The effect of the option intensity in the compensation package of a CEO on the volatility of the returns of the firm

The effect of the option intensity in the compensation package of a

CEO on the volatility of the returns of the firm

Felix Mooij 11022388 Economics and Business Finance and Organization Junze Sun

29/05/2018

Abstract: In this thesis, we analyze the effect of the option intensity (the fraction of compensation that consists of options) in the CEOs’ compensation package on the volatility of the returns of a firm. Using data on firms listed on the NYSE from 2012 through 2016, we formulate a linear regression model with four control variables: Firm size, the amount of leverage, the R&D intensity, and the firms’ growth rate. A significant positive effect of option intensity on the standard deviation of the firms’ returns is found at the 10% significance level. All control variables are also found to have significant effects on the volatility of the firm.

Statement of Originality

This document is written by Student Felix Mooij who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document are original and that no sources other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

1. Introduction

For many years, stock options have been an integral part of executive compensation within the U.S., and elsewhere. In fact, Fernandes et al. (2010) calculated that stock and stock options accounted for, on average, 39% of total CEO pay. Using stock options to incentivize executives has been criticized for its tendency to stimulate risk-taking behavior. For example, Vermeulen (2009) states that

shareholders should reduce the amount of stock CEOs receive in their remuneration packages. He argues that since the upside potential of options is unlimited but the downside is limited (for the CEO), their use leads to significantly more large losses than large gains. Furthermore, they can also lead to illegal behavior, all in an effort to increase the share price of the firm.

The volatility of the returns of a stock also affects the effectiveness of executive compensation, through its impact on the value of stock options. Due to the way that these options are valued, an increase in the volatility of the returns of the firm will, ceteris paribus, also increase the present value of the firms’ options. In this thesis we will determine whether CEOs use their executive power to increase the volatility of the returns of their firms’ stock, in an effort to increase the value of their own compensation. As such, the research question of this thesis is the following: is there is a

significant positive relation between the amount of stock options the CEO receives, and the standard deviation of the returns of his firm? Our hypothesis is that this positive relationship exists. This assumption is based on the findings of previous literature.

According to the optimal contracting view in corporate governance, executive compensation schemes are primarily designed to align the interests of the principal (the shareholders of the company) and the agent (the CEO) (Murphy, 1999; Hölmstrom, 1979). According to agency theory, both the principals’ and agents’ objectives are to maximize their own wealth. The shareholders’ objective is to maximize the share price. The CEOs might have different objectives than the

shareholders. Since the actions of CEOs are not perfectly observable to the shareholders, CEOs may act to serve their own agenda at the cost of the shareholders’ interests. In the literature, this is known as moral hazard. The key to solving moral hazard problems is to design an incentive scheme that aligns the interests of CEO (the agent) and shareholder (the principal). By granting stock options to the CEO, a portion of his wealth will be tied to the wealth of the shareholders, through the share price of the stock. This is designed to motivate the CEO to act in the best interest of the shareholders, since this also increases his own wealth. However, as was suggested by Agrawal and Mandelker (1987), and Harley and Wiggins (2010), it is possible for CEOs use their executive power to increase the volatility of their firms’ returns. This will increase the value of the CEOs’ options.

In order to study our research question empirically, a dataset is needed. We chose to use a dataset of firms listed on the New York Stock Exchange (NYSE), since this stock exchange contains a large number of firms, in all sectors of the economy. This resulted in a selection of 3012 companies which were active from 2012 through 2016. The sources of the data are the Execucomp and CRSP database, both accessed through the Wharton Research Data Service (WRDS) database. We conduct a linear regression with the option intensity of the firm as the main explanatory variable. Additionally, four control variables are used: The size, the amount of leverage, the R&D intensity, and the growth rate of the firm. The original dataset has shrunk, since not all the of the companies had data publicly available for all of the control variables. Based on research by Berk and DeMarzo (2014), leverage is expected to have a positive effect on the volatility of the firm, whilst size is expected to have a negative effect (Bodie et al., 2014). Additionally, both R&D expenditure and the growth rate of the firm are expected to have a positive effect on the volatility of the firms’ returns, consistent with research performed by Bhagat and Welch (1995), and Smith and Watts (1992), respectively. Through statistical testing, we found a significant positive effect of option intensity on the volatility of the firms’ returns. The control variables also all had significant effects in accordance with previous literature, except for R&D intensity. This control variable unexpectedly had a negative effect on the volatility of the firm.

