CEO Compensation – The Role of Stock Option Incentive and Risk-Taking.

Incentive and Risk-Taking.

MASTER THESIS

University of Groningen

Faculty of Economics and Business

Master of International Financial Management

June 2016

Camar Cousins S2766892

Supervisor:J. J. Bosma. PhD Co-Supervisor:

2

Abstract

3

1. Introduction

The concern of executive compensation and risk-taking has been a central focus in corporate governance literature over the years. Where it is vastly debated whether the structuring and level of chief executive officer ―CEO‖ incentives to maximize shareholders value in a levered firm encourages unhealthy risk taking and contribute to financial instability by increasing systemic risk. Recent event in the form of the 2007-08 financial crisis, that led public outcry and government intervention, further underline the dispute. As many believe the misalignment between executive risk incentive and firm value fostered an environment that encourage manager to engage in risky business policies, which eventually engenders negative externalities from the bank to the banking sector (Kim, Ma, Li and Song, 2013 and Bebchuk, Cohen and Spamann, 2010). The risk-taking incentives emanates primarily from the executives stock option compensation, a corporate governance mechanism that is considered by experts to better align the interest of management with those of shareholders, as it ties the wealth of executives with the performance of the firm and thereby mitigating the agency-principal

problem (Jensen and Meckling, 1976; Murphy 1999). As executives are awarded stock options, both shareholders and executives benefited whenever stock prices increases. However, the convex payoff structure inherent in stock option compensation provide substantial incentive for risk-taking as executive are fully exposed the upside of risks taken, but are totally isolated from the downside. As a consequence, stock option compensation might induce bank executives to place more emphasis on risky activities in order to inflate stock price and the value of stock their option.

This paper aims at answer the question of whether stock option incentives alter

managerial risk behavior and contribute to systemic risk. Specifically, I examine to what extent the degree CEO vega and delta influences business policy decisions at U.S. commercial

4

the period of 2004-2014. Three market based risk measure (systematic, idiosyncratic and total risk) was used in order to asses‘ executive risk-taking policy. According to Guay and Core (2010) including a large stock option in executive compensation package induces executives to select variance increasing investment and that volatility and stock return variance increase with the growth of their equity portfolio. Moreover, the option incentive of executives‘ contribution to systemic risk ―SRISK‖ formation is explored. SRISK is generally describe as risk caused by an event at the firm level that is severe enough to cause instability in the financial system Acharya, Pedersen, Philippon and Richardson, (2010). Researchers argue as bank executive seek to increase the value of their equity, investing uncertain innovative financial products becomes more attractive, further adding the inevitably of future government bailout executive might tend to aggregate risk-taking know that the consequence will be borne by the firm but the community. Since systemic risk is the risk of the collapse of an entire financial system and compensation structure is affected by firms‘ condition and firm value, the relation between systemic risk and firm value cannot be ignored. Empirical results in this paper indicate that equity compensation serve as a large portion of executives total compensation. However, equity incentives cannot be blamed for inducing risk-taking policies and does not contribute to the overall riskiness of the entire financial system over the course of the sample nor during the financial crisis.

The main research question that this study will address is: to what extent does the

incentive effect of stock option compensation contributes to managerial risk-taking activities and the formation of systemic risk in US banking industry. Data on executive compensation is

5

2. Literature Review.

There has been plenty of research conducted on the relationship between stock option incentives and management risk-taking. However, the results are inconclusive.

Researchers noted over the past two decades there has been an increased reliance on stock option as important component of executive compensation package (Murphy, 1999; Murphy, 2013; Core and Guay, 2010). However, equity compensation has always been under extreme scrutiny. This is generally due to the convex payoff structure inherent in stock option compensation, which provides executives with an incentive to undertake risky businesses and activities (Murphy, 2010). Executives that are awarded stock options benefits from then upside of risk-taking, however they are entirely isolated from the downside of such risk, this relation provides them with the incentive to take on excessive (unhealthy) amount of risk without concern for future consequences. Guay & Core (2010) noted that the 1950s saw the first backlash on whether executive option compensation provides proper incentive to maximize shareholders value, as well as whether these level are fare and appropriate. This is after the early 1900s was driven by the application of equity based compensation, a

phenomenon that continued to increase and later eventually became the largest component of executive compensation in all sector with the exception of utilities (Hall, 2002; Murphy, 1990). In addition, more recent events and scandal such as accounting fraud and option back-dating coupled with suspicions that risk-related incentives led to excessive risk-taking that triggered the financial crisis, has further seen compensation committees and regulators faced with new pay-related laws1, accounting and disclosure rules designed to stem the perceived abused.

The increased reliance on stock option is mainly attributed to pressure to align the interest of managers with those of shareholders, in other words it is a corporate governance instrument used to mitigate the agent-principal problem, that arise when managers take action that are in their own interest at the expense of shareholders (Jensen and Meckling, 1976).

1

6

However, it might not always serve as an efficient alignment of executives‘ interest with the long-term shareholders value. Instead, it may provide executives with the opportunity to reap large amounts of compensation based on short-term gain even at the expense of maintaining an excessively high risk of large losses in the future (Bebchuk et al.2010). Moreover,

alignment incentive may also contributed to riskier policies and ultimately financial instability. For example, in a cross-sectional data investigation conducted by Guay (1999), he found a significant and positive relation between stock option incentives of executive and firm risk, he further added and that stock option incentives are much higher for firms with higher

investment opportunities. In a similar research Hall & Murphy (2003) examine the pay-to-performance sensitivity of options for undiversified risk averse executives and found that executives tend to increase stock price when their wealth are tied with the firm performance. Furthermore, Coles, Daniel and Naveen (2006) show evidence consistent with a strong causal relationship between the structures of managerial compensation and managerial decisions in terms of investment policy, debt policy and risk-taking. Their study pointed out that higher sensitivity of CEO wealth to stock volatility induces additional risk taking in investment policy. Researchers further argues that the positive relation increases more in highly levered firms such as banks, due to executives ability to increase the value of their equity by

increasing the volatility of the asset because their equity are effectively options on the value of the asset (Bolton, Mehran and Shapiro, 2010). For instance, Minnick, Unal and Yang (2011) find that bank executives with higher pay per performance sensitivity are less likely to make value enhancing acquisitions. In a similar study DeYoung, Peng and Yan (2010) find bank executive vega increased substantially prior to the crisis and that vega was positively related market-based risk measures and they chose riskier firm policies.

