The effect of CEO Compensation towards Systemic risk in the
banking sector
MSc Finance Thesis
Edwin Bosman
S2998629 e.h.bosman@student.rug.nl University of Groningen Faculty of Economics and BusinessMSc Finance Supervisor: J.J Bosma
Date: 06-06-2018
Abstract: This paper investigates the relation of executive compensation to the contribution of a bank to the systemic risk. The study is divided into two periods: 2004–2006, in which the cash compensation is tested and 2006, which is examined in more detail, with the
compensation structure divided into cash, stocks and options. Compensation in options in 2006 is the only significant variable and has a negative effect on the bank’s contribution to the systemic risk. When controlling for other variables, leverages and especially volatility proved to impact the banks contribution to the systemic risk.
Words: 9841
Contents
1. Introduction...3
2. Literature review and hypothesis development ...6
2.1 Agency cost...6
2.2 Moral hazard ...7
2.3 Pay for performance...8
2.4 Compensation structure...9
2.5 Systemic risk ...11
2.6 Hypothesis development...12
3. Data and Methodology...14
3.1 Data ...14
3.2 Methodology ...17
4. Empirical results ...21
4.1 First sample ...21
4.2 Second sample...22
5. Summary and conclusions...25
Bibliography ...26
1. Introduction
It's easy to be a short-term hero. It is very easy for me to get tremendous results very short term, get that translated into compensation, and be off sailing in the Bahamas. But the goal for this company - and it's very difficult to do - the goal is to follow a four- or five-year process. (Paul Polman, April 4, 2010)
Compensation structure has been widely debated since the financial crisis started in 2007 as debtor for the banking crisis. Public opinion became that when the banks did well, their employees were paid well. However, when the banks did poorly, their employees were also paid well, and when the banks did very poorly, they were bailed out by taxpayers and their employees were still paid well (Cuomo, 2009). The storm of public critique regarding the compensation structure suggests people perceive it as one of the main causes of the financial crisis. Political pressures to change compensation have been escalated without strong evidence that compensation has been a significant factor driving excessive risk-taking in the financial sector. The beliefs that CEO compensation enforced the excessive risk-taking induced, among other actions, the Dodd-Frank Wall Street Reform and Consumer Protection Act in 2010, which limit compensations. However, many other claim that it’s not obvious that compensation is responsible for the crisis. Therefore, this research investigates the effect of compensation on the financial crisis. Most related studies examine the effect of executive compensation on firm performance. The systemic risk, which can start a financial crisis, hasn’t been investigated much. This paper studies the relation between bank CEO risk-taking incentives and the contribution of the bank to the systemic risk.
According to Danielson and Zigrand (2018):“Systemic risk arises when the failure of a
single entity or cluster of entities can cause a cascading failure, due to the size and
interconnectedness of institutions, which could potentially bankrupt or bring down the entire financial system”. Systemic risk is defined in this study as the possibility that an event at
corporate level could cause instability or breakdown of the entire financial sector, it occurs when the market decline by more than 40% in a six-month period (Brownless and
Engle,2012). Banks that are considered as a systemic risk are called too big to fail. These banks are large relative to the financial sector and significantly influence the market when they fail. Regulations concerning CEO compensation have been tightened as a consequence of the aftermath of the financial crisis, with the intention to prevent a next crisis. Taking that in mind, the focus on compensation regulations in the past banking legislation is not fully justified, as the role of CEO compensation in a bank’s excessive risk-taking is not settled (Murphy, 2009). This public opinion arose without a clear description of excessive risk-taking or how it can be separated from normal risk in the financial world. It is clear that CEO
related character, but many other factors have a larger impact, for example “too big to fail” companies, policies on loose monetary and home ownership, and poorly performing new financial instruments (exotic mortgages, securitization, and collateralized debt obligations).
When you give chief executives too much compensation in stock options, they concentrate too much on the stock price, and there is a perverse incentive to raise the stock price, particularly when the chief executive wants to exercise his own options. (George Akerlof, April 7, 2002)
The agency theory (Jensen and Meckling, 1976) states that manager risk-taking
incentives can negatively influence the relationship between shareholders and debt holders. Shareholders can diminish debt holder wealth by taking riskier investments. Incentivizing a CEO to take more risk increases the agency conflict of asset substitution between debt holders and equity holders. Compensation based on equity, like stock and options, can increase agency problems by inducing risk-averse managers to undertake risky investments in order to receive the compensation (Smith and Stulz, 1985). The deregulation in the 1990s caused that banks to allocate more weight in their compensation structure to options bonus, leading to a CEO compensation structure that encouraged risk-taking.
This paper focuses on systemic risk because it’s a sign of economic crises (Acharya et al., 2012). The importance of this research is underlined by Schliefer and Vishny (2010) that stated that the financial sector faces a lot of systemic risk and thereby the recent financial crisis that emphasized the presence of high systemic risk. To test the relationship of CEO compensation structure on systemic risk, a sample of companies between 2004 and 2006 is tested and found no relationship between cash bonuses and systemic risk. To explore the other compensation instruments, we conducted another test in a sample of companies in 2006 and found that CEO options has an increasing effect on systemic risk. Therefore, we concluded that compensation in options strengthened the financial crisis. Previous studies investigated the relationship between performance and CEO compensation. Fahlenbrach and Stulz (2011) stated that lack of alignment of bank CEO incentives with shareholder interest cannot be blamed for the credit crisis. In contrast, studies such as Mehran and Rosenberg (2009) found that the payoff structure of options bonus increases CEO incentives for taking excessive risk and therefore increase the default risk. The link between CEO compensation and systemic risk is on the other hand quite unexploited. Therefore the results of this study could provide valuable information for regulators to determine the degree to which CEO compensation played a part in the crisis and therefore whether policies need to be tightened. It could also give the society a fair image of whether the compensation aspect influenced the crisis. It could help the board of directors of banks set the best
2. Literature review and hypothesis development
Various studies have examined CEO compensation of banks. Below is a review of mostly empirical studies that examined the connection among CEO compensation, excessive risk and systemic risk. This study has probably the most similarities with Kim et al. (2016) and Choi (2014), both of which investigated the relation of compensation structure to systemic risk in the banking industry. The following topics are covered in this review: agency cost, moral hazard, pay for performance, compensation structure and systemic risk. The last section presents the hypotheses.
