University of Groningen Faculty of Economics and Business
Thesis – Msc Finance Student number: s2077108
Name: Luu Dan-Anh Nguyen
Study program: Msc Finance Supervisor: Dr. R.O.S. Zaal
PAY FOR PERFORMANCE
IN BANKS & IN NON-BANKING CORPORATIONS
PAY FOR PERFORMANCE
IN BANKS & IN NON-BANKING CORPORATIONS
DURING THE CREDIT CRISIS
Abstract: Pay-for-Performance is one of the recommended strategic solutions for the agency
problem in corporate governance, and is measured by Performance sensitivity (or Pay-Performance relationship). The higher Pay-Pay-Performance sensitivity is, the stronger the relation between firm performance and executive remuneration is. Pay-Performance sensitivity has been tested by a large body of research. However, different levels of sensitivity in the banking industry and in non-banking industries have not been widely considered. Therefore, the first objective of this thesis is to examine how Pay-Performance relationships vary in these industries. I find that Pay-Performance relationship is less significant in banks. In other words, Pay-for-Performance is more implemented by corporations in other industries. The second objective of this thesis is to investigate the effect of the current credit crisis on Pay-for-Performance. I find that in non-banking corporations CEOs’ bonuses and stock-based compensation become less sensitive to financial performance after the crisis began in September 2008. However, banks tend to experience no significant change of Pay-Performance sensitivity after 2008.
Keywords: Pay-for-Performance, Pay-Performance sensitivity, Agency Theory, banking
industry, non-banking industries, credit crisis
Table of Content
I. Introduction ... 1
II. Literature Review ... 3
1. Agency Theory and Pay-for-Performance ... 3
1.1 Agency Theory ... 3
1.2 Pay-for-Performance ... 5
2. The Current Financial Crisis and its Effect on Pay-Performance Sensitivity ... 7
2.1 The current financial crisis ... 7
2.2 Effect of the credit crisis on Pay-for-Performance ... 8
a. Non-banking sectors ... 8
b. The banking sector ... 10
III. Methodology ... 11
IV. Variable Description and Sample Selection ... 13
CEO remuneration ... 14
Financial performance ... 14
Dummy for banks ... 15
Dummy for the credit crisis ... 15
V. Results and Discussion ... 19
1. Pay-for-Performance in banks and in non-banking corporations... 19
2. Effect of the Credit Crisis on Pay-for-Performance ... 22
a. Non-banking sectors ... 22
b. The banking sector ... 25
VI. Conclusion ... 27
Appendices ... 30
1
I.
Introduction
The level of CEO compensation remains a controversial topic in business practice as well as from an academic perspective. Discussion in the media about CEO compensation and how executives should be compensated seems to be continual. For example, in April 2014 CNBC headlined the goodbye bonus of US $58 million paid by Yahoo to its COO, Henrique De Castro, after he ineffectively worked in this company for only 15 months; or in November 2013 CNN reported that Oracle shareholders voted against the pay package of the CEO Larry Ellison, which was nearly US $77 million. Generally, large corporations’ executives seem to receive tremendous amounts of money. However, not all prosperous executives do. Warren Buffet, the billionaire CEO of Berkshire Hathaway, received in 2011 only US $100,000 annual salary and did not receive bonus and stock options or grants (Bloomberg, 2011). Why is there such a difference between his compensation and that of, e.g., Henrique De Castro? From a corporate governance point of view, one possible answer is the conflicting interests between shareholders and the manager. It can be said that De Castro has more conflicting interests with shareholders than Buffet does. This is because Buffet is also the largest shareholder of Berkshire Hathaway with 33.10% of outstanding shares and 25.78% of common shares (GuruFocus - Forbes’s contributor, 2009), while De Castro only owned 0.17% of the total shares in Yahoo1. It could be suggested that Buffet relatively concerns more about maximizing shareholders’ value, while De Castro concerns more about his private benefits. This answer is based on the separation of ownership and control, which is discussed more carefully in the literature review section. While the public often finds CEOs’ compensation of millions of dollars unreasonable, analysts in corporate governance find it acceptable as long as CEOs successfully manage their companies. In academic literature of corporate governance this is referred as Pay-for-Performance, a solution for the problem caused by the separation of ownership and control. Pay-for-Performance indicates that CEOs should be paid for their effective effort, which is reflected by firm performance(Jensen and Murphy, 1990).
Another interesting point is that it is important to realize that there might be different degrees of Pay-Performance sensitivity in the banking industry versus non-banking industries. It is also
1 This percentage is calculated given that Yahoo has 1.03 billion shares totally, and De Castro owns 1.7 million
2
important to understand the reasons of this difference. Banks are more regulated and have higher leverage than other firms (John and Qian, 2003). John and Qian claim that these considerable different characteristics make Pay-Performance sensitivity in banks lower than that in non-banking firms. To my best knowledge, this is the only study that compares Pay-Performance correlation of the banking sector with that of another sector. Although many researchers have conducted studies on Pay-for-Performance of banks (e.g. Barro and Barro, 1990; Hubbard and Palia, 1995; Crawford, Ezzell and Miles, 1995) or of non-banking firms (e.g. Jensen and Murphy, 1990; Gilson and Vetsuypens, 1993; Core, Holthausen and Larcker, 1999; Brick, Palmon and Wald, 2006), they do not emphasize the difference between banks and other non-banking corporations. Therefore, the first objective of this thesis is to shed light on this difference by providing a theoretical review and by conducting an empirical analysis.
3
This thesis focuses on the US companies, because the separation of ownership and control in the US corporate governance is more frequently observed than in other countries (Moerland, 1955). Consequently, the problem caused by this separation is more severe in the US system, and Pay-for-Performance is likely to be more important in the US economy than in other economies.
The structure of this thesis is as follows. Section II discusses in more detail about the theoretical and empirical literature on Pay-for-Performance: its origin, how it is being applied in the banking industry and in other industries, and the financial crisis effect on Pay-for-Performance in these industries. Hypotheses are also constructed in this section. Section III provides the methodological approach of the empirical analysis. Section IV describes data collection and statistical characteristics of the final dataset used in the empirical study. Section V presents regression results of the analysis, and discusses these results. Finally, section VI concludes the thesis.
II.
Literature Review
This section is divided into two subsections. Subsection 1 discusses Agency Theory and the reason why Pay-for-Performance is recommended to solve the agency problem. It also provides insight into how Pay-Performance relationship can differ between the banking industry and non-banking industries. Based on this insight, hypothesis 1 is derived. Subsection 2 discusses the origin of the current crisis and its effect on Pay-Performance relationship, based on literature evidence. This subsection also specifies hypotheses 2 and 3 about the crisis effect on Pay-for-Performance, respectively in non-banking industries and in the banking industry.