7

executive incentives during the financial crisis. Their results indicate that pay per performance sensitivity was associated with lower return during the crisis, no relation was found for pay volatility. They argue that this is a result of executives having a strong incentive to maximize shareholders wealth, and thus the crisis cannot be attributed to poor incentives provided by compensation packages. They concluded that executives also suffered when their firms

collapsed, indicating that their risk-taking decisions was mainly due to failure of perceive risks and could not have been a response of incentives to increase risk-taking beyond optimal levels. Jin (2002) also yield similar outcomes, his examination of the relation between CEO‘s

incentive levels and systematic and non-systematic risk show negative results. Specifically, he found a negative association between incentive level and non-systematic risk while controlling for systematic. However, his results did not hold for systematic risk while controlling for non-systematic risk.

Although much of the debate has been focus on the alignment of incentive between managers and shareholders, research has also indicate that equity portfolio sensitivity could lead to more financial instability—the financial crisis underline the argument. Suggesting that a conflict could arise between shareholders wealth maximization and systemic risk. Acharya et al., (2010) defined systemic risk (SRISK) of a financial institution as its contribution to the total expected shortfall (SES) in the event of a systemic crisis. They propose a SRISK measure that captures the expected shortage of a firm given its degree of leverage and Marginal

8

on the relation between executive stock option incentive and systemic risk, with the exception of (Kim et al., 2013 and DeYoung and Huang, 2016). Both studies find strong evidences indicating executive pay sensitivity measured in delta and vega are positively associated with negative externalities.

The lack of consistency and complexity that exist between executive incentives and risk policies have certainly created a concern for more investigation. As such this paper seeks to provide evidence on the topic by examining the relation between executive stock option incentives and market based risk and its contribution to financial instability. Prior research mostly focused a single form of market based risk measure, however this study examine three market based risk measure in addition systemic risk. The research finding will help in the judging the arguments concerning whether stock option compensation provide the proper incentive to maximize shareholders wealth or does it incentivize executive to engage in riskier firm policies that could lead to the detriment of the firm and financial systems as a whole.

3. Hypothesis Development

In this section, hypotheses concerning the central research question will be formulated and an empirical model will be constructed.

Shareholders are the owners of the company and residual claimants to the firms‘ cash flow after it is has satisfied its financial obligations, as residual claimant to the frim cash flow, shareholders naturally would prefers that executives invest in risky value enhancement

9

linking the wealth of executive with firm performance and thus shareholders wealth,

however, the convexity inherent in stock option pay- off structure allows executive to engage in unhealthy risk-taking activities at the expenses of shareholders and reap benefits at no cost. As the payoff structure of stock option is increasing in the underlying stock volatility and stock price, including a large amount of stock option in executive compensation package diminishes the potential risk aversion and make risk more valuable to executives. Chen, Steiner and Whyte (2006) study of 68 banks covering over 70 executive found evidence that support such statement, their study showed that greater sock option compensation induces risk-taking in the banking industry, moreover their study provided limited evidence

suggesting option-based wealth enhance shareholders wealth. In addition, Guay (1999) study of option incentive, found that firm risk is positively related to the convex payoff option incentive provides and that stock option serves as an effective devise in affecting executive behavior towards risky business policies. Therefore;

H1: Option based compensation affect risk-taking. bank risk increases as executive stock option incentive increases.

Furthermore, executive compensation incentive might induced risk-taking that aggravate risk contagion and thus systemic risk.

10

capital relative to asset, knowing the consequence will be borne by society. Therefore, I think that more incentive compensation for executives may result in the manager‘s excessive risk taking for several risky projects. It also lead to higher systemic risk.

H2: Option based compensation affect risk-taking. Executive stock option incentives are positively associated with financial instability and thus systemic risk.

In summary, though including a large amount of stock option grant in CEO compensation package might have reduce the conflict of interest between executive and shareholders by linking manager wealth to the performance of the firm, it might also facilitate mangers engage in activities that are not beneficial to the firm owners. Thus I anticipate executive stock option compensation measured in delta and vega to be positive associated with all forms of risk policies including economic financial instability.

3.1 Model Development

Various factors can be included when examining the incentives to risk relation. In this paper the relation is explored by modeling three market-based risk and SRISK as a function of three incentive measure and several control variables. DeYoung et al.(2010) suggest that the relationship between and among the variables are complex, and as such they should be tested in a fully endogenous and simultaneous system equation:

Risk-Taking = f (Delta, Vega, Bonus/Salary, lnAssets, LnMarketCap, leverage, B/M) (eq.1)

The risk taking variables in (eq.1) measures are obtained from the single index model and are estimated for each year using monthly data from relevant year obtained from Center for Research in Security Price (CRSP) database. The model is given as follow:

Ri = α + βi (Rm) + μit

Where, Ri is the monthly return on bank stock, Rm is the monthly return on the S&P

11

systematic risk or beta (βi). Beta is a measure of the degree of market risk of volatility in

returns on the banks stock that cannot be eliminated by investors by holding a diversified portfolio. Two additional risk measures are computed by calculating the standard deviation of the stock return and the standard deviation of the residuals. They are referred to as total risk (σi) and idiosyncratic risk (σiμ) respectively. These risk characteristics has been used in research as broad market measure that captures the overall riskiness of banks policies and executive incentive. Researchers argue that as the equity compensation of executive increases, the interest of executive and shareholders converge. Consequently, executive inherent the risk appetite of shareholders, who have incentive to increase the risk of the firm, resulting in a transfer of wealth from bondholders to shareholders. In the banking industry bondholders are venerable to such action, as shareholders seek to increase the value of the option like equity by increasing bank risk (Chen et al, 2006). Ultimately, as the stock-based compensation of

executives grows, executive face the same incentives as stockholders and, as such, will pursue strategies that increase bank risk.

