Impact of Executive Stock Option Sensitivity
to Stock Volatility on CEO’s Risk Taking
Behaviour
MSc Accountancy & Control, variant Control
Faculty of Economics and Business, University of Amsterdam
Author:
Obeydollah Naseri
Student number:
10579184
Date:
22-06-2015
Words:
11,095
Statement of Originality
This document is written by student Obeydollah Naseri who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is 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.
Abstract
This thesis provides causal relationship between executive stock option (ESO) sensitivity to the stock price volatility, Vega, and CEO’s operational decision-making, controlled by the sensitivity of the CEO’s wealth to the performance, or Delta. The research finds a positive relation between the increase in Vega and research and development expenditure and the opposite relation of Vega with capital expenditure. This result indicates increase Vega incentivizes CEOs to implement risky investment policy. The result of the paper also shows that higher Vega causes increase in the leverage, which indicated a risky financial policy of CEOs.
Table of Contents
1. Introduction 5
2. Literature Review & Hypothesis Development 8
2.1 Employee Stock Options (ESO) 8
2.2 Managerial Risk Taking Behaviour 11
2.3 Hypothesis 12
3. Methodology 15
3.1 Sample 15
3.2 Empirical Method 16
3.2.1 Investment Policy Model 17
3.2.2 Financing Policy Model 19
4. Descriptive Analysis 19
4.1 Descriptive statistics 19
5 Multivariate Analyses 24
5.1 Investment policy 24
5.1.1 Research and Development Expenditure 24
5.1.2 Capital Expenditure 25
5.2 Financial Policy 26
5.3 Robustness Analyses 28
5.3.1 Regression of deleted missing R&D values 28 5.3.2 Regression of top 25th percentile of Vega 30
6. Conclusion 31 Appendix A 32 Appendix B 33 Appendix C 34 7. Reference 35
1. Introduction
A significant number of academic literatures have examined how the incentives provided through compensation contracts affect managerial behaviour. In order to avoid principal-agent problems within a firm, many shareholders enter into a contract with CEOs to align their interests (Jensen & Meckling, 1976). Jensen & Meckling (1976) explain in their paper the relation between the agent (e.g., CEO, executives) who performs services on behalf of other actor involved in a firm, the principal (shareholders, debt holders). As Shareholders cannot monitor the actions of the CEOs all the time, therefore, they provide CEOs with incentives to maximize the shareholder value, often done by providing CEOs stock compensation. However, this does not solve all the agency conflicts within a firm. Another problem of misalignment of interests can take place, when CEOs hold a large number of stocks in one firm, while shareholders have invested their wealth in different companies and operations. The portfolios of CEOs are undiversified as opposed to those of the shareholders, this will incentivize the CEOs to be risk averse and avoid projects, which could be profitable for shareholders but risky. To alleviate this, stock option executive stock option (ESO) is provided to the CEOs. According to Guay (1999) the CEOs are incentivized to take more risk when they are awarded with stock option compensation. When the wealth of a CEO is tied to stocks of the firm, and the CEOs' personal wealth will move in line with the firm's value. But when a CEO is provided with stock options, his wealth will increase if the stock price increases, but will remain unchanged if the stock price of the firm drops. Thus CEOs will then be incentivized to take more risk. (Knopf et al., 2002; Rajgopal & Shevlin, 2002)
Besides the alignment of the interest of the CEOs with shareholders, there are also other unwanted impacts of stock option compensations on a firm or CEO’s behaviour, such as the effect ESOs have on opportunistic behaviour of the CEOs. Also Lie (2005) raised his concerns about the vulnerability of the ESOs and the opportunistic behaviour of CEOs regarding timing of the award. He explains that ESOs are granted with a fixed exercise price equal to the stock price on the award date. If a CEO can influence the a date of grant, then the CEO might choose to have the options granted at a time when the stock price is expected to peak or dip thereby generating options with an abnormally profitable strike price. All these cases become more obvious in situations in which a stock price of a firm decreases before the stock option grants or stock price of the firm increase afterwards, which clearly shows the self-serving behaviour of a CEO. Another impact of the stock option compensation, according to Lowry and Murphy (2007), is the fact that these compensations provide opportunities for powerful CEO’s to maximize their own earnings by creating their own compensation package and by timing the option awards. They state in their paper, that most of the options parameters are provided by the CEO’s to the compensation committee’s, and subsequently one is chosen. So having the CEOs provide option parameters to the compensation committee will give them the opportunity to maximize their personal wealth and not the firm's wealth.
After examining the possible impacts of the stock option compensation discussed by several authors, I have also reviewed papers related to factors influencing managerial risk taking behaviour other than stock option compensation. According to Lewellyn et al. (2012) there is a positive relation between the power of CEOs position and their risk taking behaviour in a firm. They argue in their study that a powerful CEO would work in their own best interests rather than that of the firm since a larger amount of power would allow their influence in both the operations and decision-making process of the firm.
Akin to prior literature ( Guay, 1999; Knopf et al., 2002; Rajgopal & Shevlin, 2002; Huang et al., 2013), this study is interested in the impact of the stock option compensation, provided to CEO’s, on their risk taking behaviour. As already mentioned above, CEO’s are awarded with stock options to incentivize them to have more risky policies in a firm. Similar to the research of the Rajgopal & Shevlin (2002) and (Huang et al, 2013), I am looking at the CEO’s policies regarding investment and financing in a
firm. According to Huang et al (2013) there are two types of prominent investments in a firm, Research and Development (R&D) and Capital expenditure. In their research they explain that R&D expenditure has riskier nature than capital expenditure, since it involves higher costs and higher uncertainty of its success compared to capital expenditures. They conclude in their research that CEO stock option compensation incentivizes them to increase R&D expenditure and decreases the capital expenditure. They also found that higher leverage in a firm is influenced by the provision of higher stock option compensation to the CEO’s. Accordingly Coles et al. (2006) provide an explanation for these behaviours of CEO’s; by looking at the sensitivity of the CEO’s wealth to the stock price and stock return volatility. They explain that the sensitivity of CEO’s wealth to stock price, also called delta, aligns the incentives of CEOs with that of the shareholders. On the other hand, stock option compensation of CEOs increases the sensitivity of their wealth to the stock return volatility, also called Vega. Similarly, this research is looking at whether sensitivity of the stock option compensation to the stock return volatility, Vega, is positively related to the risky investment policy, increase in R&D investment, and risky financial policy, increase in book leverage.