1. Agency Theory and Pay-for-Performance
1.1 Agency Theory4
controlling the firm. This leads to a separation of ownership and control. This separation, in turn, causes potential conflicts of interests between investors and the manager. It is also stated by Adam Smith (1776) in his well-known book “The Wealth of Nations” that:
The directors of such [joint-stock] companies, however, being the managers rather of other people’s money than of their own, it cannot well be expected, that they should watch over it with the same anxious vigilance with which the partners in a private copartnery frequently watch over their own. Like the stewards of a rich man, they are apt to consider attention to small matters as not for their master’s honor, and very easily give themselves a dispensation from having it. Negligence and profusion, therefore, must always prevail, more or less, in the management of the affairs of such a company.
More specifically, stockholders desire to maximize the return on their investment in the company and seek for long-term stockholder value creation. In contrast, the manager tends to care more about the welfare of his own or of other stakeholders. For example, the manager may keep a loss-creating factory because it is the main workplace for villagers, or he may operate a plant at environmental standards higher than required by the law mandates (Berk, DeMarzo and Harford, 2009, p14).
5
that case the agent is not able to take the advantage of some profitable opportunities either. This results in bonding expenditures. “Residual loss” is the difference between expected returns if investors were able to control the firm without hiring a manager and realized returns when a manager is hired to control the firm. The more ownership and control are separated, the higher the agency cost is (Jensen and Meckling, 1976).
However, the separation of ownership and control is necessary. Shleifer and Vishny (1997) explain that corporations, especially large and public ones, are complicated to manage. Investors, while having a large amount of capital, are not experts in management, and they need the specialized human capital of the manager to generate returns on their capital. The manager, in turn, needs the investors’ funds because he either does not have enough capital to invest or wants to cash out his holdings. Therefore, because of this unavoidable separation within business organizations, it is essential to find solutions to minimize the agency cost it causes.
Dalton et al. (2007) mention two possible actions that investors can take to minimize the probability of misbehavior by the manager. The first one is aligning interests of the manager to their interests. By doing so, the manager’s benefits are hampered if the manager’s decisions hamper shareholders’ benefits. The second action is monitoring the manager via an independent board of directors to ensure that the manager does not misbehave. This thesis focuses on the first approach – alignment of interests.
1.2 Pay-for-Performance
Nyberg et al. (2013) state that stock ownership and performance-based compensation (or so-called Pay-for-Performance) can reduce interest conflict between the manager and investors. This is because the manager’s financial gain is now derived from the financial gain of shareholders. This outcome-contingency links the manager’s preferences and desires with those of shareholders. As a result, the manager is willing to pay attention and effort towards actions that benefit the shareholders. Outcome-based incentives are particularly effective when it is difficult or expensive for investors to monitor a manager’s behavior (Eisenhardt, 1989).
6
empirical research has found consistent results with this implication. For example, Murphy (1985), Coughlan and Schmidt (1985) and Eisenhardt (1989) find that executive compensation is strongly contingent to corporate performance in the US. Kato and Kubo (2006) also find a positive pay-performance relationship in Japan. In Italy although the sensitivity of incentive pay to firm performance is low for domestic firms, it is higher in foreign-owned firms, in listed firms, and in firms affiliated to a multinational group (Brunello, Graziano and Parigi, 2001). In the banking sector, the same trend of Pay-Performance relationship is found. Hubbard and Palia (1994) state that CEO compensation is strongly and positively related to shareholder wealth in US banks. Barro and Barro (1990) come to the same conclusion when studying about 70 large US commercial banks. There seems to be a consensus among academic literature regarding how shareholders’ interests are matched with the manager’s interests. Therefore, it could be generally concluded that an outcome-based contract is an effective remedy for the agency problem.
7
shareholders capture most of the benefits. But if these projects fail, the debt-holders bear most of the costs. In organizations with high leverage, such as banks, the CEOs should act on behalf of debt-holders to the same degree as he does on behalf of shareholders (John and John, 1993). Therefore, the relationship between shareholders and the managers should not be too closely aligned by Pay-for-Performance. In other words, low Pay-Performance sensitivity helps minimize the agency costs of debts. Therefore, the optimal Pay-Performance sensitivity in banks should be lower than that of other non-banking corporations (John and Qian, 2003). Consequently, the first hypothesis of this thesis is specified as follows:
Hypothesis 1: Pay-Performance sensitivity in the banking industry is lower than
pay-performance sensitivity in other industries.
2. The Current Financial Crisis and its Effect on Pay-Performance Sensitivity
2.1 The current financial crisis8
several banks tried to sell out of their positions, prices decreased even further. Then public’s concerns about illiquidity, which could turn to potential insolvency, arose. Finally, bank runs started, followed by the bankruptcy of Lehman Brothers in September 2008. This event triggered a global banking panic (Diamond and Rajan, 2009), and the credit crisis started spreading since September 2008.
The banking industry suffered the most from this financial crisis, as it is where the crisis was initiated. According to the Failed Bank List provided by the US Federal Deposit Insurance Corporation, bankruptcy filings were dramatically increasing in quantity. While in 2007 only 3 banks were bankrupt, 25 banks filed for bankruptcy in 2008. This number jumped to 140 banks in 2009, and peaked at 157 banks in 2010. In 2011 92 banks went bankrupt. It is noteworthy that in 2005 and 2006, no banks were reported of bankruptcy. Since banks are mutually lenders and borrowers, other banks that survived also experienced a bad time (Brunnermeier, 2008). Apart from the banking sector, other sectors were also affected by the financial crisis. A further analysis of the crisis effect on financial performance of banks and of firms in other sectors is discussed in subsection 2.2.
2.2 Effect of the credit crisis on Pay-for-Performance a. Non-banking sectors
9
2007 the number of business bankruptcy filings has been increasing annually. In 2007 there were 35292 bankruptcy filings among US business. In 2008 the number was 30741, marking an increase of 40%. In 2009 there were 49077 filings, indicating nearly 60% increase of bankrupt companies. In 2010 the bankruptcy’s acceleration slowed down with 61148 filings reported.
Generally, CEO compensation is reduced during the crisis, as recorded by the US Bureau of Labor Statistics. From 2007 to 2008, average CEO total compensation decreased by 18.5%. In 2009, the average total compensation fell down by 11.1%, and in 2010 it fell by nearly 30%. Intuitively these numbers show co-movements between CEO compensation and firm performance, which indicates that implementation of Pay-for-Performance is likely to be unchanged. However, this intuition needs affirmation by empirical evidence. While researchers have focused on the change in either CEO compensation or firm performance during the crisis, the change in their relationship has not been tested. Only in the article of Gilson and Vetsuypens (1993) Pay-for-Performance is re-evaluated under the effect of a financial distress. However, in the context of that paper a financial distress implies either filing for bankruptcy or debt restructuring event. These events are of individual firms, and can occur in different periods of time.