2.1 Agency cost
In a situation when the CEO is besides the board of director also a shareholder, there are no agency cost. However, if this is not the case a CEO and shareholder interest may differ and the CEO needs to be monitored to keep the CEO’s interest consistent with that of the shareholders. In most cases, the shareholders that own the firm don’t run it. The firm is led by a group of managers who specialize in a way to improve the firm. They are bounded by corporate governance—the system of rules by which a firm is directed and controlled. Corporate governance essentially involves balancing the interests of a company's stakeholders, like shareholders, managers, customers, employees, financiers and
government. Thus, corporate governance provides the framework for achieving a company's objectives. It includes action plans and internal controls to measure performance and
corporate disclosure.
Corporate governance can be divided in two different models: the one and two tier boards. The one-tier board system is when a company is governed by one corporate body that undertakes both the management and monitoring functions. The two-tier board system is when there are two separate bodies that operate independently: the board of directors and the supervisory board. The board of directors controls the daily activity of the firm and decides the strategy for the firm. The supervisory board monitors those actions.
interest alignment between the CEO and the shareholders and maximize the wealth of the shareholders (Jensen and Murphy, 1990). When a firm faces high growth opportunities, it’s harder for the shareholder to monitor the CEO behavior. Therefore, these firms are more inclined to give stock or options to link the CEO compensation to the value of the firm (Harjoto and Mullineaux, 2003). In contract, Holmstrom and Milgrom (1987) states that an interaction between the extent of unobservable effort from the CEO and the diminishing amount of risk that is needed to bear, and forecast the volatility of the pay for performance of CEO compensation is diminishing in the variance of the organizations performance. Overall, the previous literature indicated that CEOs get extra compensation (besides their base salary) to align their interests with the shareholders.
2.2 Moral hazard
Depositors are easier towards shareholders and CEO of a distressed bank and are inclined to stay because of the guarantee of the deposits; when the bank goes bankrupt their money is safe anyway, so why bother? This form of protection leads to excessive risk-taking of a CEO, known as the moral hazard (Kane, 1995). Jeitschko and Jeung (2005) stated that the
guarantee for deposit strengthens the moral hazard problem and they even describe it as an agency problem that increases the incentives of a CEO risk-taking. The depositors don’t mind if the bank is financially unhealthy and only focus on their own financial gain. CEOs in turn know that depositors don’t leave the bank if its financial situation worsens, so they don’t take that into account when incurring in a risky investment. Without a deposit guarantee, CEOs want to assure the depositors that their money is safe with them, so they will incur less risky investments.
This behavior is especially a problem with banks that are distressed or are too big to fail. Shareholders of distressed banks demand more risk from their CEO. Since the bank is already struggling, they incur lower losses if the bank collapses. The too big to fail banks are backed by the government, which will bail them out if they’ve taken to much risk. In contrast, Houston and James (2011) found that bank CEOs kept lower equity-based compensation than nonbanking organizations and that distressed and too big to fail banks didn’t provide larger stock options.
important place in the overall risk a bank is bearing.
Although the moral hazard problem is induced by several factor for banks, this study concentrates on bank CEO’s compensation in relation to risk-taking incentives and the contribution of the bank to the systemic risk. Many researchers, such as Jensen and Mekling (1976), agreed that CEOs’ risk incentives increase when banks have more leverage. This means that risk to the firm’s value is reallocated from debtholders to shareholders. Therefore, risk incentives increase when leverage increases.
2.3 Pay for performance
The goal of this paper is to study whether CEO risk incentives derived from compensation affect a firm’s share to the systemic risk. These incentives are the link between systemic risk and pay for performance. Next of this paragraph there will be a brief summarize of the large amount of empirical studies that study managerial compensation with relation to
performance companies. Hence, the relation to systemic risk is less investigated and therefore valuable to study. Most of these studies that examine the relation between managerial compensation and performance agree that CEO compensation increase CEO risk incentives. Crawford et al. (1995) stated that a large ratio of pay for performance are the same for low and high capitalization banks. Low capitalization banks are the 25% of banks with the lowest capital to the number of outstanding shares multiplied by the share price and high capitalization banks are the top 25%.
According to Smith and Stulz (1985), if the cash bonus grows linearly with the
performance of the bank, the remuneration linked to a compensation plan are non-convex, meaning that they aren’t jointly connected with risk-remuneration. However, if the
performance is beneath the earnings-based boundary at the point that compensation is rewarded, the compensation structure will be a call option on the value of the assets. At that point, compensation structures will be convex and compensate the concavity of the CEO’s risk averse utility function. On the other hand, if the performance of the bank is higher than the boundary for paying compensation, the change of the compensation structure is linear in relation to performance and therefore has no effect on the incentives of the CEO to increase the bank’s risk to gain higher compensation.
terrible realization of their trading strategy. Thereby, banks that rewarded their CEO with a larger cash bonus or banks with larger equity risk incentives didn’t perform poorly. They found evidence that the CEOs didn’t reduce their share before crisis contradicts the argument that CEOs foresee the crisis and only focus on the short term view of the bank. With this study we want to contribute to the existing literature by adjusting the study of Fahlenbrach and Stulz (2011). We changed the independent variable from performance to systemic risk and expanded the control variables to be more appropriate in the analysis. By filtering the different compensation factors, we zoomed in to which component of
compensation mainly influences excessive risk-taking. Giving this information one can conclude that pay for performance, especially bonus in cash, isn’t essential for the risk incentives of a CEO.
2.4 Compensation structure
performance is rewarded (Indjejikian et al., 2011).
According Cheng et al. (2010), banks with large residuary compensation, which is the mean of the compensation of the five best paying CEO, have larger betas, more return volatility and are more at risk than the average bank. These finding are in line with the assumption that heterogeneous CEOs are looking for short-term profit and therefore invest in various project and incentivize to take many risks. However, their results are not in line with the idea that CEO compensation could cause bad governance.