Prior conducted studies have researched the impact of sensitivity of ESO’s to stock price and stock return volatility on risk taking behavior of managers in different industries or different markets with different periods. Authors’ such as, Rajgopal and Shevlin (2002) researched stock option compensation provided to managers and the incentives of these compensation to invest in risky projects, for oil and gas companies for the period of 1992-1997. Another paper, Coles et al. (2006), provide strong causal relation between managerial compensation and investment and debt policy in period of 1992-2002. However, my paper is going to research the impact of the sensitivity of stock option compensation to stock return volatility on the investment policy and financial policy of the CEO’s in firms listed in S&P 1500 in the period of 2009-2013. My research is different than previously researches mentioned in several factors. First, the period I have chosen is just after latest collapse of stock market in 2008. I am hoping to see whether there are changes in impact of sensitivity of ESO’s to stock return volatility on risk taking behavior of CEO’s. Second, unlike Huang et al. (2013) & Rajgopal and Shevlin (2002), my research is based on a broad population. I have gathered data from all
firms listed in S&P 1500, while, the other two papers researched firms within a specific country, such as Huang et al (2013), or within a specific industry, such as Rajgopal and Shevlin (2002). By doing this particular research my study adds to the stream of literature of impact of sensitivity of ESO’s to stock return volatility and determinants of the risk taking behavior of the CEO’s. This paper should attract attention Therefore the aim of this paper is to produce findings that might be interesting and have practical implications for CEO’s and shareholders and other actors involved in the strategic decision making process of the firm.
2. Literature Review & Hypothesis Development
2.1 Employee Stock Options (ESO)
According to Garen (1994) executive compensation is a classical solution of the principal-agent problem in which the separation of ownership and control is the main issue. He explains in his paper that in every business there are different parties involved with different interests and goals, and the theory describes the relationship between two parties, principal and agent, where the principal passes the task to the agent. In this case the CEO is the agent and the principal is the shareholder of a firm. The attitude of the CEO in a firm is different than that of the shareholders; therefore shareholders provide equity compensation and other awards to align the interests. According to Jensen and Meckling (1976) the reduction of the agency conflict is best tackled by tying the wealth of an agent with the firm equity or performance of stock price. Implementing a compensation policy in which the CEO wealth is tied to the firm, the performance of CEO and creation of value for the firm can be managed and thereby the CEO wealth and stock price is managed. Also Gerhart et al. (2010) explain in their research that it is crucial to have alignment of interest, such as alignment of financial interest and alignment of other actions, between agent and principal, so that after the provision of equity compensation to the agents, the agents’ interest spontaneously align with those from the owners, even though it is still motivated by their own interest (Gerhart et al., 2010).
Balsam and Miharjo (2007) state in their research that the solution for the aforementioned agency problem is the equity compensation. There are several types of equity compensations with variety of impacts on the agents. The stock compensations and stock options compensations are the most prominent ones. The stock compensations are awards provided to the agents in stocks of the firms, in this way the shareholders reward agents by awarding them a part of the equity of the firm. Balsam and Miharjo (2007) have the belief that stock compensation provides direct link between executive compensation and shareholders wealth and therefore the interests of a firm’s CEO’s with those of its shareholders. Therefore, this type of executive compensation is very useful in aligning the principal and agent’s interest. However, stock compensation has negative side effects, such as an increase in risk aversion of CEO’s. By providing more stocks to a CEO, the risk attitude of CEO in a firm becomes different than that of the shareholders: whereas CEO’s will be loyal with most of their capital to their corporations trying to avoid risk while shareholders aim is to maximize their gains, and prefer more risk taking operations. Therefore, it is very important to provide the right executive compensation to motivate agent (CEO’s or executives) to act in the best interest of the principal (shareholders, debt holders) to mitigate the agency problem created by the provision of the stock compensation.
According to Low (2009) the managerial risk aversion is a serious agency problem, one that leads managers to decrease firm risk at the expense of shareholder’s wealth. He explains that executives with more value of wealth tied to stocks of the firm are not diversified, unlike shareholders. Therefore, they will not take risky projects because they will lose their wealth tied to the stocks if anything goes wrong. However, shareholders portfolios are more diversified and they are willing to take risky projects. One of the solutions to the aforementioned agency problem is executive stock option (ESO) compensation. According to Core and Guay (2002) the stock options are contracts that give the executive, employees or other holders the right to buy a specific number of shares at a prearranged price over a fixed period. Unlike stock compensations, ESO incentivize executive to take riskier decisions (Guay, 1999). The reason behind this is the fact that there is no downside risk of the executives. The shareholder wealth increases when the stock price rises, so does the wealth of the CEO in a firm, however, when there
is a drop in stock prices and in the shareholder’s wealth, then the wealth of the CEO does not drop. The reason behind this fact is that options are only exercised whenever the stock price is exceeding the exercise price, in any other cases the option would not payout anything (Core & Guay 2002). Therefore, by providing ESOs to the CEOs, they will be incentivized to yield the required results to earn the option grant. After the grant of the options, the CEOs will own larger equity of the firm. In both cases the interests of the principal and agent are aligned (Grinner, 1999). Although it has been argued that alignment of interest through different types of efficient contracting between principal and agents increases the performance of the firm, it can also have an opposite effect.