10
Based on these arguments, the second hypothesis is specified as follows:
Hypothesis 2: In non-banking corporations CEO compensation becomes more sensitive to firm
performance after 2008.
b. The banking sector
Pay-for-Performance in the banking sector can be indirectly influenced by the credit crisis via two mechanisms. The first mechanism is debt structure. Beltratti and Stulz (2012) state that financial losses force banks, especially large banks with high debt-to-equity ratio, to reduce their leverage. According to John and Qian (2003), a lower debt-to-equity ratio reduces the manager’s sympathy with debt-holders’ concerns, and implies higher alignment between shareholders’ interests and the manager’s interests. Higher interest alignment, in turn, implies that Pay-Performance relationship is magnified. Consequently, financial distress can lead to higher sensitivity of CEO compensation to bank performance due to lower leverage level desired.
Another mechanism via which the crisis can affect banks’ Pay-for-Performance is government regulation. In the aftermath of the credit crisis, all US banks are required to report their executive pay structure and to re-arrange the structure if necessary (US Department of Treasury, 2009). The purpose is to eliminate excessive risk-taking behavior of banks’ CEOs, as stated by the US Department of Treasury. This indicates a nation-wide change in Pay-Performance relationship of the US banking industry. More specifically, the US Department of Treasury aims at reforming corporate governance and compensation rules to “better promote long-term value and growth for shareholders, companies, workers and the economy at large and to prevent such financial crises from occurring again”. This attempt of the US government suggests that after the credit crisis CEO compensation is more strongly associated with shareholders’ value as well as with firm value. Therefore, Pay-Performance sensitivity is predicted to be higher after the onset of the financial crisis because of government regulation.
The two arguments mentioned above constitute my last hypothesis, which is also consistent with Jensen’s (1989) point of view. The third hypothesis is specified as follows:
Hypothesis 3: In the banking industry CEO compensation becomes more sensitive to firm
11
III. Methodology
There are two approaches to measure the sensitivity of the CEO’s wealth to the firm performance. In the context of this thesis, the first approach will be called “value-value approach”, and the second approach will be called “percentage-percentage approach”. “Value-value approach” is developed by Jensen and Murphy (1990). It indicates the change in dollar value of CEO compensation as a result of a change in dollar value of the company. “Percentage-percentage approach” is developed by Hall and Liebman (1998). It indicates “Percentage-percentage change of CEO compensation as a result of 1% change in firm value. Hall and Liebman claim that Jensen and Murphy’s estimate can result in misleading picture of pay-performance association. This is because the dollar change in firm value may be very large, which makes the change of millions of dollar in CEO compensation appears very small when being compared with the dollar change in firm value. This is the reason why Jensen and Murphy do not find a considerable relation between CEO compensation and firm performance (they find “CEO wealth changes $3.25 for every $1,000 change in shareholder wealth”). In contrast, Hall and Liebman, when attempting to find the percentage change of CEO wealth as a result of a percentage change in firm performance, observed a more comprehensible pay-performance relationship. Therefore, the percentage-percentage approach is used to examine the Pay-Performance sensitivity in this thesis. Consequently, all variables related to CEO remuneration and firm performance mentioned above are transformed to annual percentage change.
Fixed-effect regression for panel data is used to construct the empirical models. Hypothesis 1 is tested by model (1), and hypotheses 2 and 3 are tested by model (2). In both models growth of firm performance is lagged one year against CEO compensation growth. This is because a CEO is paid for his performance in the previous year.
(1) CEO Pay Growthi,t = α0 + α1Firm Performance Growthi,t-1
+ α2 Firm Performance Growthi,t-1*Dum_banki
+ γi + εi,t
12
Hypothesis 1 is tested by the results for interaction of Firm Performance Growth and Dum_bank. Dum_bank is a dummy variable that distinguishes banks and non-banking corporations. The coefficient attached to the interaction, α2, compares the difference of pay-performance sensitivity
of the banking industry and of other industries. As Dum_bank equals to 1 if the observed firm is a bank, a negative α2 implies that Pay-Performance sensitivity in the banking sector is lower than
that in other sectors. In other words, hypothesis 1 is not supported if α2 is negative. The
magnitude of α2 is the value of percentage difference between Pay-for-Performance in the
banking industry and Pay-for-Performance in other industries respectively.
In addition, α1 examines the sensitivity of CEO pay growth to firm performance growth in
non-banking industries. CEO compensation changes by α1% for every 1% change of firm value. If
firm performance and CEO compensation have perfect co-movements with each other, α1 is
equal to 1 (i.e. 2% decrease in firm value leads to 2% decrease in CEO pay, 10% increase in firm value leads to 10% increase in CEO pay, etc.). If α1 >1, CEO pay grows faster than firm value
when the firm is performing well; but α1 >1 also means when the firm is less profitable, CEO pay
decreases relatively faster than firm value. If 0 < α1 < 1, CEO pay increases relatively less than
the increase of firm value when the firm is performing well; but when firm performance is going down, CEO pay decreases relatively less than firm value does. If α1 < 0, CEO pay increases
when firm value decreases, and decreases when firm value increases. Besides that, the sum of α1
and α2 identifies the sensitivity of Pay-Performance relation in the banking industry. The
magnitude of sum (α1 + α2) is interpreted similarly to α1’s magnitude.
Model (2) tests the effect of the financial crisis, and is applied for both banks and non-banking corporations.
(2) CEO Pay Growthi,t = β0 + β1 Dum_crisist + β2Firm Performance Growthi,t-1
+ β3 Firm Performance Growthi,t-1*Dum_crisist
+ γi + εi,t
13
Hypothesis 2 and 3 are tested by the results for interaction of Firm Performance Growth and Dum_crisis. In order to examine hypotheses 2 and 3, the sample is divided into two subsamples. The first contains only observations of banks, and the second contains only observations of non-banking firms. Model (2) is run with non-banking subsample to test hypothesis 2. Because Dum_crisis equals to 1 if observations are in or after 2008, if attached coefficient β3 of the
interaction term is positive, hypothesis 2 is supported. In that case, it could be concluded that in the banking industry CEO compensation becomes more sensitive to firm performance after the start of the crisis. Model (2) is run with a non-banking subsample to test hypothesis 3. Similarly to banking industry, positive β3 generated from non-banking subsample supports hypothesis 3. If
that is the case, the conclusion is that in non-financing industries CEO compensation becomes more sensitive to firm performance after the onset of the crisis.