SRISK = f (Delta, Vega, Bonus/Salary, lnAssets, LnMarketCap, leverage, B/M) (eq.2)

In addition to market based risk, the relation of executive equity incentive and

systemic risk is explored (eq.2). Systemic risk is generally describe as risk caused by an event at the firm level that is severe enough to cause instability in the financial system as a whole indicating that an externality exists. Acharya et al. (2010) and Brownlees and Engle (2011), define systemic risk as the expected undercapitalization of bank i when the aggregate banking system as a whole is undercapitalized. They SRISK for firm i is defined as:

SRISKi = max (0, Equityi [K Leveragei – (1–K) exp(-18 MESi)])

Where Equityj denotes market capitalization of firm i, K is the minimum capital requirement

12

any given firm (Ri) for these days:

=

∑ t = system is in its 5% tail

Since systemic risk is the risk of the collapse of an entire financial system and compensation structure and equity incentive is affected by firms‘ condition and firm value, the relation between systemic risk and firm equity compensation is explored. (See Acharya et al 2010 and Brownlees and Engle (2011) for further methodology)

The manner in which executive are compensated can shape their incentives to take risk, and delta and vega are two important measures that capture these incentives (Core and Guay, 2002). In order to calculate the equity sensitivity compensation Core & Guay (2002) used the Black-Scholes (1973) formula for valuing European call options, as modified to account for dividend payouts by Merton (1973). Where, vega (pay-risk sensitivity), captures the change in the dollar value of CEO wealth for a 1 percent change in stock return volatility and delta (pay per performance sensitivity), captures the change in the dollar value of

executives wealth for a percent change in the stock price. Including a large amount of stock option grants in CEO compensation packages will typically result in a high vega, which make risk more valuable to executives and diminishes the potential executive risk aversion, whereas, including a large amount of stock grants in CEO compensation packages will result in a high delta, it intended to reduces conflicts of interest between managers and

shareholders by linking manager wealth to the value of the firm‘s stock.

13

highly levered, the leverage ratio across firms are controlled for, Coles et al. (2006) argue that managers are able to change the capital structure to adjust corporate risk.

4 Methodology.

4.1 Data and Sample

The focus of this study is to investigate: to what extent incentive effect of executive stock option compensation contributes to risk-taking activities and the formation of systemic risk. For this Standard and Poor‘s (S&P) ExecuComp2 database is used, as it offers extensive data on executive compensation and is widely used by researchers (Core and Guay, 2002; Murphy, 2010; Fahlenbrach and Stulz, 2011). ExecuComp provides data on executive salary, bonuses, equity portfolio and other compensations. Firm specific data is obtained from Compustat Capital-IQ and stock prices are gathered using the Center for Research in Security Prices (CRSP). Sample selection criteria are based on firms with Standard Industry Classification (SIC) codes between 6000 and 6300. This yield firms that are involve in the lending business as well as others that are not considered typical lending institution, for example security and brokers dealers, finance service, investment advice and investment banks. The latter‘s four were excluded so that only pure bank are in our sample. Furthermore, firms with missing data over three

calendar years were excluded. Also firms for which data is not available in ExecuComp are deleted to avoid bias resulting from differences between complete and missing data. The sample yield a total of 76 banks listed between 2004 and 2014.

4.2 Descriptive Statistics

Table 1 provides a summary statistic of selected variables for the sample of banks between 2004 and 2014. It shows that the sample cover medium, to very large financial institutions. Firms specific variables based on financial statement are listed first with for instance median, mean, standard deviation, min and max observations. The average asset value

2

14

is $133 billion, and the median asset value is $11 billion. Three bank (Bank of Internet Federal, Home Bancshares and Pinnacle Financial Partners) had asset value under $1 billion at certain point in the sample. The average market capitalization for the banks over the periods was $12 billion, while the median was $1.5 billion. The average net income over asset (ROA) is 0.72 percent. Regarding the risk-taking variables, Total risk (σi), measured by the standard monthly stock return for a given year, has an average of 0.084 with a standard deviation of

0.062, whereas idiosyncratic risk (σiμ) has an average of 0.074 with a standard deviation of 0.057. The average systematic risk (βim) is 1.03, while average systemic risk (SRISK) is $3 billion. With respect to other variables which are included as control variables, book-to-market ratio average is 0.93, with a reported standard deviation of 0.98. The average leverage ratio is 2.05, with a standard deviation of 4.93.

