimal CEO Incentives in Market Equilibrium. The Review of Financial Studies. Vol. 22. Nr. 12.

Endriss, U., Maudet, N., Sadri, F., Toni, F. (2003). Resource Allocation in Egalitarian Agent Societies. Secondes Journées Francophones sur les Modéles Formels Vol. 101-110.104.

Fama, E. F. (1980). Agency Problems and the Theory of the Firm. The journal of political economy. 88 (2), 288-307.

Fama, E.F., & Jensen, M. C. (1983). Separation of ownership and control. Journal of law and economics, 26(2), 301-325.

Faulkender, M., & Yang J. (2010). Inside the black box: The role and composition of compensation peer groups. Journal of Financial Economics. 96. 257–270.

37 Field, A. (2009). Discovering Statistics using SPSS. 3rd Edition. Sage Publications.

Franco de, G., Kothari, S.P., & Verdi, R.S. (2011). The Benefits of Financial Statement Comparability. Journal of Accounting Research. Vol. 49. 895-931.

Frydman. C., & Jenter. D. (2010). CEO Compensation. Annual Review of Financial Economics, Vol. 2 (1). 75-102.

Goel , A. M., & Thakor, A.V. (2005). Green with Envy: Implications for Corporate Investment Distortions. The Journal of Business. 78 (6). 2255-2288.

Haynes, K.T., Campbell, J.T., & Hitt, M. A. (2014). When more is not Enough: Executive Greed and its influence on shareholders wealth. Journal of Management. 1– 30.

Haynes, K.T., Campbell, J.T., & Hitt, M. A. (2015). The Dark Side of Leadership: Towards a Mid-Range Theory of Hubris and Greed in Entrepreneurial Contexts. Journal of Management Studies 52:4.

Hill C., & Phan P. (1991). CEO Tenure as a Determinant of CEO pay. The Academy of Management Journal. Vol. 34, No. 3, pp. 707-717.

Jensen, M.C., & Meckling, W.H. (1976). Theory of the firm: Managerial behavior, agency costs and ownership structure. Journal of financial economics, 3(4), 305-360.

Jensen, M.C., & Murphy, K.J. (1990). Performance pay and top-management incentives. Journal of Political Economy, 98(2), 225-264.

McCarthy, B.F. (2004). Instant gratification or long-term value? A lesson in enhancing shareholder wealth. Journal of Business Strategy, Vol. 25 Iss 4 pp. 10 – 17.