Additionally, β2 is the sensitivity of CEO compensation to financial performance before the crisis
in either banking industry or non-banking industries. The sum of β2 and β3 is the sensitivity of
CEO compensation to financial performance after 2008. The magnitudes of β2 and of the sum (β2
+ β3) are interpreted similarly to that of α1 in model (1). The sign of β1 indicates whether CEO
compensation grows or declines after 2008, and the magnitude of β1 implies the value of
percentage change in CEO compensation.
In both regressions the intercepts α0 and β0 represent effects of various omitted factors that also
may affect CEO compensation growth such as firm size, gender, age, years of experience, nationality, etc.
IV. Variable Description and Sample Selection
14
CEO remuneration of these corporations is collected from ExecuComp, and data of financial performance is collected from Datastream. The gathered dataset contains 112 listed banks and 1303 listed non-banking firms from other sectors in the period from 2005 to 2011.
CEO remuneration
Three categories of CEO remuneration are used in this study: salary, variable compensation and total compensation. This is different from the measurement of CEO remuneration in previous studies. Compensation package of a CEO comprises different elements such as annual salary, bonus, granted stocks, granted stock options and long term incentive payment. Previous studies (e.g. Gilson and Vetsuypens, 1993; Hubbard and Palia, 1995; Core, Holthausen and Larcker, 1999; Brick, Palmon and Wald, 2006) divide the total package into two categories: cash and cash compensation. Cash compensation consists of annual salary and bonus in cash. The non-cash category consists of granted stocks and granted stock options. Long term incentive payment (LTIP) is paid out irregularly and only by a few companies. Therefore, including LTIP will cause outliers in some years or in some companies. In order to avoid the negative effect of outliers, LTIP is excluded in previous studies as well as in this study. As mentioned, this study has different categories of CEO remuneration: annual salary, variable compensation and total compensation. Annual salary is contracted between the investors and the CEO for a period of time. So it is unlikely to change corresponding to the firm’s annual performance, which reflects the CEO’s effort (Gilson and Vetsuypen, 1993). In contrast, variable compensation tends to vary with the change in CEO’s effort. It comprises bonus, granted stocks and granted stock options, whose fair value are purported to be correlated with firm performance. Total compensation includes both fixed compensation and variable compensation. Total compensation is used to examine the effect of firm performance on the whole compensation package.
Financial performance
15
relations with CEO compensation are possible with different measures. Including both measures helps to observe the relationship more accurately.
Dummy for banks
This is the variable that categorizes banks and non-banking corporations. It equals to 1 if the observed company is a bank, and equals to 0 if the observed company is not a bank. This variable is necessary for the comparison between banks’ Pay-Performance sensitivity and that of other industries.
Dummy for the credit crisis
This variable controls for firm performance before and after the onset of the crisis. The dummy variable for the credit crisis helps examine the crisis effect on Pay-Performance relationship. Lehman Brother’s collapse in 2008 triggered the global credit crisis (Brunnermeier, 2008). This suggests that 2008 could be chosen to be the starting year of the financial crisis for this dummy variable. Therefore, it equals to 1 if the observed financial outcome of firms and the observed CEO compensation are of fiscal years 2008, 2009 and 2010. It equals to 0 if the observed financial outcome of firms and the observed CEO compensation are of fiscal years 2005, 2006 and 2007.
16
executive by different corporations, seems to be irrelevant for this study. The second comparison, which is between non-executive compensation and executive compensation, also seems to be inappropriate. This is because this study focuses on the change of compensation paid to the managers when they are in executive positions. These reasons suggest that compensation growth of new CEOs’ in the first year of their position, which comprises both comparisons mentioned above, should be excluded from the dataset. Consequently, the final dataset is an unbalanced panel data with 75 US banks and 830 US non-banking firms listed on US stock market from 2005 to 2011.
Table I
Definition of variables measuring Firm characteristics, CEO remuneration and Financial Performance of US listed firms in banking sector and other sectors.
Appendix A presents descriptive statistics of all variables used in this study for banking sector and other sectors. In the US change in variable compensation had lower average and median than change in fixed salary. This could be explained by the negative average change in return on assets (of banks and of firms in other sectors) and negative average change in stock price (of banks). In the period 2005 - 2011, US firms did not perform well, which led to slow growth of
Definition Source
Firm
characteristic
Dum_bank 1 if observation is a bank, 0 otherwise Bankscope
Orbis
CEO
remuneration
Change in Salary Percentage change in annual salary paid to the CEO ExecuComp Change in Variable
Compensation
Percentage change in variable compensation paid to the CEO, which consists of bonus, restricted stock granted and stock options granted.
ExecuComp
Change in Total Compensation
Percentage change in total compensation of the CEO, which consists of salary, bonus, restricted stock granted and stock options granted.
ExecuComp
Financial performance
Change in Return on Assets (ROA)
Percentage change in the annual net income divided by the book value of total assets at the end of the year
Datastream
Change in Stock Price
Percentage change in the annual stock price, which is the closing price at the end of the year
17
effort-related compensation. Growth of salary was positive, although it should be equal to zero to reflect fixed annual salary. The reason might come from different contract periods between CEO and investors in different companies. In every fiscal year of the period 2005-2011, while annual salary did not change in some companies, it could fluctuate in other companies. This happens when a contract between the CEO and the investors terminates, and a new contract is signed. Positive salary growth implies that either many CEOs successfully negotiated to have higher contracted salaries or fewer CEOs have relatively large salary increase even when other CEOs’ salaries are decreased. Changes in all three compensation categories in the banking industry were lower than those of other industries. Market-based performance (stock price) of banks also grew more slowly than market-based performance of firms in other sectors. In contrast, growth of accounting-based performance (return on assets) in banking industry is reported to be higher than that of non-banking industries.
18 Figure 1. Percentage change of variables measuring CEO compensation and firm performance in US non-banking industries. Sample consists of 830 listed firms from 2005 to 2011.
Figure 2. Percentage change of variables measuring CEO compensation and firm performance in the US banking industry. Sample consists of 75 listed banks from 2005 to 2011.