There are several unwanted side effects of the ESO in a firm. Lowry and Murphy (2007) raise concerns about the actions of powerful CEO’s in relation to the stock option compensations. If a powerful CEO choose his/her own compensation package, then he/she is able to change the award date to match with their firm's initial public offerings (IPOs) and to set the options' exercise price equal to a low offer price to maximize their option profits (Lowry & Murphy, 2007). These powerful CEO’s, by offering themselves initial public offerings in favourable terms, would affect the shareholders wealth by not only initial public offering option grants to CEO’s but also by higher level of under-pricing of stocks.
Also Chahine & Goergen (2011) argue that CEO is the most powerful executive within a firm and this executive is even more powerful when he/ she is part of the board of directors. This particular research also found a positive relation between under-pricing and the ex-ante gain from initial public offerings (IPO’s) in firms with powerful CEO who is also founders, chairmen or both. However, a good corporate governance of the firm can limit this particular problem. Because having a highly independent board of directors may prevent CEOs from being overly powerful. Another issue raised by Lie (2005) is the vulnerability of the ESO, which leads to the opportunistic behaviour of CEOs regarding timing of the award. He argues that a compensation committee officially determines the size and the timing of the stock options grants, but the CEO can affect these decisions. This could be either by close friendship of CEO with the committee members or the fact that CEO proposes the parameters of the option grants, whereas compensation committee selects among those proposals. He also explains that the
exercise price is the same as the stock price on the day of the award of the option. Because the option value declines with exercise price, CEOs are willing to let the stock price to be as low as possible on the day of the award, in order to increase the value of their option-based compensations. This preference of the CEOs arises from their opportunistic behaviour and if the options are not awarded at the same date every year, CEOs might use their influence to fix the award date when the stock prices are low.
2.2 Managerial Risk Taking Behaviour
Managerial risk taking behaviour is described as making decisions that have high uncertainty and unpredictable outcomes, as well as potential of generating large losses (Palmer & Wiseman, 1999). This risk taking behaviour is impacted by several determinants taking place in an organization. These determinants somehow motivate CEO’s to take decisions with high or low risk. One of these determinants is the power of a CEO, which is positively related to the risk taking behaviour of CEO’s, concluded by Lewellyn et al. (2012). In their research they apply social psychology and agency theory and conclude that the power of the CEO is positively related to excessive and unmanaged risk taking. A powerful CEO could work in his/ her own interest opposed to the interest of the shareholders. Also, a powerful CEO could have an impact on the board of the directors and as a whole on the corporate governance of the firm, which could lead the firm in a riskier position.
Managerial decisions of CEO can be analysed by looking at his/ her different policies within a firm, such as investment policy, financial policy and firm focus. According to Huang et al (2013) there are two types of prominent investments, research and development (R&D) expenditure and capital expenditure, in a firm. R&D expenditure is categorized as risky investment, since it involves long payoff period with a low rate of success. If the R&D investment is successful, it usually leads to higher profits, but the related risk is also very high. One of the factors that could lead to increase in investment of R&D expenditure is the age of the CEO. As Marshall et al. (2006) explain in their paper that younger managers are more willing to take R&D investments and are less conservative, thereby more likely to take risk. On the other hand, Matta and
Beamish, (2008) support the findings of Marshall et al. (2006), by arguing that CEO’s near retirement age become oriented toward short-term investment strategies and willing to reduce R&D expenditure in their last years of job. Another reason for the increase in risky investment, R&D expenditure, could be the CEO tenure. As Barker and Muller (2002) state in their research, shorter tenured CEO follows prospector strategies, which means emphasizing and pursue strategies of innovation through higher R&D spending.
Also financial policy of the managers should be considered when recognizing the risk taking behaviour of the CEO’s in a firm. Also, Huang et al (2013) state in their paper, that the higher the leverage of the firm the riskier financial policy of managers in the firm. Higher leverage policy of the firm is caused by several determinants, such as, the tax shield benefits and low risk of bankruptcy cost. If a firm borrows, she is also obliged to pay the interest on their debt. This will increase the cost of the firm and decrease the income of the firm, and it will decrease the firms tax payments. Therefore a CEO of a firm will be incentivized to have more leverage to profit from the tax shield. (Berk and Demarzo, 2014)
Summarized, it seems that there are different opinions about the effectiveness of ESO. There are several reasons to accept that this type of compensation is beneficial for the firms, however, there are also plenty of reasons to accept the fact that ESO does not always incentivize CEO for optimal risk taking. This paper will contribute by looking at the impact of ESO on the risk taking behavior of CEO’s, specifically at CEO’s operational decisions.
2.3 Hypothesis
This paper is specifically looking at the effect of ESO on investment and financial decision of CEO’s in a firm. Guay (1999) researched the impact of equity risk, or stock return volatility, on the value of options or stocks held by CEO’s. In his research he found that not common stockholding, but stock option increase the sensitivity of the CEO’s wealth to equity risk. He shows that convexity of the manager’s
wealth-performance is positively linked to the risk taking. In other words the stock option increases convexity of the relation between manager’s wealth and stock price, therefore they also increase the risk taking behaviour of CEOs. Accordingly, Coles et al. (2006) define Delta as sensitivity of CEO’s wealth to the stock price, and it is seen as aligning the interest of CEO’s with shareholders. They mention in their research that higher Delta will incentivise CEOs to work harder to increase their share gains, however, this could also affect more risk averse behaviour of the CEOs. This risk aversion of CEO is due to the fact that CEO’s wealth is tied to a firm and he/ she would not want to take decisions, which could possibly lead to lower share returns or possibly negative share returns. In addition, unlike shareholders the portfolios of CEO’s are not diversified and this may incentivise them to avoid taking risky project.