The rest of this thesis is organized as follows. Section 2 reviews the related literature. Section 3 clarifies some of the terms used in this thesis, and subsequently discusses the methodology. The regression, and the control variables used are also analyzed in this section. In section 4, the statistical results are presented, and their significance is discussed. Finally, section 5 concludes with some possible consequences, and limitations of the research performed in this thesis. Furthermore, some suggestions for future research are also provided.

2. Related Literature

In this section, we briefly discuss the relevant literature on the relation between stock option value and the volatility of stock returns. Besides reporting the general findings of that literature, we also analyze the difference between the approach this thesis takes and previous researchers have taken. In a seminal paper, written by Jensen and Meckling (1976), it is argued that that a manager holding stock options tends to, ceteris paribus, choose the investment with a higher variance. This is theorized to be done in order to increase the value of his options.

Some of the arguments provided by Haugen and Senbet (1981), as well as some of the empirical evidence found by Amihud and Lev (1981), Walkling and Long (1984) and Benston (1985), indicate that the holding of common stock and options have significant effects on managerial incentives. One especially relevant article, written by Agrawal and Mandelker (1987) was used extensively in formulating the expected relationship between stock options as a remuneration vehicle, and the volatility of the firm. In this article, research was performed regarding the effect of option holdings by managers and the firm’s choice of investment and financing decisions. A positive relationship between the option intensity, and firm variance and degree of leverage was found. Both these findings were used in this thesis, the first one being the main research subject, and the second one (between option intensity and the degree of leverage) being converted into a control variable. Another relevant article, written by Harley and Wiggins (2010), is focused on the influence of firm and managerial traits on executive remuneration. Through their research, the authors were led to believe that the usage of options within an executives remuneration package incentivizes these executives to take more risks. One way of measuring this risk-taking by executives is by comparing the investments in R&D to those in Capital Expenditures. Previous research shows that investments in R&D produce more risky and volatile returns than those in Capital Expenditures. As such, the R&D intensity of the firm is included as a control variable in this thesis. Additionally, the researchers also found that using options results in a larger volatility of returns of the firm. This finding is in

accordance with the hypothesis of this thesis.

3. Data Description & Methodology

In section 3.1 some additional information regarding the valuation of options is provided. Section 3.2 then expands on how the standard deviations of the returns of a stock are calculated. Subsequently, in section 3.3, we determine a definition for the total compensation of the CEO.In section 3.4 the regression equation will be formulated, and its variables explained. Section 3.5 elaborates on the origins of the data used. Finally, in section 3.6, some summary statistics will be exhibited.

3.1 Financial call options

A financial call option is the right, but not the obligation to purchase an asset at a predetermined price (the strike price) at some predetermined time in the future. Call options are usually only

exercised when the call option is in-the-money, that is, the current stock price is above the strike price (Berk & DeMarzo, 2014).

According to Murphy (2009), a relatively large part of executives’ compensation packages consists of stock options, and other forms of equity compensation. Fernandes et al. (2010) calculated, using data from 2006 of 1648 U.S. firms, that stock and stock options accounted for, on average, 39% of total CEO pay. Stock options are potentially an effective way to align the interests of managers with firms, and to solve agency problems. This is due the fact that they tie some of the shareholders wealth to that of the CEO (Thomson & Conyon, 2012).