-40 -30 -20 -10 0 10 20 30 40 50 60 Change in salary (%)
Change in variable pay (%)
Change in total compensation (%) Change in ROA (%)
Change in stock price (%)
Non-banking industries
P erc entag e C ha ng e re lative to pre vious y esr 2005 2006 2007 2008 2009 2010 2011 -40 -30 -20 -10 0 10 20 30 40 50 60 Change in salary (%)Change in variable pay (%)
Change in total compensation (%) Change in ROA (%)
19
In the banking sector, there are also different patterns of different compensation categories. Annual percentage changes of salary fluctuated less, compared to changes in return on assets and changes in stock price. But salary did appear to follow change in ROA or change in stock price in previous year. Change in variable compensation is more aligned with stock price growth than with ROA growth. More specifically, variable compensation decreased in 2007 after a decrease of stock price in 2006 while return on assets slightly increased in 2006. In another timeframe, variable compensation increased in 2008 after stock prices grew in 2007 but return on assets decreased in 2007. Only variable compensation in 2010 varied according to the variation of both stock price and return on assets in 2009: all of them increased. In contrast, change in total compensation is more aligned with change in return on assets. Growth of total compensation in 2007, 2010 and 2011 respectively followed the growth of return on assets in 2006, 2009 and 2010. In 2008 total compensation decreased after a decrease of return on assets in 2007. For stock price, only in 2010 and in 2011 the total compensation had the same trend with stock price change in 2009 and in 2010, respectively.
V.
Results and Discussion
1. Pay-for-Performance in banks and in non-banking corporations
Table II presents the regression results for model (1). Panel A shows estimated coefficients that relate change in three CEO compensation categories to percentage change of return on assets for all financial and non-banking companies. Panel B shows the relation between percentage changes in three CEO compensation categories to percentage change in stock price for all companies. The second coefficients in both panels attempt to compare degrees of Pay-Performance sensitivity in banking industry with that of other industries.
Hypothesis 1 is supported by the second regression in panel A, whose dependent variable is “Change in variable compensation”. The coefficient attached to the interaction term of this regression is significantly negative (β2 = -0.21). This implies that given a same growth of
20 Table II
Comparison between the banking sector and non-banking sectors
The unbalanced panel data consists of 75 banks and 830 firms in other sectors that were listed in US stock market in 2005-2011. Dependent variables are yearly percentage changes of three compensation categories: annual salary, variable compensation package and total compensation. All independent variables measuring “Firm Performance Growth” are lagged one year. Dum_bank equals to 1 if the observed company is a bank, and equals to 0 otherwise. Change in ROA and Change in stock price are the yearly percentage difference between values of ROA and of stock price, respectively. The interaction terms of Change in ROA and Dum_bank and of Change in stock price and Dum_bank measure the difference of pay- performance sensitivity between the banking sector and other sectors. The least square method with fixed-effect of cross-section is used in the regression. Standard errors of coefficients are in brackets. ***, **, * denote statistical significance at 1%, 5% and 10% respectively.
Comparison between the Banking sector and other Non-banking sectors
Dependent variables Independent variables Change in Salary Change in variable compensation Change in total compensation
Panel A: Financial Performance measured as Percentage Change of Return on Assets
Change in ROAt-1 0.00 -0.02* -0.02***
(0.00) (0.01) (0.01)
Change in ROAt-1*Dum_bank 0.03 -0.21*** -0.004
(0.02) (0.06) (0.05) Constant 4.63*** -0.39 2.3*** (0.17) (0.68) (0.59) Number of observations 4583 3883 4169 R-squared 0.240 0.187 0.197
Panel B: Financial Performance measured as Percentage Change of Stock Price
Change in Stock Pricet-1 0.02*** -0.12*** -0.13***
(0.01) (0.02) (0.02)
Change in Stock Pricet-1*
21
Insignificant interaction’s coefficients of the first and the third regressions in panel A indicate that between the banking industry and other industries there is no considerable difference in the sensitivity of salary and total compensation of CEOs to firm value. In panel B the estimated interaction’s slopes imply that the sensitivity of all three compensation categories to stock price in banks is similar to that in non-banking corporations. Therefore, it is concluded that with accounting-based performance, only sensitivity of CEOs’ bonuses and stock-based compensation in the banking industry is lower than that in other industries. But with market-based performance, the sensitivity of all compensation categories to firm value does not vary among these industries.
In addition, the variable compensation and total compensation are negatively related to firm performance in both panel A and panel B. This applies for both banks and non-banking corporations. More specifically, panel A shows that a 1% decline in return on assets increases CEO variable compensation by 0.02%, and increases total compensation in non-banking firms by 0.02%. For the banking industry, a 1% decrease in return on assets increases CEO variable compensation and total compensation by 0.23%2 and by 0.02%3 respectively.
Regarding market-based firm performance, 1% decline of stock price leads to 0.12% growth of CEO variable compensation and 0.13% growth of CEO total compensation in both the banking sector and other non-banking sectors. While for-Performance implies that the slope of Pay-Performance should be positive, the estimated slopes of Variable Compensation-Stock Price and Total Compensation-Stock Price are significantly negative. This finding is contrast to Pay-for-Performance implication. A reason for this result can come from the arguments of Core, Holthausen and Larcker (1998) and Brick, Palmon and Wald (2006). These researchers argue that firms with weak governance structure have bigger agency problem, and therefore, managers of these firms receive higher compensation despite poor financial performance. This is concluded after they find a significant negative relation between components of CEO compensation with subsequent firm operating and stock return performance.
2 This slope is obtained by adding the coefficient of interaction term α
2=-0.21 to the coefficient of lagged change in
ROA α1=-0.02.
3 This slope is equal to -0.02, which is the coefficient of “Change in ROA
t-1”, because the coefficient of interaction
22
In the context of this thesis, the compensation that is not correlated to firm performance tends to be bonus. Variable compensation comprises bonus and stock-based compensation. Because stock-based compensation is positively correlated with stock price (Jensen and Murphy, 1990; John and Qian, 2003), the negative sign of variable compensation is likely to result from negative relation between bonus and stock price. Furthermore, the value of bonus seems to be large enough to outweigh the association between salary and firm performance. This statement is drawn from negative sign of total compensation (α1=-0.13) and positive sign of salary to stock
price (α1=0.02). This finding could signal inefficient governance of observed corporations in this
study, or at least most of them. One example of bad governance could be Yahoo’s situation mentioned in the introduction: the COO received $58 million bonus after working inefficiently for 15 months.