On the other hand, Vega of ESO has a positive relation with research and development expenditure in a firm. Coles et al. (2006) define Vega as sensitivity of the CEO stock option holding to stock return volatility. They explain that sensitivity to stock return volatility will decrease the risk aversion of the CEO’s caused by the high Delta. So the compensations granted as stocks do not have effect on the risk taking behaviour of the CEO’s, since stocks add more ownership of the CEO’s in a firm, therefore they are incentivised to take less risk. Also, Haugen and Senbet (1981) conclude in their research that executive stock options (ESO) increases risk-taking incentives of a CEO because increase in stock return volatility increases the value of options. Also the literature mentioned above provides evidence of a relation between risk taking behaviour and compensation structure, and the fact that ESO sensitivity to stock price has influence on risk taking behaviour of a CEO. Therefore, this research will focus on the impact of sensitivity of ESO to stock price volatility on risk taking behaviour of the CEO’s, by looking at their investment and financial policies. Shen and Zhang (2013) explain in their paper that Delta, the sensitivity of CEO compensation portfolio value to stock price, and Vega, change in the value of a compensation portfolio for a 1% change in stock return volatility, are two main characteristics in measuring the influence of the compensation on risk taking behaviour. Vega is more noticeable when the result of the investment is positively related to the stock return volatility, as mentioned above the R&D investment and higher leverage policies are more suitable for these kind of
situations. Delta on the other hand has also impact on the investment and financial policy, however, these effects are somehow ambiguous and unclear, and therefore I use Delta only as a control variable in my model.
In addition, in order to analyse the risk taking behaviour of the CEOs, I am going to research the operational decisions, investment policy & financing policy, of the CEOs. According to Coles et al. (2010) there are two types of main investment policies, Research and Development (R&D) investment and net capital expenditure, and they define the former riskier than later. They argue that R&D investment involves long payoff with usually low rate of success. So if it is successful then it will have higher profits, if not, then it can lead to high risk for the firm. Also, prior studies explain that investors have difficulties in categorizing R&D investment by its qualities because of their intangible characteristic. (GU and Wang, 2005). So an increase in the R&D investments increases the number of different opinions between investors because of the nature of this particular investment, and these differences in opinions of investors in return has a positive effect on the stock return volatility. Also, Rajgopal and Shevlin (2002) and Coles et al. (2006) conclude in their research that ESO’s sensitivity to stock price volatility will persuade executives to invest in high-risk and high-return projects.
Based on the discussion above, I assume that CEO risk taking incentives, which is mainly derived by the ESO’s provided to them, increases a firm’s investments in R&D and decreases the capital expenditure.
H1: Executive Stock option sensitivity to stock price volatility is positively related to R&D investment.
Unlike R&D expenditure, the capital expenditure is far less risky. Therefore, capital expenditure is seen as less risky investment. According to Coles et al. (2006) one way to increase risk is to reallocate investment dollars away from capital expenditure, towards more R&D investment. Therefore;
H2: Executive stock option sensitivity to stock price volatility is negatively related to capital expenditure.
to the value of the firm. However, leverage can cause costs in a firm, such as financial distress and bankruptcy costs. Therefore, a higher leveraged firm is more exposed to shocks in the market and therefore have higher risk. Also according to Knopf et al. (2002) managers who own portfolios with higher sensitivity to stock return volatility tend to hedge less than mangers with portfolios with lower sensitivity. Also, Coles et al. (2006) conclude in their paper that sensitivity of CEO’s wealth to stock return volatility is related to higher leverage.
H3: Executive Stock Option sensitivity to stock price volatility is positively related to the increase in Book Leverage.
3. Methodology
3.1 SampleSimilar to previous researches on the executive compensation (Coles et al., 2006; Shen & Zhang, 2013) this research will use the data of listed companies in the S&P 1500. I have gathered data on the compensation of CEO’s and the firm specific data for the period 2009-2013. The data on corporate CEO’s salary, bonus and stocks and option holdings is gathered from Execucomp and the firm specific data is gathered from the CRSP (Center for Research in Security Prices)/Compustat merged. The data contains different periods related to each CEO, and therefore the data could be less than 5 years for some CEO’s. Aligned with the research of Huang et al. (2013) this research will exclude the financial firms (SIC codes between 6000 and 6999) from the data collection because of their unique characteristics. According to Lin et al. (2010) it is difficult to compare their financial data with the firms related to other industry. Another reason for excluding these financial firms from my data collection is the fact that they lack the data I need for my research because of their special nature of business activities.
My initial sample consists of 8,774 observations from Execucomp, after merging with CRSP/Compustat the sample decreases to the 8,554 observations. Besides, 1,669 observations related to financial firms are excluded from my sample. Finally, after
deleting some missing values related to the CEO compensation and other variables I have reached to the final sample of 3,582 observations.
As already mentioned the research period (2009-2013) is part of my contribution to the field of my research. There is no research about the impact of CEO compensation on CEO’s risk taking behaviour based on the latest data, especially after latest financial crisis. Therefore, this research will be able to see the changes, if there are any, in the period just after financial crisis.
3.2 Empirical Method
As already explained the research will mainly explore the influence of stock option compensation on CEO risk taking behaviours. To start, I have to find out about the Vega and Delta. Delta is known as change in CEO wealth for a change of 1% in stock returns, and Vega is the change in option value for a 0.01 change in stock-return volatility. Vega and Delta are calculated by following, Core & Guay (1999) research, in their research they used the Black, Scholes and Merton (1973) formula. This formula is widely used in research world for calculating the option value. Further details on the formula are provided in Appendix A.