Table 2 provides a summary statistic summary statistic of key compensation variables for the sample of 76 bank holding companies for fiscal year 2004-2014. All data are gathered from ExecuComp database. The summary statistic is separated into three sections; annual compensation, equity portfolio and equity portfolio sensitivity. The annual compensation section shows the structure of CEO compensation package, on average CEO‘s earned $751 thousand and has a median of $718 thousand. Bonus which is generally large a component of executive compensation package is on average $1 million, whereas the median is $338

thousand. On average cash bonus was 1.2 times that of base salary. The total average of equity grant which comprises of stock and option grant is $2.45 million, with an average stock grant of $1.51 million. The Black-Scholes (1973) model was used to compute the dollar value of option grant. The average value of option granted is $942 thousand, whereas the median is $366 thousand. For the majority of CEO‘s in the sample compensation emanate from performance based pay, as equity compensation amount to 55 percent of the total compensation. However, there has been a decline in stock and option grant post-crisis,

15

percent of their company shares. Core and Guay (2002) approximation methodology was used to compute the value of previously granted exercisable and un-exercisable options. The

average value exercisable and un-exercisable options were $6.5 and $2.2 million respectively. The value of equity portfolio is relatively large to the total compensation. The average ratio value of the overall equity value divided by the total compensation is 18 for CEOs.

Table 1

Sample summary statistic for fiscal year 2004-2014.

The table shows summary statistic for key variables for a sample of 76 bank holding companies for fiscal year 2004-2014. Sample selection criteria are based on firms with Standard Industry Classification (SIC) codes between 6000 and 6300. This yield firms that are involve in the lending business as well as others that are not considered typical lending institution, for example security and brokers dealers, finance service, investment advice and investment banks. The latter were excluded so that only pure bank are in our sample. A list of the sample will be available in the appendix of the paper. The data are gathered from the Compustat and Center for Research in Security Price (CRSP) database. Beta (βim) is used as a proxy for systemic risk and is calculated by regressing the monthly stock return on the monthly market (S&P 500) return. The standard deviation of residual from the regression provides also a measures of risk; Idiosyncratic risk (σiμ). Total risk is the standard deviation of the stock monthly return. Those data are provided by CRSP, the remaining variables are provided by Compustat. all accounting variables are measured in thousands of dollars.

Variable Number Minimum Lower

Quartile Median Upper Quartile Maximum Mean

Standard deviation Total Asset 76 405,0 5,455.2 11,779.5 50,524.2 2,573,126.0 133,073.8 399,468.3 Total Liabilities 76 373.3 4,849.3 10,105.4 35,500.9 2,341,061.0 106,565.0 344,384.3 Market Capitalization 76 0,0 717.8 1,484.3 4,478.0 283,438.5 12,503.2 36,374.4 Leverage 76 0.0 0.59 1.00 1.72 93.59 2.05 4.93 Book-to-Market Ratio 76 -0.51 0.50 0.73 1.04 11.76 0.93 0.98 ROA 76 -0.07 0.006 0.009 0.012 0.05 0.007 0.01 Beta (βim) 76 -2.91 0.54 0.94 1.44 4.96 1.03 0.83

Idiosyncratic Risk (σiμ) 76 0.02 0.04 0.05 0.08 0.43 0.07 0.06

Total Risk (σi) 76 0.0 0.04 0.06 0.10 0.43 0.08 0.06

Systemic Risk (SRISK) 76 0.0 105,409.2 389.9 1,245.6 178,961.3 3,220.1 17,599.6

16

portfolio value if stock price volatility increases by 1 percent, the dollar expression is $186 thousand on average.

Table 2

Executive compensation and equity for fiscal year 2004-2014.

The table shows summary statistic for key compensation variables for a sample of 76 bank holding companies for fiscal year 2004-2014. All the data are gathered from Compustat Execucomp database. Values are reported in thousands of dollars. The average and median yearly compensation package of CEO from our sample are reported under annual compensation, the Black-Scholes (BS) model was used to compute the dollar value of option grant. The equity portfolio of CEO's in our sample is reported under equity portfolio value, previous granted options (exercisable and un-exercisable) were calculated using the BS model and Core and Guay (2002) approximation. Under Equity portfolio sensitivity Delta ($) refers to the dollar change in the executive's stock and option portfolio value for a 1% change in the stock price. Vega ($) is equal to the dollar change in the executive equity portfolio value for a 1% change in stock volatility. Variables are reported in thousands of dollar.

CEO Mean Median Annual Compensation Total Compensation 4.425,2 2.193,5 Salary 751,3 718,8 Cash Bonus 1.005,0 338,3

Dollar value of annual stock grant 1.509,4 366,9

Dollar value of annual option grant (BS) 942,4 -

Other compensation 217,1 81,0

Cash Bonus/Salary 1,16 0,56

Equity portfolio value

Total value of equity portfolio 66.644,3 13.206,9

Value of shares 55.320,6 8.221,6

Value of exercisable option (BS) 6.506,6 976,3

Value of un-exercisable option (BS) 2.243,6 85,0

Value of unvested restricted stock 2.870,9 503,4

Percentage of ownership from shares 0,8 0,2

Value of total equity portfolio/total compensation 17,7 5,1

Value of shares/salary 68,8 11,9

Equity Portfolio Sensitivity

17 Figure 1. 2004-2014 pay for CEO‘s in 76 US banks.

The figure show he average compensation structure over the sample period. The figure confirm that prior to the financial crisis, 2004-07, stock option indeed was the largest component of executive compensation package reaching a peak of 40 percent in 2006. Post crisis this reduce drastically, reaching a minimum of 4 percent and was replace by stock grant (max 51 percent). Overall executive compensation decrease post financial crisis.

4.3 Empirical Specification

According to the model mentioned in previous section, we are able to construct the two simultaneous equations as follows:

Riskit =

α

α

α

α

α

α

α

α

μ

SRISKit =

β

β

β

β

β

β

18

Where,

Risk = systematic, idiosyncratic and total risk (market based measures risk)

SRISK = the expected undercapitalization of a bank when the aggregate banking system as a whole is undercapitalized.