According to Murphy (1999), most cases the stock-bases compensation can be divided in two types. Generally, the composition for equity compensation consists of options and stocks. Options are an agreement that gives CEOs the right, but not the obligation, to earn a pre-specified amount of stocks of the bank. Because it’s a call option, it gives the CEO limited risk and unrestricted potential profit. As mentioned before, this form of compensation can create agency problems. Bad governance can cause CEOs to take excessive risks to earn the stock compensation and benefit them in short run, while the company performs worse in the long run. Option stocks are the obligation for the CEO to buy the stocks of their company for a price which is established in advance. Before exercising the option, the value is based on the underlying asset, time and volatility.
CEOs can impact their compensation volatility factor. Rajgopal and Shevlin (2002) stated that the volatility of a stock price and the return influence the CEO causes convexity in the compensation structure. The stock price influence a CEO on whether to enter into
investments with a positive net present value. Stock return volatility influences the risk-taking incentives when the compensation will increase when risk-taking more risk. Rajgopal and Shevlin (2002) refer stock option as risk incentive effect, giving incentives to CEOs to take more risky projects to increase stock return volatility. If stock return volatility increases, the value of the options will increase as well.
To alleviate these agency problems, there are other ways to give CEOs compensation. One of them is a stock bonus, giving part or all of the compensation in the form of corporate stock. Stocks are generally restricted stocks, which is a form of compensation that is
performance in the long term and might even negatively affect the performance. Giving CEOs lower bonuses in cash instead of more options or stocks results in lower compensation cost (Balachandran et al., 2010). Furthermore, cash-based compensation includes the
standard salary and is based on the firm’s past performance.
Most of the executive compensation in the US consists of compensation based on equity and cash. The compensation based on equity tries to bring the interest of the shareholder and the CEO together. After that the US government decided to deregulate in the 1990s, the standard salary decreased and the bonuses linked to performance increased (Cuñat and Guadalupe, 2007). Chen et al. (2006) argued that the use of options-based compensation has become more widespread after deregulations in the banking industry. Thereby, the structure of CEO compensation, which has an increased option based compensation, lead to an
increase in risk taking in the banking industry. They used four measures of risk: total, idiosyncratic, systematic, and interest rate risks. Fortin et al. (2010) stated that banks that gave their CEOs more stock options and cash bonus in 2005 were more exposed to risk in 2006. In contrast, banks that reward their CEOs with a higher base salary in the same period were exposed to lower risk.
On the other hand, Bebchuck and Spamann (2009) stated that the principal-agent conflict between shareholders of a bank and CEO is embodied to remuneration, therefore the compensation structure has large influence on the risk preferences of managers. If CEOs manage their banks to projects that shareholders appreciate and for which CEO got rightfully compensated, the interest of the CEO and the shareholder were correctly aligned.
Therefore, looking at former papers, the main elements to define CEO risk-taking incentives are base salary, stock and options bonus. It’s crucial for banks to know how these elements can be structured to increase firm value and lower the agency costs.
2.5 Systemic risk
the big bank holds many unprofitable tranches.
To measure systemic risk, there are number of theories that might be suitable for this research. In the past year, many studies on systemic risk produced a number of risk measures, but there is no agreement on which measures are best. Bisias et al. (2012) overviewed the various risk measures. Many previous studies in the last year of risk measurement focus more to the banks’ exposure than to share of the total systemic risk. Acharya et al. (2010) provide a systemic risk measurement model. They call the measure systemic expected shortfall (SES), which can determine a bank’s share in the total systemic risk. This measures a bank’s tendency to be low capitalized when the whole financial world is low capitalized. It shows the bank’s vulnerability to systemic risk. To expand their theory, they developed another measure: long run marginal expected shortfall (LRMES). This is an indicator of the expected equity loss when the market is in its left tail and can predict systemic risk. LRMES can gauge the bank’s share of the overall risk, so it can be used to determine a bank’s losses during an economic depression.
Another theory by Brownlees et al. (2017) involved the systemic risk of a bank as its share of the total capital shortfall of the financial system that can be anticipated in a next crisis. They designed a risk measure (SRISK) that measures the capital shortfall of a firm conditional on a severe market decline as a function of its size, leverage and risk. To test whether there is a relationship between compensation and systemic risk, we use LRMES and SRISK.
2.6 Hypothesis development
Based on the literature in this chapter, this paper analyzes whether a bank’s share in the systemic risk is related to the compensation structure. As mentioned before, there has been a shift of traditional banking to nontraditional banking, where banks generate their income from non-interest activities. Brunnermeier et al. (2012) stated that nontraditional banking is more sensitive to compensation based on performance, which leads to higher systemic risk. Dittman and Maug (2007)stated that to minimize the compensation costs and increase the banks value, compensation should consist of a low base salary and variable compensation based on performance. Similarly, the agency theory states that CEO incentives to take risk, which increases the systemic risk, is as an increasing function of compensation.
H1: CEO risk incentives in 2004–2006 derived from total compensation increase the bank’s contribution to the systemic risk.
H2.1: CEO risk incentives in 2006 derived from cash bonus as compensation increase the bank’s contribution to the systemic risk.
H2.2: CEO risk incentives in 2006 derived from options bonus as compensation increase the bank’s contribution to the systemic risk.
3. Data and Methodology
In this chapter, the research data and the general regression specification are discussed. First, the information about the dataset is provided. To obtain the inputs for the regression, several data sources are used. The regression was used to investigate the impact of
compensation on the bank’s contribution to the systemic risk. Finally, the dependent, independent and control variables are explained.