Before calculating the Vega and Delta all the variables needed are gathered from several databases. The variables needed are, volatility of stock return, stock price, exercise price, expected dividend of option, risk free rate of related options and time to maturity of the options (Core & Guay, 2002; Cole et al., 2006; Huang et al., 2013). The volatility of the returns, which is defined as logarithm of the variance of daily stock returns, is gathered from the CRSP/ Compustat merged by the name of (OPTVOL). Also price of the underlying stock (PRCCF) and exercise price of the option (OPTPRCWA) are gathered from the same database. The exercise is the weighted average price of exercisable options. Another variable needed for the calculation of option value is the risk free interest rate (OPTRFR), I used the average risk free rate used in the fair value calculation of stock options, provided by the CRSP/Compustat merged. Time to maturity of the option in years is gathered from CRSP/Compustat merged as well, which is the expected life of the option used in fair value calculations provided by this particular database, (OPTLIFE). The last component needed for the formula is the expected
dividend rate over the life of option. This variable is calculated by Ln(1+dividend rate), in which dividend rate is calculated by taking the product of annual dividend per share divided by the year end stock price. Further details on these variables are provided in Appendix A.
In addition to components used in the Black Scholes formula, I have also gathered information about the CEO wealth in order to calculate the Vega, Delta and Cash compensation related to each CEO and then wensorized the Vega, Delta and Cash compensation of CEOs by 1st and 99th percentile (Core & Guay, 1999; Coles et al., 2006). By winsorizing these variables the outliers of the data are not deleted from my sample of data but take on value of specific percentile. The CEO wealth is measured by looking at their stock option and stock portfolios. Stock option portfolio of the CEO is calculated by adding the unexercised exercisable options and unexercised unexercisable options of the CEO. And stock portfolio is measured by multiplying the stock price and stockholdings of CEO (Coles et al., 2002). Cash compensation is calculated by taking sum of the salary and bonus variable provided in the Execucomp database.
After gathering all relevant information, the Vega of single option is calculated using the Black Scholes formula and then multiplied by the total option holding of the CEO. Delta is calculated by multiplying the option delta with option holding of CEO and then multiplied by change in total stock value for a 1% change in the stock price.
In this study the independent variables are Vega and Delta of the compensation wealth of the CEOs. The dependent variables related to the investment policy are the capital expenditures and research & development investment in investment policy model. The dependent variable related to the financing policy model is book leverage.
3.2.1 Investment Policy Model
Parallel with the previous research (Huang et al., 2013) about the impact of the executive equity compensation on the risk taking behaviour of CEOs in Chinese firms, this research is also implementing a model to see the effect of ESO and stock compensations on the investment policy and financing policy of the CEOs of the S&P 1500 listed firms. In this model I expect that there is an increase in the research & development of a firm by increase in the Vega of the CEO compensation and I expect the opposite effect in the
capital expenditure. In other words, whether a CEO of a firm is choosing a risky investment policy, by increasing R&D expenditures or a safer investment policy, by increasing capital expenditures. The R&D expenditures and capital expenditures are used as dependent variables. Research and Development is calculated by dividing investments in research and development by the total assets of the firm, and the missing values of research and development expenditure are set to zero (Coles et al., 2002 & Huang et al., 2013). And Capital expenditure is calculated by deducting sale of property, plant and equipment from capital expenditures and then divided by total assets.
The control variables used in investment policy model represent forces that influence Vega and Delta together with risk measures and it is based on earlier conducted studies. (Servaes, 1994; Opler et al., 1999; Huang et al, 2013; Coles et al, 2006). Control variable Size represents the firm size and it is calculated by taking a logarithm of sales. Executive age, as the name of control variable already explains, it represents the age of CEOs. Surplus Cash is the amount of cash available for the financing of new projects divided by total assets (Richardson, 2002). Cash compensation is used to proxy for the CEO’s level of risk aversion, it is the cash compensation of the CEOs, which is sum of salary and bonuses provided to the CEO (Coles et al., 2006). According to Berger et al. (1997) CEOs with longer tenures and higher cash compensation are more likely to avoid risk. However, I did not include the CEO tenure because for the most of companies in my sample there was no change for employed CEOs. Return on Asset (ROA), defined as earnings before interest, taxes, depreciation and amortization (EBITDA) divided by total assets. MTB, stands for the Market to book ratio and it is calculated by (total asset - common equity + price-fiscal year *common shares) / total asset. This particular control variable is used as proxy of investment opportunities. Net PPE is the investment in property, plant and equipment divided by total assets. Z-score is used to proxy the bankruptcy of the firm and it is calculated as followed; Z-score = 0.012 (working capital/total assets) + 0.014 (retained earnings/total assets) + 0.033 (earnings before interest and taxes) + 0.006 (market value of equity/book value of total liabilities) + 0.999 (sales/total assets) (Altman, 1968).
Ø Model of Hypothesis 1: R&Dit = β0+ β 1Vega it + β 2 Deltait + β 3
7SurplusCashit + β 8SalesGrowthit + β 9BookLeverageit +Eit
Ø Model of Hypothesis 2: CAPEX it = β0 + β 1Vega it + β 2 Deltait + β 3
ExecutiveAgeit + β 4Cashcompensationit + β 5Sizeit + β 6MTBit + β
7SurplusCashit + β 8SalesGrowthit + β 9BookLeverageit +Eit
3.2.2 Financing Policy Model
In this part of the research I am also looking at the risk taking incentives of CEOs resulting from their stock options compensation, to see whether these compensations has an impact on the financing policy of the CEOs in a firm. The book leverage is used as dependent variable in this case. As earlier mentioned increase in leverage is positively related to Vega, so a higher Vega should cause a higher leverage within a firm. Therefore, I expect an increase in book leverage by increase in Vega. Unlike Vega, Delta has a negative effect on book leverage, so I expect by increase in Delta there is a decrease in book leverage of firm (Coles et al., 2006).