2. Effect of the Credit Crisis on Pay-for-Performance
a. Non-banking sectors23 Table III
Regression result for listed US firms in non-banking sectors
The unbalanced panel data consists of 830 firms in non-banking industries that were listed in US stock market in 2005-2011. Dependent variables are yearly percentage changes of three compensation categories: annual salary, variable compensation package and total compensation. All independent variables measuring “Firm Performance Growth” are lagged one year. Dum_crisis equals to 1 if subscript t-1 is 2008, 2009 and 2010, and equals to 0 otherwise. Change in ROA and Change in stock price are the yearly percentage difference between values of ROA and of stock price, respectively. The interaction terms of Change in ROA and Dum_crisis and of Change in stock price and Dum_crisis measure the change of pay-for-performance relationship after the onset of the crisis. The least square method with fixed-effect of cross-section is used in the regression. Standard errors of coefficients are in brackets. ***, **, * denote statistical significance at 1%, 5% and 10% respectively.
Non-banking industries Dependent variables Change in Salary Change in variable compensation Change in total compensation Independent variables
Panel A: Financial Performance measured as Percentage Change of Return on Assets
Dum_crisist-1 -4.41*** 10.433*** -0.91
(0.39) (1.69) (1.41)
Change in ROAt-1 0.002 0.03 -0.02
(0.004) (0.02) (0.01)
Change in ROAt-1*Dum_crisist -0.004 -0.04** -0.01
(0.004) (0.02) (0.02) Constant 7.73*** -7.77*** 3.12*** (0.32) (1.47) (1.17) Number of observations 4298 3565 3869 R-squared 0.264 0.196 0.194
Panel B: Financial Performance measured as Percentage Change of Stock Price
Dum_crisist-1 -4.32*** 13.99*** -0.84
(0.42) (1.86) (1.54)
Change in Stock Pricet-1 -0.002 0.12** -0.1**
(0.01) (0.06) (0.05)
Change in Stock Pricet-1*
24
A more detailed evaluation of these results shows that three compensation packages have different degrees of sensitivity to firm performance in non-banking corporations. Salary is not likely to be related to firm performance before and after 2008. All the values for β2 and sum (β2
+ β3) of salary growth are small and insignificant in both panel A and panel B. In terms of
variable compensation, CEOs are awarded according to Pay-for-Performance strategy before the crisis. This is reflected by β2 = 0.12 in panel B, which means variable compensation changes by
0.12% for every 1% change of stock price. However, after the crisis began, the relationship between variable compensation and market-based performance is reversed. If stock price decreases by 1%, variable compensation increases by 0.13%4. This reversed result can be again explained with the effect of bonuses. It could be explained that because value of stock-based compensation follows stock price’s trend, the dominating effect of high bonuses renders the category variable compensation not corresponding to market-based performance. The corporate governance of non-banking firms seems to be poor.
The last category, total compensation, is not paid corresponding to growth of return on assets before and after 2008. However, this category is negatively related to stock price before the crisis, it decreases by 0.1% if stock price increases by 1%. This relationship did not likely to change after the crisis began, as implied by insignificant coefficient β3. This result suggests that
pay-for-performance is not strictly implemented in many non-banking corporations. In addition, the first regressions in panel A and panel B show significant and negative coefficients of Dum_crisis. This means that salary growth is reduced in the aftermath of the crisis. In contrast, variable compensation grows faster after the crisis began, as reflected by significant and positive Dum_crisis’s slopes in the second regressions of both panels. Again, this finding signals poor governance within non-banking corporations. While salary and stock-based compensation are in line with financial difficulty of companies, great amounts of bonuses are received by the CEOs. Another interesting finding is that the total compensation, which is the sum of salary and variable compensation, is not statistically affected by the credit crisis. Coefficients β1 in the regressions of
total compensation are insignificant, although negative. This finding suggests that the bonuses
4
This slope is obtained by adding the coefficient of interaction term β3=-0.25 to the coefficient of lagged change in
25
given to CEOs are just large enough to offset the decreased amounts of salary and stock-based compensation.
b. The banking sector
Table IV presents the estimated regression results for effect of the financial crisis on Pay-for-Performance in US banks. These results render hypothesis 3 rejected. Both panel A for accounting performance and panel B for market-based performance show insignificant coefficients of interaction term. In other words, there seems to be no noticeable change in the relationship between bank performance and different categories of CEO compensation after the credit crisis. Therefore, it could be concluded that the crisis has no significant effect on Pay-for-Performance in the banking industry. An explanation for this finding is that during the financial distress corporations have to face strategic or regulatory constraints. One of these constraints is the limitation of CEO compensation, regardless of his or her performance or professional skill level (Jensen, 1991; DeAngelo and DeAngelo, 1991).
Additionally, before the crisis only salary is paid correspondingly to stock price, but not to return on assets. Statistically, salary grows by 0.05% when stock price increases by 1%. Because of no significant effect of the financial crisis on this relationship, post-crisis salary’s sensitivity can be interpreted similarly to pre-crisis salary’s sensitivity. Other compensation categories, which are variable compensation and total compensation, are not related to both accounting-based and market-based bank performance. It could be concluded that pay-for-performance does not seem to be implemented within the US banking industry, both pre-crisis and post-crisis.
26 Table IV
Regression result for listed US banks
The unbalanced panel data consists of 75 banks listed in US stock market in 2005-2011. Dependent variables are yearly percentage changes of three compensation categories: annual salary, variable compensation package and total compensation. All independent variables measuring “Firm Performance Growth” are lagged one year. Dum_crisis equals to 1 if subscript t-1 is 2008, 2009 and 2010, and equals to 0 otherwise. Change in ROA and Change in stock price are the yearly percentage difference between values of ROA and of stock price, respectively. The interaction terms of Change in ROA and Dum_crisis and of Change in stock price and Dum_crisis measure the change of pay-for-performance relationship after the onset of the crisis. The least square method with fixed-effect of cross-section is used in the regression. Standard errors of coefficients are in brackets. ***, **, * denote statistical significance at 1%, 5% and 10% respectively. Banking sector Dependent variables Change in Salary Change in variable compensation Change in total compensation Independent variables
Panel A: Financial Performance measured as Percentage Change of Return on Assets
Dum_crisist-1 -1.27** -9.26 4.25
(0.64) (5.7) 5.05
Change in ROAt-1 0.03 -0.12 0.32
(0.03) (0.24) 0.25
Change in ROAt-1*Dum_crisist -0.002 -0.14 -0.35
(0.03) (0.25) 0.25 Constant 4.59*** -2.76 -3.53 (0.51) (4.45) 4.12 Number of observations 285 318 300 R-squared 0.459 0.204 0.242
Panel B: Financial Performance measured as Percentage Change of Stock Price
Dum_crisist-1 -1.06* -5.04 3.99
(0.61) (5.76) 5.13
Change in Stock Pricet-1 0.05* -0.33 0.09
(0.03) (0.3) 0.26
Change in Stock Pricet-1*
27
VI. Conclusion
With a dataset covering 75 banks and 830 non-banking corporations that were listed in US stock market in 2005-2011, this thesis addresses two research questions related to Pay-for-Performance. The first question is related to different degrees of Pay-for-Performance implementation between banks and non-banking corporations. The second question is about how Pay-for-Performance implementation is influenced by the financial crisis. I employ fixed-effect (cross-section) regression for panel data to capture the sensitivity of CEO compensation to financial performance, which approximates the CEO’s effort to maximize shareholders’ value. All specifications mentioned below fit well with the dataset, as implied by acceptable levels of R-squares, which range from 12.7% to 47%.