One problem with valuing stock options is that the end-of-contract value is not known at the time that the option is granted. The customary method for determining the price of a call option is known as the Black-Scholes equation (Thomson & Conyon, 2012). The Black-Scholes equation for

determining the present value C for an European call option on a dividend paying stock is as follows:

𝑂𝑝𝑡𝑖𝑜𝑛 𝑉𝑎𝑙𝑢𝑒 = 𝐶 = 𝑆𝑒

𝑁(𝑑

) − 𝑋𝑒

_𝑁(𝑑

)

𝑊ℎ𝑒𝑟𝑒 𝑑

=

𝑎𝑛𝑑 𝑑

= 𝑑

− 𝜎√𝑡

The term C is the value of the call option. S is the stock price when the option is granted, X is the exercise or strike price, t is the time to maturity, r is the risk-free interest rate, q is the dividend yield, and

𝜎

standardized normal variable.

Within this equation, the present value of the option increases when the volatility of the underlying stock increases. This is known as the Vega of the option. It is the derivative of the option value with respect to the volatility of the underlying asset.

3.2 Standard Deviation of the Returns of a stock

The standard deviation of the returns of a stock can be used as a proxy for the volatility of that particular firm.

This thesis focuses on stock options. Since the value of these options depends on the stock price, the method to calculate the risk and return of a certain stock is the same as used by an ordinary investor. Two commonly used measures of risk within finance are the variance and the standard deviation of a stock. The variance is the squared deviation from the mean, and the standard deviation is then the square root of the variance (Berk & DeMarzo, 2014).

The variance of a stock is calculated using the follow formula: 𝑉𝑎𝑟 (𝑅) = 1 𝑇−1 ∑ (𝑅𝑡− 𝑅̅) 2 𝑇 𝑡=1

To calculate the standard deviation, the square root of the variance is taken: 𝑆𝐷(𝑅) = √𝑉𝑎𝑟(𝑅)

Where 𝑉𝑎𝑟(𝑅) is the variance of the return of a particular stock, 𝑆𝐷(𝑅) is the standard deviation, 𝑅𝑡 is the realized or total return from date 𝑇 − 1 to 𝑡, and 𝑅̅ is the average return.

3.3 Total Compensation

The dependent variable in our regression is calculated as the value of stock options as a fraction of total compensation. In order to compute this dependent variable, we need a definition for total compensation. The variable SEC (Securities and Exchange Commission, the regulatory agency tasked with monitoring and regulating the financial markets in the U.S.) Total Compensation is used as the measure of total compensation, which consists of the base salary, any bonus paid within the fiscal year, the present value of any stock awards granted within the fiscal year, any non-equity incentive plan compensation (such as phantom stock, stock appreciation rights, and management carve-outs), and all other compensation (such as pension payments or contributions to the executives 401-K account). This variable is found within the Execucomp database.

Our reasoning behind using these variables is the following: Companies listed in the U.S. are required by the SEC to report their CEOs’ total annual compensation. Additionally, they are also required to report the fair value of any stock options granted within the fiscal year. This fair value is calculated on the grant date of the option. As such, by using two measures which both follow the SEC’s reporting guidelines, the value of the stock options will never exceed the total compensation, and their value is calculated with respect to the total annual compensation.

3.4 Empirical Model

The following empirical model is used:

𝑉𝑜𝑙𝑎𝑡𝑖𝑙𝑖𝑡𝑦𝑖= 𝛽0+ 𝛽1𝑂𝑝𝑡𝑖𝑜𝑛𝑖+ 𝛽2𝑀𝑎𝑟𝑘𝑒𝑡𝐶𝑎𝑝𝑖+ 𝛽3𝐿𝑒𝑣𝑒𝑟𝑎𝑔𝑒𝑖+ 𝛽4𝑅&𝐷𝑖+ 𝛽5𝐺𝑟𝑜𝑤𝑡ℎ𝑖+ 𝜀𝑖 Within this model, the dependent variable is 𝑉𝑜𝑙𝑎𝑡𝑖𝑙𝑖𝑡𝑦𝑖, which measures the volatility of the returns (the standard deviation) of firm 𝑖.

The main explanatory coefficient of the regression is 𝛽1, which is the effect the intensity of options in the remuneration package of the firm has on the dependent variable. The associated independent variable is 𝑂𝑝𝑡𝑖𝑜𝑛𝑖. This variable consists of the amount of stock options, as a percentage of total compensation, of firm 𝑖. These options are valued trough the Black-Scholes method, and calculated according to the guidelines set out by the SEC.