Ø Model of Hypothesis 3 BookLeverageit = β 0 + β 1Vegait + β 2 Deltait + β 3
ExecutiveAgeit +β
4Cashit + β 5Sizeit + β 6MTBit + β 7ROAit + β 8NetPPEit + β
9R&Dit + β 10 Z-scoreit +Eit
4. Descriptive Analysis
4.1 Descriptive statisticsIn this part of the research the details about the data analysed and the results related are discussed. In the table below all the variables statistics are summarized. CEO characteristics related data, Vega, Delta and Cash compensations related data is very similar to the research of Coles et al. (2006). The mean of Vega 68.56 thousand dollar, this number does not differ much from the findings of the Coles et al. (2006) paper, in which the mean of Vega is 80 thousand dollar. However, This value cannot be compared
with the paper of the Huang et al. (2013), because there values are indicated by Chinese currency, since the research is based only on China. Mean of the Delta is much higher than mean of Vega, this is of course because Delta is the change in dollar value of CEO’s stock and stock option portfolio for one percentage point change in stock price, But Vega is the change in dollar value of CEO’s stock options portfolio for one percentage point change in volatility of stock returns.
On the other hand the data statistics differ from the research of Huang et al. (2013), since the research field related to this particular paper is China and the characteristics of firms and CEO compensation related to this particular country are different than the ones listed in S&P 1500. Difference between mean of the Delta and Cash compensation is smaller than the previous researches (Coles et al., 2006), the reason for this matter could be increase in popularity in equity compensation and decrease in cash compensation payments. Another reason is off course, the impact of the crisis on the strategy of the firms regarding the CEO payments, since the firms are more willing to give the equity compensation and less cash compensation if they are short on cash. Considering the investment policy of the firms my data produce different results than Coles et al. (2006) results. The difference between the mean of R&D and the mean of Capital expenditure in the table below is 0.005, however, the difference in mean of these two variables in Coles et al. (2006) is 0.030. The mean capital expenditures of the firms are quiet lower in 2009-2013 period compared to the period (1992-2002) researched by
Mean Standard deviation Min 10th percentile 50th percentile 75th percentile Max Observations CEO Compensation Vega ($000s) 68.56 147.10 2.3 15.35 49.50 165 3934.7 3582 Delta ($000s) 604.91 917.69 41.25 44.18 256.70 656.10 4500 3582 Cash Compensation ($000s) 922.49 627.08 41.25 420.70 800 1055 4500 3582 Dependent Variables R&D 0.035 0.067 0 0 0.0053 0.047 0.8923 3582 CAPEX 0.040 0.044 0.00010 0.008 0.027 0.049 0.209 3582 Book Leverage 0.218 0.224 0 0 0.192 0.319 0.8593 3582 Control variables Firm size 7.525 1.679 3.64 5.458 7.483 8.647 11.051 3581 Surplus cash 0.092 0.100 -‐0.148 -‐0.002 0.081 0.139 0.3625 3544 Sales growth 0.128 1.061 -‐0.33 -‐0.09 0.070 0.161 1.1290 3582 Return on Assets 0.138 0.109 -‐0.161 0.054 0.133 0.182 0.422 3582
Z-‐score 3.553 35.469 0.213 0.415 0.955 1.554 49.50 2955
Market to book 1.958 1.208 0.77 1.033 1.603 2.217 7.837 3582
Executive age 55 7 31 46 55 59 75 3580
Net PPE 0.247 0.219 0.0082 0.036 0.172 0.349 0.870 3579
Table 1: descriptive statistics of dependent variables (Research and Development, Capital expenditure, Book leverage) and independent variables (Vega, Delta,
Cash compensation) and control variables (Firms size, Surplus cash, Sales growth, Return on Assets, Z-score, Market to book, Executive age, Net PPE.) Vega is the dollar change in the CEO’s wealth for a 0.01 change in the annualized volatility returns and Delta is the change in CEO wealth for a change of 1% in the stock price. Cash compensation is the sum of salary and bonuses provided to each CEO. Similar to other studies, Vega, Delta, Market to book and Cash compensation variables are winsorized at the 1st and 99th percentile.