For the first research question, I find that Pay-for-Performance tends to be less implemented in the banking industry, specifically with regard to the compensation package that is likely to vary with CEOs’ effort (bonus and stock-based compensation). This finding confirms the expectation of lower sensitivity of CEO compensation to financial performance in the banking sector than in other sectors. This could be explained by two reasons: (1) banks are more regulated than firms in other sectors and (2) banks have higher leverage than firms in other sectors. However, the sensitivities of salary and of total compensation to firm value are indifferent between banks and non-banking corporations.
28
could have significant effect on Pay-for-Performance. These factors can be related to the credit crisis, or firm characteristics or executive characteristics. Future research on Pay-for-Performance is encouraged to identify those factors in order to evaluate the effect of the crisis more accurately. The second explanation for unexpected result tends to come from the measurement method. The method to measure Pay-for-Performance used in this study tend not to be the most relevant one, although the reasons for this hitherto is unknown. Therefore, these reasons need to be investigated by researchers so that Pay-for-Performance can be measured more accurately.
Based on the finding of this thesis, it is reasonable to assume that US corporations, both banks and non-banking ones, tend to have ineffective strategies for CEO incentives during the period 2005-2011. Recommended Pay-for-Performance strategy indicates the co-movements of CEO compensation and firm value. However, this study observes mostly either opposite movements or unrelated movements of CEO compensation and firm value. In non-banking industries both salary and total compensation tend to be unrelated to market-based performance of firms. In the banking industry only salary is paid corresponding to market-based performance of banks. These findings suggest that within US firms the practice of Pay-for-Performance is not likely to be widely applied as recommended by Agency Theory.
Although this study has important findings, it also has an unavoidable limitation, which results from the dataset. In order to ensure normal distribution for the least square methodology, some observations for CEO compensation and financial performance of companies are deleted. Although this helps avoid outliers that could lead to biased regression results, it also renders the study ignore some observations that might also be helpful to examine Pay-Performance relationship and the crisis effect on this relationship. Therefore, researchers are urged to improve this limitation in order to capture the implementation of Pay-for-Performance more adequately.
Pay-for-29
Performance in multinational corporations. Future research is also encouraged to examine the applicability of this thesis’s findings by replicating the models with data from companies in economies different from the US. These companies can operate in the same industry but different countries, or they can operate in the same country but in different industries.
30
Appendices
Appendix A
Descriptive statistics of variables measuring CEO remuneration and Financial Performance of US listed firms in the banking sector and other sectors.
The unbalanced panel data consists of 75 banks and 830 firms in other sectors that were listed in US stock market in 2005-2011. Definition of variables is provided in table I. Observations for all variables are transferred to percentage growth relative to previous year.
Mean Median Maximum Minimum Std. Deviation
Number of Observations Banking sector
Change in salary (%) 3.51 2.53 18.93 -16.67 4.94 368
Change in variable pay (%) -6.36 -4.71 99.42 -100 4.61 378
Change in total compensation (%) 0.88 0.06 107.23 -99.64 3.56 382
Change in ROA (%) -10.99 -3.06 160 -18.72 4.91 381
Change in stock price (%) -1.26 0.04 58.82 -5.87 2.19 415
Other sectors
Change in salary (%) 5.33 3.5 100 -100 2.52 5052
Change in variable pay (%) -0.33 0.77 100 -100 4.66 4204
Change in total compensation (%) 3.64 2.8 100 -100 3.61 4529
Change in ROA (%) -14.17 -4.16 495.14 -49.77 9.59 4781
31 Appendix B
Annual mean of variables measuring CEO remuneration and Financial Performance of US listed firms in the banking sector and other sectors.
The unbalanced panel data consists of 75 banks and 830 firms in other sectors that were listed in US stock market in 2005-2011. Dependent variables are yearly percentage changes of three compensation categories: annual salary, variable compensation package and total compensation. Independent variables are yearly percentage change of return on assets (ROA) and yearly percentage change of stock price.
Mean
2005 2006 2007 2008 2009 2010 2011
Banking sector
Change in salary (%) 5.45 4.87 3.71 4.20 1.55 2.99 2.79
Change in variable pay (%) 49.42 0.01 -7.81 -4.91 10.76 17.29 -59.02
Change in total compensation (%) 6.41 -15.74 -2.69 -6.78 -3.68 10.18 17.41
Change in ROA (%) 2.72 3.16 -10.70 -35.24 -28.18 -3.29 0.34
Change in stock price (%) 16.28 0.85 9.31 -19.15 -11.41 -6.39 0.16
Other sectors
Change in salary (%) 10.46 8.40 6.41 5.78 0.65 3.21 4.66
Change in variable pay (%) -8.58 -22.22 -4.15 -2.78 -1.38 15.27 4.50
Change in total compensation (%) 8.20 0.92 4.83 -2.83 -2.87 15.36 4.59
Change in ROA (%) -3.71 -1.60 -7.58 -27.543 -34.44 -10.60 -1.45
32 Appendix C
Correlation matrix
The unbalanced panel data consists of 75 banks and 830 firms in other sectors that were listed in US stock market in 2005-2011. Dependent variables are yearly percentage changes of three compensation categories: annual salary, variable compensation package and total compensation. Independent variables are yearly percentage change of return on assets (ROA) and yearly percentage change of stock price.
Variables -1 -2 -3 -4 -5
Banking Industry
(1) Change in Salary 1
(2) Change in Variable Compensation -0.11 1
(3) Change in Total Compensation 0.18 -0.17 1
(4) Change in ROA (lag) 0.18 -0.15 -0.04 1
(5) Change in Stock Price (lag) 0.1 -0.05 -0.12 0.3 1
Other sectors
(1) Change in Salary 1
(2) Change in Variable Compensation 0.17 1
(3) Change in Total Compensation 0.15 0.88 1
(4) Change in ROA (lag) -0.01 -0.01 -0.03 1
(5) Change in Stock Price (lag) 0.09 -0.09 -0.1 0.16 1
All sectors
(1) Change in Salary 1
(2) Change in Variable Compensation 0.16 1
(3) Change in Total Compensation 0.15 0.8 1
(4) Change in ROA (lag) -0.01 -0.01 -0.03 1
33
References
Barro, J.B. and Barro, R.J. (1990), Pay, Performance, and Turnover of Bank CEOs, Journal of
Labor Economics 8 (4): 448-481.