The third variable, 𝛽2, is a control variable. This variable is the size of the firm, as measured by market capitalization. Smaller firms often have more (endogenous) risk than larger firms. Small stocks (companies with a market capitalization at the bottom 20% of the NYSE from 1926 until 2011) had the most variable returns in this period, compared to Treasury Bills, Corporate Bonds and an investment in the S&P 500. The average standard deviation of returns of these small stocks was 39.2%, compared to the S&P 500 with 20.3%, and corporate bonds with an average of 7% (Berk & DeMarzo, 2014). Additionally, small stocks are often less liquid, encounter more difficulty in securing bridge financing, and lack performance history and large amounts of available public information for investors when compared to larger, older companies. By using this control variable, part of this endogenous difference in the volatility of small and large firms is controlled for.

The following three other control variables, are all factors that are all partially within the CEO’s control. As such, it is necessary to include these three factors in the regression in order to accurately measure the CEOs’ possible influence on the volatility of his firm.

The fourth variable, 𝛽3, is the effect the level of leverage has on the volatility of the firm. It is measured as debt divided by debt plus assets. Leverage increases the risk of the equity of a firm. Investors who invest in levered equity require a higher expected return, as compensation for the increase in risk. Furthermore, leverage increases the risk of the equity of a firm, even when that firm has no risk of default (Berk & DeMarzo, 2014).

The fifth variable, 𝛽4, is the effect of the R&D intensity of the firm, as compared to Capital

Expenditures. As was shown by Bhagat and Welch (1995), an investment in R&D contains more risk, when compared to an increase in Capital Expenditures. Additionally, it is also harder to value due to the length of its horizon (Bange and DeBondt, 1998) and contains more uncertainty (Pisano, 1989). The associated coefficient 𝑅&𝐷𝑖, is the 𝑖𝑡ℎ firms R&D expenditure, divided by the sum of the R&D and Capital Expenditures.

The sixth variable, 𝛽5, is the growth rate of the firm. Consistent with previous work by Smith and Watts (1992) and Bizjak et al., (1993) this thesis uses the market-to-book value of assets, 𝐺𝑟𝑜𝑤𝑡ℎ𝑖, which is defined as the market value of equity plus the book value of debt divided by the book value of assets, of the 𝑖𝑡ℎ firm, to proxy for the availability of growth opportunities. The reasoning behind

using this variable is that the outcomes of these growth opportunities are uncertain (Harley and Wiggins, 2001). As such, more uncertainty about the cash flows of the firm might lead to more volatility of the returns in the future.

Variable Coefficients Expected effect on volatility

Standard Deviation 𝑦𝑖 -

Constant 𝛽0 -

Option Intensity 𝛽1 Positive

Market Capitalization 𝛽2 Negative

Leverage 𝛽3 Positive

R&D Intensity 𝛽4 Positive

Growth Opportunity 𝛽5 Positive

Error Term 𝜀𝑖 -

Table 1: Variables used and their expected effects

Some additional evidence for the role the growth rate (𝛽5) and the size of the firm (𝛽2) might have on the standard deviation of the returns of a firm is found by Bodie et al. (2014). They calculate that companies classified as ‘small/growth’ (firms in the bottom half by capitalization and bottom third by book-to-market value) and ‘small/value’ (firms in the bottom half by capitalization and top third by book-to-market value) have an average standard deviation of returns of 29.49, and 28.93

respectively.

In contrast, companies classified as ‘big/value’ (firms in the top half by market capitalization and top third by book-to-market value) and ‘big/growth’ (firms in the top half by market capitalization and bottom third by book-to-market value) have an average standard deviation of returns of 24.08 and 20.93, respectively. For their research, a CRSP dataset with the returns of all U.S. companies listed from January 2000 until September 2012 was used. The differences in volatility between these different types of firms has to be accounted for by the control variables, as it would otherwise cause omitted variable bias.

The final variable is 𝜀𝑖. It is the error term of the regression.