Table 2 shows that the correlation between Vega and R&D expenditure is unexpectedly negative (-0.03). According to previous researches (Coles et al, 2006 & Huang et al, 2013) R&D expenditures of the firms are positively related to the Vega. On the other hand, consistent with previous researches Vega has a negative correlation (-0.04) with capital expenditure of the firms. Also Delta has a positive correlation with R&D expenditure and negatively correlated to the capital expenditure of the firms, which is not similar to previous research results. The negative correlation of the capital expenditure with Vega and Delta could be affected by a decrease in capital expenditure as a whole, as shown in table 1, the capital expenditure mean is lower compared to the previous research (Coles et al., 2006), this particular research was also focused on the firms listed in S&P 1500 but in the period of 1992-2002. Besides the investment policy indicator, if I look at the financial policy dependent variable, book leverage, I can see in my data that this variable is positively correlated (0.061) with Vega and negatively correlated (-0.07) with capital expenditure. This outcome is in line with the previous research quotes, since increase in Vega increases the book leverage in a firm and Delta decreases the book leverage.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 1.Vega 1 2.Delta 0.21 1 3.R&D -0.03 0.02 1 4.Sales growth -0.02 -0.01 0.04 1 5.Executive’s age 0.07 0.14 -0.04 -0.01 1 6. Cash compensation. 0.16 0.31 -0.11 -0.003 0.18 1 7. Firm size 0.27 0.29 -0.32 -0.10 0.05 0.35 1 8. Surplus cash 0.029 0.20 0.43 0.005 -0.04 0.004 0.011 1 9. Book leverage 0.061 -0.07 -0.04 0.007 0.11 0.075 0.103 -0.22 1 10. Market to book 0.020 0.32 0.32 0.045 -0.05 -0.019 -0.106 0.43 0.02 1 11. capital expenditure -0.04 -0.01 -0.15 0.033 -0.02 -0.031 0.019 -0.01 0.05 0.004 1 12. Z-score -0.02 -0.01 0.05 0.011 -0.01 -0.006 -0.107 0.09 -0.13 0.082 -0.03 1 13. ROA 0.05 0.19 -0.26 -0.039 -0.03 0.074 0.258 0.44 -0.16 0.25 0.125 0.04 1 14. Net PPE -0.01 -0.08 -0.28 -0.019 0.04 0.009 0.119 -0.20 0.21 -0.174 0.63 -0.07 0.038 1
Table 2: In the table above the correlation of dependent variables (Research and Development, Capital expenditure, Book leverage) and independent variables (Vega, Delta)
5 Multivariate Analyses
5.1 Investment policy5.1.1 Research and Development Expenditure
In the table below the results of the OLS regression of the dependent variable, R&D expenditure, with independent variables and control variables is presented. In these results only the control variables are taken which could also have impact on R&D expenditure of the firm besides the impact of Vega and Delta. As already explained the control variables represent forces that drive Vega and Delta together with investment policy. In the table below Vega is statistically significant at 1% and has a positive coefficient (0.000176) with R&D expenditure. From this result I can conclude that a higher Vega result in a higher R&D expenditures. This result is consistent with prior literature, Coles et al. (2006) finds a positive relation between the R&D expenditure and Vega. Considering the Delta part of analysis, my result indicates that there is a negative coefficient (-0.0000141) with R&D expenditure. However, this value is not statistically significant (p-value 0.233) and therefore I cannot conclude that Delta has a negative impact on R&D expenditure in a firm. So this result is inconsistent with previous researches, such as, Huang et al (2013) finds that Delta is negatively related to the R&D expenditure and their results are statistically significant. Marshall et al. (2006) mentioned in their research that younger CEOs are more willing to take R&D investment than older. However, my findings show that executive age has negative coefficient (-0.000162) but it is not statistically significant. I also find that firm size has negative coefficient (-0.0146) and statistically significant (0.00). This result is inconsistent with the findings of Huang et al (2013), in which they find a positive coefficient. However, my findings are similar to Coles et al. (2006) paper, in which they find a negative coefficient of firm size.
The results mentioned above indicate that CEO’s with higher Vega are incentivised to invest in risky investments, R&D expenditures. However, my result does not indicate that CEO’s with higher Delta are more reluctant to invest in risky investment, R&D expenditures. As expected, the ESO sensitivity to stock return is positively related to the R&D expenditure, therefore I can reject my null hypothesis and accept hypothesis number (1).
OLS regression with R&D expenditure as dependent variable
Variables Coefficient t-statistics p-value N 3487
Vega 0.000176 2.69 0.007 F-value 188.37
Delta -0.0000141 1.19 0.233 (Pr.>F) 0.0000
Sales growth 0.00051 0.59 0.559 Adjusted
R-squared 0.326 Firm size -0.0146 -22.08 0.000 Surplus cash 0.3065 27.30 0.000 Book leverage 0.0115 2.25 0.025 Market to book -0.00235 -2.47 0.013 Executive age -0.000162 -1.22 0.224 Cash compensation 0.000016 0.94 0.347
Table 3: Regression on Research and Development expenditure as dependent variable and Independent
variable is Vega and Delta and Control variables. R&D expenditure is scaled by assets. Vega is the change in dollar wealth of CEO for a 0.01 change in volatility of stock return, and Delta represents the change of CEO’s wealth for delta of 1% in stock returns.
5.1.2 Capital Expenditure
In this part of the research the results of the OLS regression with capital expenditure as dependent variable and Vega and Delta as independent variable and related control variables are presented. In the table below the result shows that Vega is statistically significant at 5% and it has a negative coefficient (-0.000116) with capital expenditures. This result indicates that Vega is negatively influencing the capital expenditure. Similar to the paper of Coles et al. (2006), in which they find that higher Vega is negatively correlated to capital expenditure. Also, Huang et al. (2013), a research based on Chinese firm, finds that Vega is has negative impact on the capital expenditure of the firm. This means the higher Vega incentivizes CEOs to reduce capital expenditures and it is statistically significant. On the other hand, Delta has a positive coefficient (0.00000008), however, this result is statistically insignificant (0.937). Therefore one cannot conclude from this result that higher Delta incentivizes CEOs to invest more or less in capital expenditure. Also firm size and sales growth
has positive coefficient, the values are (0.00177) and (0.001) respectively. Both of these results are consistent with the paper of Coles et al. (2006), however, it is not consistent to the paper of Huang et al. (2013), in which the sales growth has a negative coefficient.
Overall the result of regression is partly consistent with prior research, since higher Vega has a negative coefficient with capital expenditure. As expected the ESO sensitivity to stock price volatility, Vega, is negatively related to the capital expenditure in a firm, therefore I reject my null hypothesis and accept hypothesis number (2). The Delta part of the regression does not explain much, since it is statistically not supported.
OLS regression with capital expenditure as dependent variable
Variables Coefficient t-statistics p-value N 2480
Vega -0.000116 -2.00 0.046 F-value 2.30
Delta 0.00000008 0.08 0.937 (Pr.>F) 0.0144
Market to book 0.0010 1.26 0.208 Adjusted
R-squared 0.0083 Sales growth 0.00115 1.67 0.095 Surplus cash 0.0049 0.51 0.611 Book leverage 0.00641 1.34 0.181 Firm size 0.00177 2.86 0.004 Executive age -0.0000152 -0.13 0.899 Cash compensation -0.000038 -2.37 0.018
Table 4: Regression on capital expenditure as a dependent variable and Independent variable, Vega
and Delta, and Control variables. Capital expenditure is calculated by taking the difference between capital expenditure and sale of Net PPE and divided by assets. Vega is the change in dollar wealth of CEO for a 0.01 change in volatility of stock return, and Delta represents the change of CEO’s wealth for delta of 1% in stock returns.