Beltratti, A. and Stulz, R.M. (2012), The Credit Crisis Around the Globe: Why Did Some Banks Perform Better?, Journal of Financial Economics 105: 1-17.
Berk, J., DeMarzo P. and Harford J. (2009), Fundamentals of Corporate Finance, Pearson
International Edition.
Brick, I. E., Palmon, O. and Wald, J.K. (2006), CEO compensation, director compensation, and firm performance: Evidence of cronyism?, Journal of Corporate Finance 12: 403-423.
Brunello, G., Graziano, C. and Parigi, B. (2001), Executive compensation and firm performance in Italy, International Journal of Industrial Organization 19: 133–161.
Brunnermeier, M. K. (2008), Deciphering the Liquidity and Credit Crunch 2007-08, US National
Bureau of Economic Research, NBER working paper, No 14612.
Campello, M., Graham J.R. and Harvey C.R. (2010), The real Effects of Financial Constraints: Evidence from a financial crisis, Journal of Financial Economics 97(3): 470-487.
CNBC (2014, April 17), Fired Yahoo exec gets $58M for 15 months of work, retrieved from http://www.cnbc.com
Core, J.E., Holthausen, R.W. and Larcker, D.F. (1999), Corporate governance, chief executive officer compensation and firm performance, Journal of Financial Economics 51: 371-406.
Coughlan, A.T. and Schmidt, R.M. (1985), Executive Compensation, Management Turnover, and Firm Performance: An Empirical Investigation, Journal of Accounting Economics 7: 43-66.
Crawford, A., Ezzell J., and Miles J. (1995), Bank CEO pay performance relations and the effects of deregulation, Journal of Business 68, 231-256.
Dalton, D.R., Hitt, M.A., Certo, S.T. and Dalton, C.M. (2007), The fundamental agency problem and its mitigation, The Academy of Management Annals 1: 1-64.
DeAngelo, H. and DeAngelo L. (1991), Union negotiations and corporate policy: A study of labor "givebacks" in the domestic steel industry during the 1980s, Journal of Financial
Economics 30: 3-44.
Diamond D.W. and Rajan R. (2009), The Credit Crisis: Conjectures about Cases and Remedies,
34
Duchin R., Ozbas, O and Sensoy B.A. (2010), Costly external finance, corporate investment, and the subprime mortgage credit crisis, Journal of Financial Economics, 97 (3): 418–435.
Eisenhardt, K.M. (1989), Agency Theory: An Assessment and Review, The Academy of
Management Review 14(1), pp. 57-74.
Feldstein, M. (2008, October 4), The Problem is still Falling House Prices, The Wall Street
Journal, retrieved from http://online.wsj.com
Forbes Annual Executive Compensation Reports (2011, December), Historical Chart: Two decades of CEO Pay, Forbes, retrieved from http://www.forbes.com
Gilson S.C. and Vetsuypens M.R. (1993), CEO compensation in Financially Distressed Firms: an Empirical Analysis, The Journal of Finance, Vol XLVIII, No2.
Goldman, D. (2013, November 1), Oracle shareholders say “no” to Larry Ellison’s pay, CNN, retrieved from http://money.cnn.com
Hall, B. and Liebman, J. (1998), Are CEOs Really Paid Like Bureaucrats?, Quarterly Journal of
Economics 103: 653-91.
How many Berkshire Hathaway shares does Warren Buffet still own? (2009, July 14), retrieved from http://www.gurufocus.com/news/60051/how-many-berkshire-hathaway-shares-does-warren-buffett-still-own
Hubbard, R.G. and Palia, D. (1995), Executive pay and performance: Evidence from the US banking industry, Journal of Financial Economics 39: 105-130
Kato, T. and Kubo, K. (2006), CEO Compensation and Firm Performance in Japan: Evidence from New Panel Data on Individual CEO Pay, Journal of The Japanese and International
Economies 20: 1-19.
Jensen, M. and Murphy, K. J. (1990), Performance Pay and Top Management Incentives,
Journal of Political Economy 98: 225-264.
Jensen, M.C. (1989), Active Investors, LBOs, and the privatization of bankruptcy, Journal of
Applied Corporate Finance, 2: 235-244.
Jensen, M.C. (1991), Corporate control and the politics of finance, Journal of Applied Corporate
Finance, 4: 13-33.
Jensen, M.C. and Meckling W.H. (1976), Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure, Journal of Financial Economics 3(4), pp.305-360.
Jensen, M.C. and Murphy K.J. (1990), Performance pay and top-management incentives,
35
John K. and Qian Y. (2003), Incentive Features in CEO Compensation in the Banking Industry.
Economic Policy Review
John, T.A. and John K. (1993), Top Management Compensation and Capital Structure, Journal
of Finance, 48 (3): 949-974.
Lee, A. (2011, March 11), Buffet retains $100000 salary after faulting pay excesses, Bloomberg, retrieved from http://www.bloomberg.com
Mathiason, N. (2008, December 28), Three weeks that change the world, The Guardian, retrieved from http://www.theguardian.com
Moerland, P.W. (1995), Alternative Disciplinary Mechanisms in Different Corporate Systems,
Journal of Economic Behavior and Organization 26: 17-34.
Murphy, K.J. (1985), Corporate performance and managerial remuneration: An empirical analysis, Journal of Accounting and Economics 7 (1-3): 11-42.
Nyberg, A. J., Fulmer, I. S., Gerhart, B., & Carpenter, M. A. (2010), Agency theory revisited: CEO return and shareholder interest alignment. Academy of Management Journal, 53(5): 1029-1049.
Redmond, K. (2010, May 14), Bankruptcy Filings highest since 2006, US Courts, retrieved from http://www.uscourts.gov/news
Shleifer A. and Vishny R.W. (1997), A Survey of Corporate Governance, the Journal of
Finance, Vol LII, no2.
Smith, A. (1776), The Wealth of Nations, Cannan Edition (Modern Library, New York, 1937), p. 700.
The Economist journal (2013), The Origins of the Financial Crisis, published on September 7 2013.
US Department of Treasury (2009), Treasury Announces New Restrictions on Executive Compensation, available at http://www.treasury.gov/press-center/press-releases/Pages/tg15.aspx
Wessels, R.E. (2014) University of Groningen, Guest lecture on May 6th 2014, lecture slides are