3.1 Data
To test the hypothesis of the relationship between systemic risk and CEO compensation, a sample was collected from Standard and Poor’s database, with standard industry codes 6000–6300 (except for 6282) in 2004–2006. The appendix A lists the companies in the
sample. The data of systemic risk were obtained from Datastream provide information about returns, leverage, beta’s, volatility and correlation. Data about CEO compensation was mainly generated from Execucomp and provide information about the company and CEO. Unfortunately, the compensation in cash, stock and options bonus were only specified for 2006. Combining the data sets and the sample leaves 61 firms containing 176 firm-year observations, is normally distributed and the sample can be characterized as an unbalanced panel. Table 1 presents the descriptive statistics of the first sample. It shows that the median of the banks’ asset value is $14.3 billion, mean asset value is $127 billion, the average
market capitalization is $17.5 billion and the median is $2.6 billion. The statistics
Table 1 Descriptive statistics over the fiscal years 2004-2006
Variable Number Minimum Lower quartile Median Upper quartile Maximum Mean Standard deviation
Total assets 270.000 1,004.809 6,198.478 14,275.400 53,413.800 1,884,318.000 127,109.900 300,366.600
Total liabilities 270.000 918.376 5,615.386 12,791.400 49,270.020 1,764,535.000 117,924.500 279,618.500
Market capitalization 270.000 48.928 1,105.705 2,589.447 10,642.320 273,691.200 17,452.940 41,088.940
Net income/total assets 270.000 0.028% 0.938% 1.218% 1.465% 2.143% 1.214% 0.409%
Net income/book equity 270.000 0.326% 11.019% 13.780% 16.841% 26.650% 14.048% 5.029%
Cash/total assets 270.000 0.260% 1.571% 2.234% 2.804% 8.171% 2.329% 1.240%
Dividend per share 270.000 0.000 0.432 0.866 1.435 3.544 1.008 0.750
Book-to-market ratio 270.000 0.190 0.388 0.491 0.614 31.573 0.821 2.936
Tier 1 capital ratio 216.000 5.990% 8.410% 9.655% 11.260% 20.540% 9.918% 2.223%
Tangible common equity ratio 270.000 0.000% 4.785% 5.970% 6.964% 34.524% 6.225% 3.311%
Total compensation 268.000 332.865 1,214.970 2,546.633 6,584.957 46,375.350 6954.332 10,046.410
Salary 270.000 0.000 510.000 700.000 975.000 2,866.667 738.847 350.714
Table 2 Descriptive statistics over fiscal year 2006
Variable Number Minimum Lower quartile Median Upper quartile Maximum Mean Standard deviation
Total assets 93.00 1,885.96 6,431.80 14,669.73 57,064.91 1,884,318.00 141,342.60 337,641.100
Total liabilities 93.00 1,753.69 5,760.72 13,365.04 50,783.81 1,764,535.00 131,222.00 314,958.100
Market capitalization 93.00 64.24 1,190.28 2,685.52 12,290.66 273,691.20 19,222.07 45,425.050
Net income/total assets 93.00 0.00 0.01 0.01 0.01 0.03 0.01 0.457%
Net income/book equity 93.00 0.00 0.10 0.13 0.17 0.27 0.13 5.356%
Cash/total assets 93.00 0.00 0.02 0.02 0.03 0.08 0.02 1.231%
Dividend per share 93.00 0.00 0.40 0.81 1.28 3.54 0.97 0.718
Book-to-market ratio 93.00 0.27 0.42 0.50 0.64 31.57 0.87 3.224
Tier 1 capital ratio 80.00 0.06 0.09 0.10 0.11 0.19 0.10 1.984%
Tangible common equity ratio 93.00 0.00 0.05 0.06 0.07 0.22 0.06 2.838%
Total compensation 93.00 332.87 1,211.96 2,403.17 6,761.12 46,375.35 7,563.94 11,482.750
Salary 93.00 0.00 530.00 741.60 978.53 2,866.67 755.67 350.541
3.2 Methodology
To test the relationship, two risk measures are used, LRMES and SRISK (described in Section 2.5). LRMES is necessary in order to calculate the SRISK therefore it will be explained first. Acharya et al. (2012) developed LRMES to explain the average of the fractional returns of the firm's equity during times of distress. This model represents the bank’s marginal expected shortfall in the long run and determines how much risk a certain bank is adding to the total systemic risk. In other words, it can be used to determine a bank’s losses when financial system as a whole is in distress. It measures the amount of capital missing to meet the credit requirements when the entire industry is in distress as a consequence of decreasing market equity. This metric measures whether the amount a financial institution’s equity is below a certain level in a given period when there is economic distress and is based on institution-specific and systemic risk. LRMES is used to place the measurement of risk in a smaller perception, it gives information on corporate level instead of the whole sector.
Knowing how much each bank adds to total risk of the financial world is essential in determining the systemic risk. To calculate the LRMES, a market-based approach was used. Using market data to analyze the stock price market return in period of distress lets the LRMES determine each bank’s share of the systemic risk. This approach indicates that a bank’s share of the systemic risk is based on more than how a bank is connected to system. The size and how levered a bank is are important components of this model. When a bank has enough capital buffer to survive distress, losses won’t be sensible in the system. When the situation of bank is less stable and there is less capital to survive those periods,
bankruptcy will sensible for other banks, thus a larger effect on the systemic risk.
𝐿𝑅𝑀𝐸𝑆(𝑡:𝜏 )𝑖 = 1 ‒ 𝑒(log (1 ‒ 𝑑)𝛽)
where denote the time, reflect the period of crisis, the correction factor and is the 𝑡 𝜏 𝑑 𝛽
CAPM Beta of the firm. According to Brownless and Engle (2012), the correction factor should be 40%.
undercapitalization in a crisis. These firms are considered to have the most impact on the systemic risk.