3.5 Data

We collected data on the standard deviation of the stock returns and other characteristics, and the composition of the CEOs’ remuneration packages, of 3012 companies listed on the NYSE. All of the data is from a single source: the WRDS database. This service provides data on different financial markets and on both listed and unlisted companies. The WRDS database contains different sources of data, such as Reuters, Bureau van Dijk and others.

For details on the remuneration packages of the CEOs, we used Execucomp. This is a database containing information on many of the different forms of remuneration that CEOs can receive. For information regarding the firm characteristics, such as the market capitalization or the amount of R&D expense, the Compustat Annual Fundamentals database is utilized. All of the data of the controlling variables is found in this database. Finally, for the standard deviations of the returns of the firms, the CRSP Portfolio Statistics and Assignments database is used. These standard deviation values were computed daily, from the first of January of 2016, until the 31st _{of December.}

Out of the 3012 NYSE listed firms, 1614 remained, after they had been filtered on their fiscal year. In order to use the Execucomp database, which uses the calendar year, only companies with a fiscal year identical to the calendar year are used. Data on the control variables was only available per calendar year. After only selecting firms that granted their CEOs a positive amount of stock options, 545 companies remained.

Of these 545 companies, 437 had the same CEO in 2012 as in 2016. The reason that we used this time spread is that most stock options received by CEOs have a three-year vesting period in which they cannot be exercised. Stock options are used as incentives for the medium term, which is usually three years. Cash bonuses and Long-Term Incentive Plans (LTIP’s) are then used for the short (one to two years) and long (more than three years) term, respectively. The CEOs would normally receive their stock options at the end of the fiscal year 2012, and would then have to wait for three years to exercise them. As such, most of the effects of their actions to increase the volatility of the firm are expected to occur in 2016, as this is the first year that they are able to exercise their options. Data on all of the control variables, and the main explanatory variable was found for every firm, for the calendar year 2016. One exception was the amount of R&D expenditures, as many firms did not actually spend any funds on R&D. One firm was removed as it did not have any data on its stock price for 2016. As such, the final data set consists of 436 companies. Below is a summary of the data used in our regression.

3.6 Summary Statistics

Variables Full Sample Number of Firms = 436

Obs. Mean SD Min. Max.

Standard Deviation 436 0.021 0.013 0.008 0.129 Option Intensity 436 0.373 0.176 0.031 0.928 Market Capitalization 436 1.877 3.958 0.007 3.743 Leverage 436 0.185 0.123 0 0.753 R&D Intensity 436 0.148 0.123 0 0.931 Growth Opportunity 436 1.323 1.230 0.008 1.374

Table 2: Summary statistics

4. Results/Analysis

In section 4.1 the statistical results are shown. Subsequently, in section 4.2 their significance is analyzed. We examine the direction in which they influence the volatility (positively or negatively) in section 4.3. In conclusion, in section 4.4 some possible consequences of the findings of this thesis are discussed.

4.1 Statistical results The statistical results are as follows:

The data values for Market Capitalization were first divided by a factor of 10000 before being used in the regression. All standard errors are robust to heteroskedasticity. Standard errors are reported within parentheses. Since the coefficients of Market Capitalization and Growth

Opportunity are smaller than three decimals, the complete regression is provided below. Table 3: Regression results

This results in the following regression:

𝑦𝑖 = .0149136 + .0076011𝑋1𝑖− .0008398𝑋2𝑖+ .030139𝑋3𝑖− .0062816𝑋4𝑖+ .000698𝑋5𝑖+ 𝜀𝑖 (. 0017426) (. 0038917 ) (. 000168) (. 0060444) (. 0021046) (. 0004076)

4.2 Significance of results

Option intensity is significant at the 10% level, with a p-value of 0.051. All of the other control variables are significant at the 1% level, except for the growth opportunity of the firm. Robust standard errors were used, since it was not likely that the errors within the regression were homoscedastic.