LnDelta = natural logarithm of the dollar change in the value of equity portfolio for a 1% change in the stock price

LnVega = the dollar change in CEO's equity portfolio value for a 1% increase in stock volatility (Formula in the appendix)

Bonus/Salary = the sum of cash bonus divided by the sum of salary

LnTotalAsset = the natural logarithm of total Asset of the bank for year t

LnMarketCap = the natural logarithm the sum of common shares outstanding multiply by the year end share price of the company stock for year t.

Leverage = the sum of long-term debt and debt in current liability divided by the market value of equity for year t

B/M= book value of equity to market value of equity for year t

19

5. Results and Analysis

In this section the results derived from the regressions are discussed. I examined whether the ratio of bonus to salary, delta and vega induces managers to take on a more risky policies. The dependent variables are Systematic, Idiosyncratic, Total and Systemic Risk.

In table 3, the results of an analysis of the relation between the importance of equity incentive and market based risk are reported. Column 2, 4 and 6 shows the regressions including control variables and the fixed effect of risk measures on bonus salary ratio and equity

20

Table 3

Regression of risk-taking measures on cash bonus and stock option sensitivities.

The table shows results from panel regression of risk-taking activities; systematic, idiosyncratic and total risk on cash bonus, stock option sensitivity measures and control variables. Beta "βim" is defined as the company monthly stock return regressed on the monthly market return. The standard deviation of residual from the regression provides a measures of risk–Idiosyncratic risk "σiμ". Total risk "σi" is the standard deviation of the stock monthly return. Delta ($) is defined as the dollar change in the value of CEO's equity portfolio for a 1% change in the stock price and is the sum unvested stock and un-exercisable and exercisable options. Vega ($) refers to the dollar change in CEO's equity portfolio value for a 1% increase in stock volatility. For both Delta ($) and Vega ($) a log transformation was applied, to avoid extreme value. The percentage change is also included in the regression Delta (%) and Vega (%). The control variables include the natural logarithm of total asset, market capitalization, leverage, book-to-market ratio, all measured over a period of 11 years (2004-2014). Column 1, 3 and 5 report the regression excluding control variable, while column 2,4 and 6 includes control variables. Column 2 and 6 are regress with fixed effects and column 4 with random effects. Standard errors are reported in parentheses. Statistical significance at the 1%, 5% and 10% level is indicated by ***,** and *, respectively.

(βim) (σiμ) (σi)

Independent Variables _{(1) (2) (3) (4) (5) (6)} Cash bonus/salary -0.001 0.007 -0.003*** -0.002*** -0.003*** -0.003*** (0.011) (0.010) (0.001) (0.001) (0.000) (0.000) Delta ($) 0.018 0.018 0.003** 0.002 0.003* 0.002 0.024 (0.023) (0.002) (0.001) (0.002) (0.002) Vega ($) -0.040 -0.044 -0.003* -0.002 -0.003* -0.003 0.029 (0.027) (0.002) (0.002) (0.002) (0.002) Ln (Total Asset) 0.370*** 0.416*** 0.005 0.006 0.012*** 0.013*** (0.090) (0.091) (0.006) (0.006) (0.006) (0.006) Ln (Market Cap) -0.017 -0.017 -0.005*** -0.005*** -0.005*** -0.005*** (0.658) (0.026) (0.002) (0.002) (0.002) (0.002) Leverage 0.011 0.010 0.003*** 0.003*** 0.003*** 0.0033*** (0.008) (0.008) (0.000) (0.000) (0.000) (0.001) B/M 0.193*** 0.190*** 0.016*** 0.015*** 0.021*** 0.019*** (0.039) (0.039) (0.002) (0.002) (0.003) (0.003) Number of Observations 836 835 836 835 836 835 R-squared 15% 37% 23% 38% 26% 38%

21

both measures. This supports the intuition that the degree of risk aversion decreases as the wealth of the company increase. The natural logarithm of market capitalization confirms the intuition at the one percent level significance that the amount of risk tolerance increases with the size of the company. With regard to leverage and B/M the results provides support to previous studies Guay (1999), which showed that leverage and investment opportunities are important determinants of firm risk. Overall, results of the control variables supports the

arguments Fama and French (1992) who showed that the value and size of the company when it comes to risk-taking.

Turning the attention systemic risk and stock option incentive incentives, Table 4 show the results of the analysis over the period of the sample. Column 4 shows the regression

including control variables and the fixed effect of risk measures. First, the coefficients (-0.221) of bonus is significantly negative at the 1 percent level. This finding is contrary with the studies of Kim et al. (2013) that CEO incentives induced from cash bonuses increase firms‘ riskiness and the default risk. Second, regarding the incentives of equity portfolio, the coefficients of delta and vega are insignificant in the model, suggesting that CEO risk incentives, captured by delta and vega, have no significant impact on a bank‘s contribution to systemic crash risk. All which suggest that stock option, might indeed be the solution of providing long-term incentive to mitigate negative externalities. Finally, the results for the controls are generally consistent with those reported by others. Specifically, bank size and book- to-market ratio which are positively associated with a bank‘s contribution to systemic crash risk, while the coefficients of leverage and market capitalization are insignificant…

5.1 Equity Incentives and the financial crisis.

22

the regression including control variables and the fixed effect of risk measures. Cash bonus/salary- ratio shows a positive (0.053) and significant at the one percent level with systematic risk (beta), this provides some indication that leading up to the financial crisis executives may have engage

Table 4

Regression of Systemic Risk on cash bonus and stock option sensitivities.