The next formula represents the SRISK for a bank i:
𝑆𝑅𝐼𝑆𝐾𝑖𝑡= max
(
0,𝑘𝐷𝐼𝑡‒ (1 ‒ 𝑘)𝑊𝑡𝑖(
1 ‒ 𝐿𝑅𝑀𝐸𝑆𝑡:𝑇𝑖)
)
This equation gives the SRISK for company i on day t. K is capital target level, the book 𝐷𝐼𝑡 value of debt and 𝑊𝑡𝑖the market value of equity. A restriction of zero is set because a bank cannot have a negative share of the systemic risk. Based on the previous expression, the total amount of systemic risk in the sector is stated as:
𝑆𝑅𝐼𝑆𝐾𝑡= 𝑛
∑
𝑖 = 1
𝑆𝑅𝐼𝑆𝐾𝑖𝑡
In this expression is n denoted as the total number of firms in the sector. Leading to the relative systemic risk where the risk of firms risk can be determined separately by considering the next formula:
𝑆𝑅𝐼𝑆𝐾%𝑡𝑖=
𝑆𝑅𝐼𝑆𝐾𝑡𝑖 𝑆𝑅𝐼𝑆𝐾𝑡
where SRISK% denotes the proportional contribution of each firm's SRISK to the total positive SRISK of the financial system. As the measure involves the long run, the model can be explained as the capital that a bank would gather to survive a systemic crisis. 𝑆𝑅𝐼𝑆𝐾𝑡 is the capital needed when there is bailout if there is a systemic crisis and 𝑆𝑅𝐼𝑆𝐾%𝑖𝑡 is the percentage of the amount of capital that is needed for a bailout by a bank. To fill the model, the market is assumed efficient, that is, stock prices are always incorporate and reflect all relevant information. This assumption was made because of the important role of equity in this theory.
As mentioned before, CEO compensation consists of a number of variables, like salary, cash bonus, stock and options. We omitted salary because it’s a predefined amount and therefore doesn’t influence the incentives of a CEO. The independent variables reflects the compensation components. The first independent variable is cash bonus, which is the amount in cash (in dollars) a CEO receives when the company reaches certain performance measures. The second is the stock, which is the amount (in dollars) a CEO receives in stocks of the firm as compensation. The last independent variable is options, which is the amount of stocks that the CEO can receive if the options is exercised.
be related to both systemic risk and CEO compensation. Ensure that the control variables do not influence the dependent variables isolated the effect of compensation on systemic risk. Following related literature (Aebi et al., 2011; Chen et al., 2001; Pontiff et al., 1998), the most important control variables were added to model. First, firm size has the most
influence on the CEO compensation. Jensen and Murphy (1990) stated that firm size is highly significant along all explanatory variables in their outcomes. Larger banks have more
financial leverage, because they are expected to have lower costs of financial distress. Larger firms with more influence on the market are more likely be diversified and therefore more resistant for a financial shock. On the other hand, large firms can decide to take excessive risk because the government won’t let them collapse.
To control for firm size and market power, we included total assets as variable. Because the data of this variable is skewed, we used the natural log of total asset. Second, we included Tier 1 capital ratio as control variable to exclude the investments banks and test only the effect of depository bank. Those investment banks have other restrictions and therefore it’s useful to investigate the effect without them. Aebi et al. (2011) stated that when banks held a larger capital ratio, they face a lower debt overhang problem during an economic shock. Third, we included book to market ratio to control for stock returns. Pontiff et al. (1998) stated that book to market ratio can predict future return and that it can link performance to risk incentives. Fourth, Bebchuk and Spamann (2010) stated that executive compensation structure can increase risk-taking at financial firms due to implicit risk-shifting incentives resulting from high leverage. Fifth, the amount of shares held by the CEO
influences his incentives therefore ownership is included as control variable. When a CEO held a low amount of shares of their company the risk incentives could be different than when a large amount of shares is held. Not the compensation of that particularly years plays a part but more their wealth in shares earned in the previous years. Sixth, we control for leverage because it can be a call option of equity. If a bank holds more leverage, it results in an increase of the value of equity, hence equity can be a call option on the value of the assets for levered firms. On the other hand, leverage can also have costs, for example, senior claims that induce the CEO to reduce risk because of the seniority to other stakeholders. Leverage is the debt ratio (debt divided by debt + equity), which shows to which extent the firm makes use of debt. And as last, Chen et al. (2001) stated that stock return volatility can predict firm crash risk, so volatility was used as control variable. The market volatility
variable controls for undiversifiable risk. Based on the literature, we regressed the individual bank systemic risk contribution on the three compensation incentives (independent
To test the hypotheses, we ran two sets of regressions. The first tested the effect of compensation in 2004–2006:
SRISK = 𝛽0+ 𝛽1𝐵𝑜𝑛𝑢𝑠 + 𝛽2𝐵𝑜𝑜𝑘 + 𝛽3𝑇𝑖𝑒𝑟1 + 𝛽4𝑂𝑤𝑛𝑒𝑟𝑠ℎ𝑖𝑝 + 𝛽5𝑇𝑎 + 𝛽6𝑇𝑎2+ 𝛽7𝐿𝑒𝑣 + 𝛽8𝑉𝑜𝑙 + 𝜀
The second set of regressions gives a detailed view of different aspects of compensation in 2006: SRISK = 𝛽0+ 𝛽1𝐵𝑜𝑛𝑢𝑠 + 𝛽2𝑆𝑡𝑜𝑐𝑘 + 𝛽3𝑂𝑝𝑡𝑖𝑜𝑛 + 𝛽4𝐵𝑜𝑜𝑘 + 𝛽5𝑇𝑖𝑒𝑟1 + 𝛽6𝑂𝑤𝑛𝑒𝑟𝑠ℎ𝑖𝑝 + 𝛽7𝑇𝑎
+ 𝛽8𝑇𝑎2+ 𝛽9𝐿𝑒𝑣 + 𝛽10𝑉𝑜𝑙 + 𝜀
4. Empirical results
In this chapter, the empirical results are discussed. We conducted several regressions to test the hypotheses given in Chapter 2. In the first section, the regression is conducted in the sample of 2004–2006. In the second section, a detailed view presented about the
compensation in 2006.