4.3 Discussion of the sign of the regression coefficients

Option intensity has a significant positive effect on the volatility of the firm. This is in line with the hypothesis outlined in this thesis. It could be argued that the CEO would use the power at his disposal (e.g. by initiating takeovers, or by investing in riskier projects) to raise the volatility of his firm, thereby increasing the present value of his stock options. In fact, according to an influential article by Jensen and Murphy (1990), this was one of the main reasons for the increase in variable compensation after the 1990s, when shareholders wanted CEOs to be paid less like bureaucrats. Before this increase, executives’ compensation depended more on the stability of operating incomes than on the stock value of the firm.

Market Capitalization having a negative effect on the volatility of the returns of the firm was expected. On average, larger firms are inherently less volatile. Leverage having a positive effect was also expected, since a higher degree of leverage increases the risk of the levered equity of the firm. Growth Opportunity having a positive effect on volatility was also within expectations, as a firm which relies more heavily on future investment projects faces more uncertainty about its future cashflows.

R&D expenditure was significant, but its negative effect on firm volatility was unexpected. We expected that if firms spent more on R&D, they would expose themselves to more risk. This

expectation was based on previous literature, mainly Harley and Wiggins (2001). In their article, they reason that R&D expenditures increase the firms’ dependence on future (uncertain) cashflows. The statistical results show that an increase in the amount of R&D expenses, as compared to Capital Expenditures, would actually predict a decrease in volatility. No readily apparent explanation for this anomaly was found.

The central question in this thesis is if the value of the stock options that a CEO receives, positively affects the volatility of the returns of that particular firm. According to the statistical results, there is a small but significant (10%) positive effect of option intensity on the dependent variable. One possible consequence of this result is that investors might be more or less inclined to buy shares of a firm that rewards its CEO with a relatively large amount of stock options. If the CEO does receive a large amount of this form of renumeration, he might be incentivized to take actions that increase the volatility of the stocks’ returns. This inclination also depends on the investors’ risk preferences. Additionally, compensation committees might also take into account that their remuneration package design might incentivize the CEO to take more or less risk. This level of risk taking would depend on, among other factors, the fraction of stock options in the remuneration package. Since many of the world’s largest and most important firms use stock options to reward their executives, better understanding how managers are incentivized to take certain (risky) actions, is important.

5. Discussion & Conclusion

In this chapter, we analyze some of the limitations of the research performed in this thesis. Additionally, we also provide some suggestions for future research.

As is often the case with research performed outside of a controlled experiment, it is difficult to verify how much of the variation in the dependent variable can be ascribed to the independent variables. CEOs do borrow funds for business reasons, not exclusively to increase the present value of their options. They do invest in R&D not solely because this will increase their own wealth, but also because in their industry heavily investing into the development of new products might be the only way to stay competitive. Further research might incorporate additional variables to try and control for these firm or industry characteristics.

Additional research might also involve including some control variable for the risk aversion of these CEOs, as this might have a significant effect on their risk taking. A highly risk averse CEO values the increase in the present value of his stock options much less than a risk-loving CEO. As such, we expect that the option intensity will have a smaller effect on the volatility of the firms returns, when a firm is managed by a risk averse CEO. An accurate proxy for risk aversion could likely explain some of the remaining omitted variable bias.

Another issue within the complicated framework of valuing options, is the fact that stock options granted to employees are different to the regular options traded by investors. Since employees usually only receive options from the firm they work at, which are restricted in the sense that they

cannot sell them on the open market, diversification becomes practically impossible. As such, employees are exposed to not only the systemic risk, but also the idiosyncratic risk of the stock. This is theorized to lead to a reduction in the perceived (and thus true) value of stock options for

employees, especially those who are relatively more risk averse (Hall and Murphy, 2002). If this is true, this would imply that the actual effect of option intensity on volatility might be larger than is shown by the main explanatory variable, 𝛽1. This increase in the effect of option intensity is due to the fact that the perceived value of the options is lower but the increase in volatility is unchanged. In conclusion, through our research we have concluded that the option intensity does have a positive effect on the volatility of the returns of a firm. However, additional research should incorporate the suggestions for further research, in order to provide a more realistic depiction of how options affect risk taking within corporations.

6. References

Agrawal, A., Mandelker, G.N. (1987). Managerial Incentives and Corporate Investment and Financing Decisions, The Journal of Finance, 42 (4), 823-837.