The table shows results from panel regression of Systemic (SRISK) on cash bonus, stock option sensitivity measures and control variables. SRISK is defined as Systemic Expected Shortfall (SES) as the expected undercapitalization of a bank when the aggregate banking system as a whole is undercapitalized, Acharya et.al (2010).Hence, it is the amount of money this is required to bailout a firm if there was a crisis. Delta ($) is defined as the dollar change in the value of CEO's equity portfolio for a 1% change in the stock price and is the sum unvested stock and un-exercisable and exercisable options. Vega ($) refers to the dollar change in CEO's equity portfolio value for a 1% increase in stock volatility. For both Delta ($) and Vega ($) a log transformation was applied, to avoid extreme value. The percentage change is also included in the regression Delta (%) and Vega (%). The control variables include the natural logarithm of total asset, market capitalization, leverage, book-to-market ratio, all measured over a period of 11 years (2004-2014). Column 1 through 3 report the regression excluding control variable, while column 4 includes them. Fixed effects were used in this regression. Standard errors are reported in parentheses. Statistical significance at the 1%, 5% and 10% level is indicated by ***, ** and *, respectively.

(SRISK) Independent Variables (1) (2) (3) (4) Cash bonus/salary -0.273*** -0.221*** (0.065) (0.061) Delta ($) -0.024 -0.121 (0.041) (0.133) Vega ($) -0.03 -0.231 (0.048) (0.157) Ln (Total Asset) 4.769*** (0.513) Ln (Market Cap) 0.017 (0.146) Leverage -0.012 (0.043) B/M 0.904*** (0.221) Number of Observations 836 836 836 835 R-squared 35% 34% 34% 45%

23

stock option compensation serves as an effective contract in linking executive wealth with those of shareholder. Such contract also expose executives to firm specific risk, idiosyncratic risk, that cannot be diversified without undoing the incentive to exert effort. Therefore it may cause friction between diversified shareholder and undiversified, resulting in a difference in point of view regarding optimal investment and financial decisions and prompt managers to take actions that benefits their personal interest. With regards to delta the results did not showed any significant association with any of the risk measures.

Table 5

Regression of risk-taking measures on cash bonus and stock option sensitivities. Pre-crisis (2006-2007).The table shows results from cross-section analysis of risk-taking activities; systematic, idiosyncratic and total risk on cash bonus, stock option sensitivity measures and control variables. Beta "βim" is defined as the company monthly stock return regressed on the monthly market return. The standard deviation of residual from the regression provides a measures of risk– Idiosyncratic risk "σiμ". Total risk "σi" is the standard deviation of the stock monthly return. Delta ($) is defined as the dollar change in the value of CEO's equity portfolio for a 1% change in the stock price and is the sum unvested stock and un-exercisable and exercisable options. Vega ($) refers to the dollar change in CEO's equity portfolio value for a 1% increase in stock volatility. For both Delta ($) and Vega ($) a log transformation was applied, to avoid extreme value. The percentage change is also included in the regression Delta (%) and Vega (%). The control variables include the natural logarithm of total asset, market capitalization, leverage, book-to- market ratio. Standard errors are reported in parentheses. Statistical significance at the 1%, 5% and 10% level is indicated by ***, ** and *, respectively.

(βim) (σiμ) (σi)

Independent Variables (1) (2) (3) Cash bonus/salary 0.053*** 0.001 0.000 (0.016) (0.000) (0.001) Delta ($) 0.023 -0.002 -0.001 (0.050) (0.001) (0.001) Vega ($) -0.054 0.002* 0.001 (0.060) (0.001) (0.001) Ln (Total Asset) 0.318 0.033*** 0.041*** (0.500) (0.010) (0.011) Ln (Market Cap) -0.438 -0.014** -0.022*** (0.345) (0.008) (0.009) Leverage 0.102 0.003 0.004** (0.085) (0.002) (0.002) B/M -0.680** 0.002 -0.005 (0.294) (0.007) (0.007) Number of Observations 228 228 228 R-squared 48% 61% 62%

24 5.2 Stock Option incentive international implications

The convex pay structure of executive compensation indulge executive with the incentive to implement risky business policies and engage in uncertain investment projects. Moreover, because option value increases monotonically with stock return volatility (Murphy, 2010) executive are more incline to take on non-interest income generating activities engage in innovative financial products and keep high levels of short-term debt and maturity mismatch. These operations are inherently risky and might exacerbate banks specific distress risk and stock price crash, which can eventually trigger sector and/or economy wide financial crisis when banks are contagious Kim et al. (2013). The authors further argues that option-based compensation promote executive risk herding, and herding in risk-taking engender risk contagious and spill over in the financial sector by increasing commonality in risky assets or debt among banks. Thereby, also exacerbating common risk exposure and banks contribution to systemic risk. To examine whether option of executive at US financial firms create such externality an adjustment was made to the systemic risk calculated for domestic US firms. The Adjustment involve adding some Europe of largest financial. These are a total of 68 banks with a mean asset value of $835 billion and a mean market value of $ 28 billion. Specifically I calculated the SISK of European financial institutions over the period of 11 years (2004-2014) same as the sample period. I then calculated the weighted average (WA_SRISK) of European SRISK with those of US sample, this will allow to draw inference on the link between executive option incentive and systemic risk formation both domestic and international.