4.1 First sample
The first set of regressions focused on the bonus in cash compensation during 2004–2006 to estimate the impact on the systemic risk variable. Table 3 presents the result of the
Table 3 CEO compensation and systemic risk (1) (2) Bonus in cash ($) -12.768 -10.16 [0.909] [0.891] Ownership shares -0.19 [0.127] Book to market 3.8 [1.082] Market value 6.85 [1.173]
Market value squared -0.35
[0.054] Tier 1 capital -0.12 [0.081] Leverage 2.21*** [0.057] Volatility 62.69** [3.194] Number of obervations 176 176 R-squared 0.0112 0.148 Adjusted R-squared 0.0055 0.128
This table present the effects of CEO compensation on bank’s contribution to the systemic risk. It contains results of regression during the fiscal years of 2004–2006 and consists of 176 banks. The results are from a linear regression. The standard errors are robust and displayed in parentheses. The dependent variable is systemic risk and is the bank’s contribution to the total systemic risk. The first variable is the independent variable, which contains the compensation characteristic of the CEOs. ‘Bonus in cash’ is the annual amount in dollars received as a bonus divided by the total
compensation. Ownership shares controls for a CEO characteristic; a log transformation is applied to this variable. The other control variables contain firm characteristics that might influence the systemic risk. These variables are the Tier 1 capital, log of the total asset, log of the total assets squared, leverage and volatility. Results significantly different from zero at 10, 5 and 1% are denoted by *, ** and ***, respectively.
4.2 Second sample
The second set of regression provided a detailed view of the relation between compensation and systemic risk by adding more compensation variables. While cash bonus represents a short term incentive, options and stock bonuses indicate a longer-term relationship with risk incentives. Table 4 presents the result of the regressions of the three compensation
control determinants (see regression 4 and 5).
The second regression tested the relationship of bonus paid in stocks with the systemic risk. The coefficient is positive, meaning that if stock bonuses increase the bank’s share of systemic risk increase as well. However, this result is insignificant. The third regression tested the relationship of options bonuses and systemic risk. The coefficient is highly significant, so options bonuses have a large impact on systemic risk. The coefficient is positive, meaning that when a CEO gets a higher bonus of options, it will increase the systemic risk contribution of the bank. This measure is robust to the inclusion of other compensation variables and control determinants (see regression 4 and 5). In the fourth column the three variables are regressed together. This regression shows that option is still significant but bonus in cash is not.
In the last regression, we included determinants besides compensation variables to control for other effects. Option remains significant, although with a higher significant level but still with 5%. The coefficient leverage is positive and highly significant. This suggests that when banks increase their leverage, their systemic risk increases. The coefficient of volatility is positive and significant which means that a higher volatility leads to more systemic risk. According to recent literature, non-traditional activities are important at the big banks after deregulation. Therefore, many projects have no incentives with the traditional way of banking, given that compensation in stocks or options encourage CEOs incentives to take project with more risk and align the interest between the CEO and shareholders. So when CEOs are compensated with stocks or options, this could lead them to maximize their own and shareholders wealth. It could even encourage CEOs to pursue nontraditional banking investments to maximize bank value and CEOs own wealth. This type of banking can improve both firm value also the turnover volatility and systemic risk. Therefore, investments that improve banks’ value could lead to a situation that banks become negligent to the systemic risk and choose for high amount of stock and options bonus in their CEO compensation structure, even though it could lead to more risky investments.
Table 4 CEO compensations and systemic risk (1) (2) (3) (4) (5) Bonus in cash ($) -4.896** -3.207 -2.501 [0.215] [0.215] [0.215] Stocks ($) 1.047 1.557 0.836 [0.177] [0.169] [0.161] Options ($) 6.124*** 5.756*** 4.854** [0.198] [0.206] [0.229] Ownership shares -0.115 [0.026] Book to market -0.491 [0.229] Market value -0.02 [0.018]
Market value squared -3.42
[0.258] Tier 1 capital 0.156 [0.012] Leverage 0.419*** [0.013] Volatility 12.224* [0.676] Number of obervations 60 60 60 60 60 R-squared 0.0818 0.006 0.141 0.195 0.4403 Adjusted R-squared 0.066 0.0112 0.1262 0.1519 0.3561
This table present the effects of CEO compensation on bank’s contribution to the systemic risk. It contains results of regression during the fiscal years of 2006 and consists of 60 banks. The results are from a linear regression. The standard errors are robust and displayed in parentheses. The dependent variable is systemic risk and is the bank’s contribution to the total systemic risk. The first variables are the independent variables, which contain compensation characteristics of the CEOs. ‘Bonus in cash’ is the annual amount in dollars received as bonus divided by the total compensation. ‘Stocks’ is the annual amount in dollars received in stocks divided by the total compensation. ‘Options’ is the annual amount in dollars received as options divided by the total compensation. Ownership shares controls for a CEO characteristic; a log
5. Summary and conclusions
The goal of this master thesis was to determine the influence of CEO compensation on a bank’s share in the systemic risk. The literature gave no conclusive answer regarding the CEO incentives and the contribution to risk. Therefore, we conducted empirical tests to
determine this relationship. In the first sample, we couldn’t prove the relationship between bonus in cash with systemic risk. However, we found evidence in the second sample for this variable. Because the second sample is smaller and the variable was only significant if there were no other independent and control variable included, we concluded that this is not strong evidence and therefore does not show a relationship.
We found that compensation provided in options bonuses influenced the contribution of bank to the systemic risk. A higher options bonus leads to higher systemic risk in the fiscal year 2006. This could be explained by the volatility of the options, when the stock price increases the options become more valuable. This could incentivize a CEO to push up the stock price as much as possible with risky investments and therefore earn a higher bonus. A consequence of this procedure is that it most likely increases the stock price in the short run with risky investments that won’t be profitable in the long run and can create financial distress for the bank. Based on the result of the second sample option as compensation helped causing the financial crisis and even made it worse.
As mentioned in Chapter 2, this paper has similarities with Jeong-Bon Kim et al. (2016) and Choi (2014). Those papers used other variables, methods or data, studied the same topic. Jeong-Bon Kim et al. (2016) found the same results as the current study—that options are the main variable that influenced the systemic risk. In contrast, Choi (2014) found no such evidence, so this research should be repeated and expanded.
It’s difficult to establish causality, so the result and the aim of this thesis is to present associations and ideas for follow-up research. Further limitations of this research are the lack of data and time. Having access to a larger database to gain more information about the companies and therefore have more dependent information will strengthen the conclusion. Having more information about the independent variable, to test for stock and options compensation in 2004 and 2005 would also strengthen this research. Other
Bibliography
Acharya, V., Engle, R., & Richardson, M., (2012). Capital Shortfall: A New Approach to Ranking and Regulating Systemic. American Economic Review 102, 59–64.