Bange, M., DeBondt, W. (1998). R&D budgets and corporate earnings targets, Journal of Corporate Finance, 4, 153-184.

Bebchuk, L.A. and Fried, J.M. (2004). Pay Without Performance: The Unfulfilled Promise of Executive

Remuneration, Harvard University Press: Cambridge, MA.

Benston, G. (1985). The Self-Serving Management Hypothesis: Some Evidence, Journal of Accounting and

Economics, 7, 67-84.

Berk, J., DeMarzo, P. (2014). Corporate Finance. Essex, England: Pearson Education Limited.

Bhagat, S., Welch, I. (1995). Corporate research and development investments: international comparisons,

Journal of Accounting and Economics, 19, 443-470.

Bizjak, J., Brickley, J., Coles, J. (1993). Stock-based incentive compensation and investment behavior, Journal

of Accounting and Economics, 16, 349-372.

Black, F., Scholes, M. (1973). The Pricing of Options and Corporate Liabilities, Journal of Political Economy, 81 (3), 637-654.

Bodie, Z., Kane, A., Marcus, A. J. (2014). Investments. Berkshire, England: McGraw-Hill Education Limited.

Fernandes, N.G., Ferreira, M.A., Matos, P.P. and Murphy, K.J. (2011) Are US CEOs Paid More? New International Evidence (October 12, 2012). EFA 2009 Bergen Meetings Paper; AFA 2011 Denver meetings Paper; ECGI – Finance Working Paper No. 255/2009.

Gormley, T. A., David, A. M., Milbourn, T. (2013). CEO compensation and corporate risk: Evidence from a natural experiment. Journal of Accounting and Economics, 56, 79-101.

Hall, B., Murphy, K. J. (2002). Stock options for undiversified executives. Journal of Accounting and

Economics, 33(1), 3-42.

Harley, E.R., Wiggins, R.A. (2000). The influence of firm- and manager-specific characteristics on the structure of executive compensation. Journal of Corporate Finance, 7(2), 101-123.

Haugen, R., Senbet, L. (1981). Resolving the Agency Problems of External Capital through Options, Journal of

Finance, 36, 629-647.

Hölmstrom, B. (1979). Moral hazard and observability, Bell Journal of Economics, 10(1), 74-91.

Jensen, M. C., Murphy, K. J. (1990) Performance pay and top-management incentives, Journal of Political

Economy, 98 (2), 225-264.

Jensen, M., Meckling, W. (1976). Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure, Journal of Financial Economics, 3, 305-360.

Merton, R. C. (1973). Theory of Rational Option Pricing, The Bell Journal of Economics and Management

Science, 4(1), 141-183.

Murphy, K. J. (2009). Compensation Structure and Systemic Risk, Marshall School of Business Working Paper No. FBE 34-09.

Murphy, K. J. (1999). Executive Compensation, in Ashenfelter, O. and Card, D. (eds.), Handbook of labor

Economics, Volume 3, 2485-2563, North Holland, Elsevier: Amsterdam.

Pisano, G. (1989). Using equity participation to support change: evidence from the biotechnology industry,

Journal of Law, Economics, and Organization, 5, 109-126.

Ryan, H. E. Jr., Wiggins, R. A. (2001). The influence of firm- and manager-specific characteristics on the structure of executive compensation, Journal of Corporate Finance, 7(2), 101-123.

Smith, C., Watts, R. (1992). The investment opportunity set and corporate financing, dividend and compensation policies, Journal of Financial Economics, 31, 263-292.

Stock, J. H., Watson, M. W. (2015). Introduction to Econometrics. Essex, England: Pearson Education Limited.

Thomson, S., Conyon, M. (2012). Corporate Governance: Mechanisms and Systems. Berkshire, England: McGraw-Hill Education Limited.

Vermeulen, F. (2009, April). Why stock options are a bad option. Harvard Business Review.

Walkling, R., Long, M. (1984). Agency Theory, Managerial Welfare, and Takeover Bid Resistance. The Rand

Journal of Economics, 15, 54-68.