25

Table 6

Regression of weighted average RSISK (WA_SRIK) on cash bonus and stock option sensitivities. The table shows results from panel regression of Systemic (WA_SRISK) on cash bonus, stock option sensitivity measures and control variables. Delta ($) is defined as the dollar change in the value of CEO's equity portfolio for a 1% change in the stock price and is the sum unvested stock and un-exercisable and exercisable options. Vega ($) refers to the dollar change in CEO's equity portfolio value for a 1% increase in stock volatility. For both Delta ($) and Vega ($) a log transformation was applied, to avoid extreme value. The percentage change is also included in the regression Delta (%) and Vega (%). The control variables include the natural logarithm of total asset, market capitalization, leverage, book-to-market ratio, all measured over a period of 11 years (2004-2014). Column 1 report the regression excluding control variable, while column 2 includes them. Fixed effects were used in this regression. Standard errors are reported in parentheses. Statistical significance at the 1%, 5% and 10% level is indicated by ***, ** and *, respectively

WA_SRISK Independent Variable (1) (2) Cash Bonus/Salary 0.010 0.001 (0.015) (0.010) Delta ($) 0.126*** 0.095 (0.445) (0.439) Vega ($) -0.078** -0.009 (0.066) (0.052) Ln (Total Asset) 0.005*** 0.121*** (0.013) (0.046) Ln (Market Cap) 0.014* 0.082*** (0.006) (0.003) Leverage 0.008 0.015*** (0.620) (0.622) B/M 0.315*** -0.113*** (0.039) (0.401) R-squared 3% 9%

6. Conclusion/Discussion

Over the years there has been constant debate on whether the compensation incentive scheme that is designed to align the interest of shareholders with those of CEO provided the proper incentive. Crash of the financial market in 2007-08 further heighten these debates and created concern for governments regulators to devise new regulations to mitigate the problem. The debate primarily evolve around the convex payoff structure inherent in option

26

could have aggregate risk contagion and systemic risk by directly increasing bank fundamental uncertainties through implementing risky business policies. This research seeks to add to that debate by focusing on how CEO compensation structure would exert impact on managerial risk taking policies and systemic risk. The main independent variable of interest in the CEO

compensation structure are the sensitivity of the CEO wealth to stock return volatility (vega) and the sensitivity of CEO wealth to incremental change in the stock price, measure in (delta). Whereas, the dependent variables are; systematic, idiosyncratic, total and systemic risk. Simultaneous equations approach is adopted to deal with complexity in possible endogeneity between incentives and risk measures. During the study we found that a neither vega nor delta incentives induce CEO of US banks to implement riskier policies over the sample period. From the results we reject the first hypothesis, that stated stock option compensation is positively related to market based risk measures. For the second hypotheses little evidence is obtained to accept the hypothesis that stock option compensation incentive induces executive to engage innovative uncertain financial product that increase the banks default risk and thus creates financial instability. In addition the study does not provide any evidence that option compensation incentive aggregate risk contagion and cause spillover from domestic financial market to the international market. Moreover, prior research suggest that short-term

compensation incentivize managers to implement risky policies, due in part that is allows managers to reap short-term gain at futures expenses, but this study provides no evidence of such impact of over the sample period. Plausible explanation of the would be that, contrary to popular believe stock option may have serve it purpose in aligning the interest of shareholders with those of managers, and may incentivize managers to focus overall wealth creation. Leading up to the financial crisis (2006-2007), however, the results provided some indication stock option incentive might have impacted risk policies. Vega was positively linked

27

Appendix.

For measuring the sensitivity of stock options (Core and Guay, 2002) adoption of Black and Scholes model, modified by Merton (1973) to account for dividends, was used. The

measurements are based on the valuation of a European call option; therefore option value is calculated using:

Option value = [S e-dT _{N(Z) - X e}-rT _{N(Z - σ T}(1/2) _)]

Where Z = [ln(S/X) + T (r - d + σ2 /2)] / σ T(1/2) ,

N = cumulative probability function for the normal distribution S = price of the underlying stock

X = exercise price of the option

σ = expected stock return volatility over the life of the option r = natural logarithm of risk-free interest rate T = time to maturity of the option in years

d = natural logarithm of expected dividend yield over the life of the option

The sensitivity with respect to a 1% change in stock price (Delta) is defined as:

[∂(option value) / ∂(price)] * (price/100) = e-dT N (Z) * (price/100)

The sensitivity with respect to a 0.01 change in stock-return volatility (Vega) is defined as:

28 Company Name

ANCHOR BANCORP WISCONSIN INC HUNTINGTON BANCSHARES

ASTORIA FINANCIAL CORP INDEPENDENT BANK CORP/MA

BANCORPSOUTH INC SIGNATURE BANK/NY

BANK MUTUAL CORP INTL BANCSHARES CORP

BANK OF AMERICA CORP JPMORGAN CHASE & CO

BANK OF HAWAII CORP KEYCORP

BANK OF NEW YORK MELLON CORP M&T BANK CORP

BB&T CORP NBT BANCORP INC

BOFI HOLDING INC NATIONAL PENN BANCSHARES INC

BOSTON PRIVATE FINL HOLDINGS NORTHERN TRUST CORP

BROOKLINE BANCORP INC OLD NATIONAL BANCORP

CAPITAL ONE FINANCIAL CORP PEOPLE'S UNITED FINL INC

CASCADE BANCORP PINNACLE FINL PARTNERS INC

CATHAY GENERAL BANCORP PNC FINANCIAL SVCS GROUP INC

CITIGROUP INC POPULAR INC

COLUMBIA BANKING SYSTEM INC PRIVATEBANCORP INC

COMERICA INC PROVIDENT FINANCIAL SVCS INC

COMMERCE BANCSHARES INC REGIONS FINANCIAL CORP

COMMUNITY BANK SYSTEM INC SCHWAB (CHARLES) CORP

CULLEN/FROST BANKERS INC SOUTHSIDE BANCSHARES INC

CVB FINANCIAL CORP STATE STREET CORP

DIME COMMUNITY BANCSHARES STERLING BANCORP

EAST WEST BANCORP INC SUNTRUST BANKS INC

F N B CORP/FL SYNOVUS FINANCIAL CORP

FIFTH THIRD BANCORP TCF FINANCIAL CORP

FIRST BANCORP PR TEXAS CAPITAL BANCSHARES INC

FIRST COMMONWLTH FINL CP/PA TRUSTCO BANK CORP/NY

FIRST FINL BANCORP INC/OH TRUSTMARK CORP

FIRST HORIZON NATIONAL CORP U S BANCORP

FIRST MIDWEST BANCORP INC UMPQUA HOLDINGS CORP

FIRST NIAGARA FINANCIAL GRP UNITED BANKSHARES INC/WV

FIRSTMERIT CORP UNITED COMMUNITY BANKS INC

FLAGSTAR BANCORP INC VALLEY NATIONAL BANCORP

FULTON FINANCIAL CORP WEBSTER FINANCIAL CORP

GLACIER BANCORP INC WELLS FARGO & CO

GOLDMAN SACHS GROUP INC WILSHIRE BANCORP INC

HANMI FINANCIAL CORP WINTRUST FINANCIAL CORP

29 Company Name

Aareal Bank AG Erste Group Bank AG

ABN AMRO Group NV HSBC Holdings plc

Aldermore Group PLC ING Groep N.V.

Alior Bank Spólka Akcyjna Intesa Sanpaolo S.p.A. Allied Irish Banks p.l.c. JSC VTB Bank

Banca Monte dei Paschi di Siena S.p.A. Jyske Bank A/S Banca popolare dell'Emilia Romagna KBC Group NV

Società cooperativa KBL European Private Bankers S.A. Banca Popolare di Milano Scarl Liberbank S.A.

Banco Bilbao Vizcaya Argentaria S.A. Lloyds Banking Group plc

Banco BPI S.A. mBank S.A. Natixis

Banco Comercial Português S.A. NIBC Holding N.V. Banco de Sabadell S.A. Nordea Bank AB (publ) Banco Espírito Santo S.A. OneSavings Bank Plc

Banco Popolare Societa Cooperativa Scarl Paragon Group of Companies plc Banco Popular Espanol S.A. permanent tsb Group Holdings p.l.c. Banco Santander S.A. PJSC Bank Saint Petersburg

Banif - Banco Internacional do Funchal S.A. Powszechna Kasa Oszczednosci Bank Polski Spolka Akcyjna Bank Pekao S.A. Raiffeisen Bank International AG

Bank Zachodni WBK S.A. Sberbank of Russia OJSC

Bankia S.A. Shawbrook Group plc

Bankinter S.A. Skandinaviska Enskilda Banken AB (publ)

Barclays PLC SNS Bank N.V.

BGEO Group plc Societe Generale Group

BNP Paribas SA Standard Chartered PLC

CaixaBank S.A. Svenska Handelsbanken AB (publ)

Commerzbank AG Coöperatieve Rabobank U.A. Swedbank AB (publ)

Credit Agricole S.A. The Governor and Company of the Bank of Ireland Crédit Industriel et Commercial The Royal Bank of Scotland Group plc

Danske Bank A/S TSB Banking Group plc

Deutsche Pfandbriefbank AG UniCredit S.p.A.

Deutsche Postbank AG Unione di Banche Italiane S.p.A.

Dexia SA Van Lanschot NV

30

Reference.

Aggarwal, Rajesh K., and Andrew A. Samwick, 1999, The other side of the trade-off: The impact of risk on executive compensation, Journal of Political Economy 107, 65–105.

Acharya, V., Pedersen, L., Philippon, T., & Richardson, M. (2010). Measuring systemic risk. Bebchuk, L. A., Cohen, A., and Spamann, H., (2010). The Wages of Failure: Executive:

Compensation at Bear Stearns and Lehman 2000-2008, Yale Journal of Regulation, 27, 257-282. Bolton, P., H., Mehran, and J., Shapiro, 2011. Executive compensation and risk taking. Working paper.

Coles, J., Daniel, N., and Naveen, L., (2006). Managerial incentives and risk-taking. Journal of Financial Economics 79, 431-468.

Core, J., and Guay, W., (2010). Is there a case for regulating executive pay in the financial services industry? Working paper, University of Pennsylvania.

DeYoung, R., E. Y. Peng and M. Yan (2010), "Executive Compensation and Business Policy Choices at U.S. Commercial Banks," Working paper, Federal Reserve Bank of Kansas City.

Fahlenbrach, R., and Stulz., (2011). Bank CEO incentives and the credit crisis. Journal of Financial Economics 99, 11-26

Guay, W., (1999). The sensitivity of CEO wealth to equity risk: an analysis of the magnitude and determinants. Journal of Financial Economics 53, 43–71.

Hall, B. J., and Murphy, J. K., (2003). The Trouble with Stock Options. Journal of Economic Perspective 17: 49–70.

Hall, B., and Leibman, J., (1997). Are CEOs Really Paid Like Bureaucrats? Harvard University. Smith, C., and R. Stulz., (1985). The Determinants of a Firm's Hedging Policies, Journal of Financial and Quantitative Analysis, 391-406.

31

Jensen, M., and Murphy, K., (1990). Performance pay and top management incentives, Journal of Political Economy 98, 225-263.

Jensen, M. C., Meckling, W.H., (1976). Theory of the firm: managerial behavior, agency costs and ownership structure. Journal of Financial Economics 3, 305-360.

Jin, L., 2002. CEO compensation, diversification, and incentives. Journal of Financial Economics 66, 29-63.

Kim, J.-B., Li, L., Ma, M. L., Song, F. M., 2013. CEO option compensation, risk-taking incentives, and systemic risk in the banking industry. Working Paper.

Murphy, K.J., (1990). Performance Measurement and Appraisal: Motivating Managers to Identify and Reward Punishment, Managerial Economics Research Center, Papers 90-10. Murphy, KJ. 2013. Executive compensation: where we are, and how we got there. In Handbook of the Economics of Finance, ed. G Constantinides, M Harris, R Stulz, Vol. 2. pp. 211–356. Amsterdam: Elsevier.