Acharya, V., Pedersen, L., Philippon, T., & Richardson, M., (2011). Measuring systemic risk. The Review of Financial Studies 30, 2-47.
Aduda, J., (2011). The Relationship Between Executive Compensation and Firm Performance in the Kenyan Banking Sector. Research Paper, Department of Finance and Accounting, School and Business.
Balachandran, S., Kogut, B., & Harnal, H., (2010). The Probability of Default, Excessive Risk, and Executive Compensation:A Study of Financial Services Firms from 1995 to 2008. Working Paper, Columbia University.
Bebchuk, L. A., & Spamann, H., (2009). Regulating bankers’ pay. Discussion Paper 641.
Benmelech, E., Kandel, E., & Veronesi, P., (2010). Stock-based Compensation and CEO (Dis)Incentives. Journal of Economics 125, 1769–1820.
Bhagat, S., & Bolton, B., (2014). Financial crisis and bank executive incentive compensation. Journal of Corporate Finance 25, 313–341.
Biais, B., Flood, M., Lo, A., & Valavanis, S., (2012). A Survey of Systemic Risk Analytics. Annual Review of Financial Economics 4, 255–296.
Black, L. K., & Hazelwood, L. N., (2013). The effect of TARP on bank risk-taking. Journal of Financial Stability, 790-803.
Bolton, P., Mehran, H., & Shapiro, J., (2010). Executive Compensation and Risktaking. Federal Reserve Bank of New York Staff Report No. 456.
Brownlees, C. T., & Engle, F. R., (2017). A Conditional Capital Shortfall Measure of Systemic Risk. ESRB Working Paper Series 37.
Brownless, C. T., & Engle, F. R., (2012). Volatility, Correlation and Tails for Systemic Risk Measurement. NYU-Stern, Working paper.
Brunnermeier, M. K., Dong, G., & Palia, D., (2011). Banks’ non-interest income and systemic risk. Princeton University and Rutgers University.
Brunnermeier, M., & Oehmke, M., (2012). Bubbles, financial crises, and systemic risk. Technical report.
Chen, C. R., Steiner, T. L., & Whyte, A. M., (2006). Does stock option-based executive compensation induce risktaking? Journal of Banking and Finance 30, 915–945.
Chen, J., Hong, H., & Stein, J., (2011). Forecasting crashes: trading volume, past returns, and conditional skewness in stock prices. Journal of Financial Economics 66, 345-381.
Cheng, I. H., Hong, H., & Scheinkman, J., (2010). Yesterday’s heroes: compensation and creative risk-taking. ECGI Working Paper No 285.
Coles, J. L., Daniel, N. D., & Naveen, L., (2006). Managerial incentives and risk-taking. Journal of Financial Economics, Vol. 79, 431-468.
Crawford, A. J., Ezzell, J. R., & Miles, J. A., (1995). Bank CEO pay-performance relations and the effects of deregulation. The Journal of Business 68, 231–256.
Cuñat, V., & Guadalupe, M., (2007). Executive compensation and competition in the banking and financial sectors. Journal of Banking & Finance, 495-504.
Cuomo, A. M., (2009). No rhyme or reason: The ‘Heads I Win, Tails You Lose’ bank bonus culture. New York: New York State Attorney.
Daníelsson, J., & Zigrand, J., (2018, May Monday). HUB. Retrieved from http://www.systemic-risk-hub.org/: http://www.systemic-risk-hub.org/
DeYoung, R., Peng, E. Y., & Yan, M., (2013). Executive compensation and business policy choices at commercial banks. Journal of Financial and Quantitative Analysis Vol 48, 165-196.
Dittmann, I., & Maug, E., (2007). Lower salaries and no options? On the optimal structure. The Journal of Finance 62, 303-343.
Fahlenbrach, R., & Stulz, R. M., (2011). Bank CEO incentives and the credit crisis. Journal of Financial Economics 99, 11-26.
Feng, Y., & Tian, Y., (2009). Option Expensing and Managerial Equity Incentives. Financial Markets, 195-241.
Fortin, R., Goldberg, G. M., & Roth, G., (2010). Bank risk taking at the onset of the current banking crisis. Financial review Vol. 45, 891–913.
Harjoto, M. A., & Mullineaux, D. J., (2003). CEO compensation and the transformation of banking. Journal of Financial Research 26, 341–354.
Holmstrom, B., & Milgrom, P., (1987). Aggregation and linearity in the provision of intertemporal incentives. Econometrica 55, 303–328.
Houston, J. F., & James , C., (2011). CEO compensation and bank risk. Is compensation in banking structured to promote risk taking? Journal Monetary Economics, 405-431.
Indjejikian, R., Matêjka, M., Merchant, K., & Van Der Stede, W., (2011). Earnings Targets and Annual Bonus Incentives. Arizona State.
Jeitschko, T. D., & Jeung, S. D., (2005). Incentives for risk-taking in banking: A unified approach. Journal of Banking & Finance 29, 759-777.
Jensen, M. C., & Murphy, K. J., (1990). erformance Pay and Top-Management Incentives. The Journal of Political Economy 98, 225-226.
Jensen, M., & Meckling, W., (1976). Theory of the firm: managerial behavior, agency costs and ownership structure. Journal of Financial Economics 3, 305–360.
John, K., Mehran, H., & Qian, Y., (2010). Outside Monitoring and CEO Compensation in the Banking Industry. Journal of Corporate Finance 16, 383–399.
Kim, J.-B., Li , L., Ma, L. Z., & Song, F. M., (2016). CEO option compensation and systemic risk in the banking industry. Journal of Accounting & Economics 23, 131-160.
Knopf, J. D., Nam, J., & Thornton, J. H., (2002). The Volatility and Price Sensitivities of Managerial Stock Option Portfolios and Corporate Hedging. The Journal of Finance 57, 801-813. Mehran, H., & Rosenberg, J., (2009). The Effect of CEO Stock Options on Bank Investment Choice,
Borrowing, and capital. Working Paper, Federal Reserve Bank of New York.