Y. Amihud and B. Lev. (1981) Risk Reduction as a Managerial Motive for Conglomerate Mergers, The Bell

Appendix

Variable Descriptions:

Option intensity: Value of options (Black-Scholes) divided by the total compensation (calculated under the SEC guidelines) in the compensation package of the CEO of the firm in the fiscal year 2012 of the firm.

MarketCap: Calculated as the amount of shares outstanding, multiplied by the share price, at the end of 2016.

Leverage: Calculated as the amount of debt of the firm, divided by the amount of debt plus assets at the beginning of 2016.

R&D: Calculated as the amount of R&D expenses, divided by the amount of R&D expenses plus the amount of Capital Expenditures, over 2016.

Growth: Calculated as the market value of equity plus the book value of debt divided by the book value of assets, over 2016.

Dependent Variable: The volatility of the firms returns, calculated daily over the year 2016. List of firms used: 436 NYSE tickers

ABBV ABG ABT ABX ACC ADS AEE AEL AEP AES AET AGCO AHC AIN AIZ AKS ALB ALE ALEX ALK AME AMN AMP AMT AON AOS APC ARNC AROC ARW ASB ASGN ATI ATR AVA AVB AVD AVP AVY AWR AXE AXL AXP B BA BANC BAS BC BGS BHE BIO BK BKH BLD BLL BMI BMS BOH BSX BWA BXS BYD C CBM CBRE CBS CBU CCI CCK CDE CF CFR CHE CI CKH CL CLB CLD CLW CMA CMI CNA CNC CNX COF COG COP CPE CPF CR CRI CRK CRL CRR CSL CTB CTL CUBI CVG CVS CVX CW CWT CXO CXW CYH D DAL DAR DF DGX DHR DHX DIN DNOW DNR DOV DPS DPZ DRQ DTE DUK DVA DVN DWDP EBS ECL EDR EE EFX EIX ELY EME EMN EOG EQT ERA ES ETN EVHC EVR EW EXC F FAF FBHS FBP FCFS FCX FE FHN FIS FIX FLO FLR FLS FMC FNB FNF FOE FSS FTK GATX GCI GD GEO GLT GLW

GM GPC GPI GS GVA GWR GWW GXP HAL HBI HE HES HFC HII HL HLX HOG HOS HPR HSY HUBB HVT HYH IBM ICE IDA IEX INGR INT IO IP IPG IPI IR IT ITGR ITT ITW JNJ JNPR K KEX KEY KMB KN KND KO KOP KRA KS KSU KWR LCII LDL LEA LEG LH LII LLL LMT LNT LPX LUV LYB LYV MA MAN MATX MCO MD MGM MHK MLI MLM MMC MMM MO MPC MRK MS MSA MSCI MSI MTB MTDR MTG MTH MTRN MTX MTZ MUR MXL NBL NC NCI NCR NE NEE NEM NEU NFX NI NLS NOC NOV NP NPK NPO NR NRG NSC NSP NUE NUS NWE NWL NWN NYCB NYT OC OFG OGE OGS OII OIS OKE OLN PB PCG PEG PES PFE PFS PGR PHM PII PKD PKG PKI PM PNM PNR PNW POL PPG PPL PRA PSX PWR PX PXD QHC R RBC RCL RDC RDN RF RGA RGR RHI ROG ROL ROP RRC RS RSG RTEC RYAM SBOW SCG SCI SCL SEE SF SHW SIX SJI SLCA SM SMP SNA SNV SO SON SPN SRE SSD SSP STI STT SWK SWM SWN SWX SXC SXT SYK T TAP TBI TCF TDS TDY TEN TER TFX TGNA THC THS TKR TMK TMO TNC TPH TREX TRN TROX TSS TTI TUP TWX TXT UFS UHS UNH UNM UNP UNT UPS URI USG USPH UTX UVE VAC VFC VLO VLY VMC VMI VSH VVC VVI VZ WAB WBS WCN WD WDR WEC WEX WFT WHR WLH WM WMB WNC WR WSO WWW X XEC XOM XRX Y ZBH ZTS