Murphy , K. J., (2009). Compensation structure and systematic risk. Working Paper, University of Southern.
Murphy, K. J., (1999). Executive compensation. Handbook of Labor Economics 3, 2485–2563. Murphy, K. J., (2000). Performance standards in incentive contracts. Journal of Accounting and
Economics 30, 245-278.
Nguyen, A. P., & Enomoto, C. E., (2009). The Troubled Asset Relief Program (TARP) and the financial crisis of 2007-2008. Journal of Business & Economics Research 7, 91.
Pontiff, J., & Schall, L. D., (1998). Book-to-market ratios as predictors of market returns. Journal of Financial Economics 49, 141-160.
Rajgopal, S., & Shevelin, T., (2002). Empirical evidence on the relation between stock option compensation and risk taking. Journal of Accounting and Economics 33, 145-171. Reilly, F., & Brown, K., (2009). Investment Analysis and Portfolio Management. South-Western
Cengage Learning.
Shleifer, A., & Vishny, R. W., (2010). Unstable Banking. Journal of Financial Economics 97, 306-318. Smith, C. W., & Stulz, R. M., (1985). The Determinants of Firms’ Hedging Policies. The Journal of
Financial and Quantitative Analysis 20, 391–405.
Appendix
A List of companies Company name
1 ANCHOR BANCORP WISCONSIN INC 41 GREATER BAY BANCORP
2 ASSOCIATED BANC-CORP 42 HUDSON CITY BANCORP INC
3 ASTORIA FINANCIAL CORP 43 HUNTINGTON BANCSHARES
4 BANK MUTUAL CORP 44 INDEPENDENT BANK CORP/MI
5 BANK OF AMERICA CORP 45 INDYMAC BANCORP INC
6 BANK OF HAWAII CORP 46 INVESTORS FINANCIAL SVCS CP
7 BANK OF NEW YORK MELLON CORP 47 IRWIN FINANCIAL CORP
8 BB&T CORP 48 JEFFERIES GROUP LLC
9 BEAR STEARNS COMPANIES INC 49 JPMORGAN CHASE & CO
10 BOSTON PRIVATE FINL HOLDINGS 50 KEYCORP
11 BROOKLINE BANCORP INC 51 LEHMAN BROTHERS HOLDINGS INC
12 CASCADE BANCORP 52 M & T BANK CORP
13 CATHAY GENERAL BANCORP 53 MAF BANCORP INC
14 CENTRAL PACIFIC FINANCIAL CP 54 MARSHALL & ILSLEY CORP
15 CHITTENDEN CORP 55 MERCANTILE BANKSHARES CORP
16 CITIGROUP INC 56 MERRILL LYNCH & CO INC
17 CITY NATIONAL CORP 57 NATIONAL CITY CORP
18 COLONIAL BANCGROUP 58 NEW YORK CMNTY BANCORP INC
19 COMERICA INC 59 NORTHERN TRUST CORP
20 COMMERCE BANCORP INC/NJ 60 PNC FINANCIAL SVCS GROUP INC
21 COMPASS BANCSHARES INC 61 POPULAR INC
22 COUNTRYWIDE FINANCIAL CORP 62 PROSPERITY BANCSHARES INC
23 CULLEN/FROST BANKERS INC 63 PROVIDENT BANKSHARES CORP
24 DIME COMMUNITY BANCSHARES 64 SOUTH FINANCIAL GROUP INC
25 DOWNEY FINANCIAL CORP 65 STERLING BANCORP/NY -OLD
26 EAST WEST BANCORP INC 66 STERLING BANCSHARES INC/TX
27 FANNIE MAE 67 STERLING FINANCIAL CORP/WA
28 FIFTH THIRD BANCORP 68 SUNTRUST BANKS INC
29 FIRST BANCORP P R 69 SUSQUEHANNA BANCSHARES INC
30 FIRST COMMONWLTH FINL CP/PA 70 SVB FINANCIAL GROUP
31 FIRST FINL BANCORP INC/OH 71 SYNOVUS FINANCIAL CORP
32 FIRST HORIZON NATIONAL CORP 72 TCF FINANCIAL CORP
33 FIRST INDIANA CORP 73 TD BANKNORTH INC
34 FIRST MIDWEST BANCORP INC 74 TRUSTCO BANK CORP/NY
35 FIRST NIAGARA FINANCIAL GRP 75 U S BANCORP
36 FIRSTFED FINANCIAL CORP/CA 76 UCBH HOLDINGS INC
37 FLAGSTAR BANCORP INC 77 UMPQUA HOLDINGS CORP
38 FRANKLIN BANK CORP 78 UNITED BANKSHARES INC/WV
39 GLACIER BANCORP INC 79 WACHOVIA CORP
81 WASHINGTON MUTUAL INC 82 WEBSTER FINANCIAL CORP 83 WELLS FARGO & CO
84 WESTAMERICA BANCORPORATION 85 WILMINGTON TRUST CORP
86 WILSHIRE BANCORP INC 87 WINTRUST FINANCIAL CORP 88 ZIONS BANCORPORATION 89 HANMI FINANCIAL CORP 90 MORGAN STANLEY 91 CORUS BANKSHARES INC 92 FIRSTMERIT CORP
B) List of variables
Srisk Capital shortfall of a bank can be explained when there is a systemic crisis
Bonus in Cash($) Bonus as total annual compensation
Stock($) The value of shares in dollar granted in given year as a percentage of total annual compensation
Option($) The value of options granted in given year as a percentage of total annual compensation
Book The ratio of the firm book value compared to the market value.
Tangible common equity ratio Ratio of a firm's tangible equity in terms of the firm's tangible assets. Tier 1 Capital The core capital of the firms.
Ownership shares The ratio of shares hold by the CEO divided by total number of shares
Market value Natural logarithm of total assets
Market value squared Natural logarithm of total assets squared
Leverage Ratio of which extent the firms is financed with debt