5.2 Financial Policy
dependent variable, book leverage, and independent variables, Vega and Delta, and control variables. Vega is statistically significant at 5% and has a positive coefficient (0.0000488) with book leverage, however, delta has a negative coefficient with book leverage and the results are also statistically significant (0.000). These results indicate that a higher Vega of CEOs leads them to take higher leverage, and the higher Delta of CEOs incentivize CEOs to take less leverage. These results are consistent with prior literature and also it is similar to what I have hypothesized. As Cohen et al. (2000) finds higher Vega leads to the higher book leverage. These results are also consistent with the findings of Coles et al. (2006), in which they find a higher Vega has positive correlation with leverage. Market to book coefficient is positive (0.0572) and statistically significant, unlike prior researches, as Yartey (2009) finds in his research that market to book is negative and significantly related to the book leverage. As expected earlier in the paper that ESO sensitivity to stock return, Vega, has positive relation to the book leverage in a firm, therefore I reject my null hypothesis and accept the hypothesis number (2). I can conclude that higher Vega incentivizes CEO to implement risky financial policy.
OLS regression with Book leverage as dependent variable
Variables Coefficient t-statistics p-value N 2924
Vega 0.0000488 1.95 0.051 F-value 32.02
Delta -0.000347 -7.08 0.000 (Pr.>F) 0.0000
Firm size 0.0098 3.33 0.001 Adjusted
R-squared 0.095 Market to book 0.0572 0.0045 0.000 Executive age 0.0003 0.06 0.955 Cash compensation 0.000032 4.56 0.000 ROA -0.2705 -6.02 0.000 Net PPE 0.17022 9.54 0.000 R&D -0.4295 -5.25 0.000 Z-score -0.00048 -4.48 0.000
Table 5: Regression on book leverage as a dependent variable and Independent variable, Vega and
Delta, and Control variables. Book leverage is sum of long-term debt and debt in accrued expenses scaled by assets. Vega is the change in dollar wealth of CEO for a 0.01 change in volatility of stock return, and Delta represents the change of CEO’s wealth for delta of 1% in stock returns.
5.3 Robustness Analyses
5.3.1 Regression of deleted missing R&D values
Similar to the previous research paper I have set all the missing values of R&D expenditure equal to zero in my data analyses, but in robustness check I delete all the missing R&D expenditures from the data I have gathered from Execucomp. By doing so, I would like to see whether my choice has major influence on the outcome of my data regression. All the missing values are deleted from the data, which results in final observations total of 2286. This data is regressed and available in the table below. On the left side of the table the R&D robust represents the result from the regression with all the missing values deleted and on the right the results from the result from the regression in which missing values of R&D expenditure are set in zero. In the robust R&D coefficient I see a change compared to previously regressed R&D. The coefficient has increased to (0.00028) but it is still positive. Also the p value in the new regression is slightly changed compared to the previous (0.003), but it is still statistically significant. On the other hand, the Delta coefficient is changed to the positive number (0.0000008), however, the p value is statistically insignificant, similar to the previous regression (0.270), so I cannot conclude anything from this result. The observations in new regression, which excludes missing values of R&D from data, are decreased to 2286 compared to previous regression observations of 3487; this is of course result of the deleted missing values of R&D expenditures. Also, the Adjusted R-squared is increased to 36.0% from 32.6%.
Table 6: OLS regression of robust R&D expenditure and R&D expenditure
Summary of two OLS regression with deleted missing R&D and undeleted missing R&D as dependent variable
Deleted missing R&D R&D Deleted missing R&D R&D Deleted missing R&D R&D Deleted missing R&D R&D
Variables Coefficient t-statistics p-value N 2286 3487
Vega 0.00028 0.000176 2.94 2.69 0.003 0.007 F-value 143.93 188.37
Delta 0.0000008 -0.0000141 1.10 1.19 0.270 0.233 (Pr.>F) 0.0000 0.0000
Sales growth 0.0055 0.00051 2.14 0.59 0.032 0.559 Adjusted
R-squared 0.360 0.326 Firm size -0.0184 -0.0146 -21.29 -22.08 0.000 0.000 Surplus cash 0.35 0.3065 23.21 27.30 0.000 0.000 Book leverage 0.035 0.0115 4.69 2.25 0.000 0.025 Market to book -0.0038 -0.00235 -3.26 -2.47 0.137 0.013 Executive age -0.00032 -0.000162 -1.49 -1.22 0.296 0.224 Cash compensation -0.000001 0.000016 1.05 0.94 0.000 0.347
5.3.2 Regression of top 25th percentile of Vega
In order to check whether the results are driven by the subset of CEOs with very high values of Vega, I run the regression mentioned above again, but this time with the top 25th percentile of Vega and compare the results with the regressions mentioned above. First, all the Vega values are divided in two groups, group 1 consist 25 % of observations with highest values and group 2 consist of remaining observations of Vega. Second, Group 2 is dropped out of the data and only the top 25% values of Vega is regressed with R&D expenditure, capital expenditure and book leverage, together with control variables, Delta, sales growth, surplus cash, firm size, executive age and cash compensation, market to book and book leverage.
In the regression of R&D expenditure as a dependent variable, the coefficient of Vega is stronger than the coefficient of the regression mentioned above (0.00067) and also positive. However, the results are 10% statistically significant compared the previous regression it is higher. The results indicate that top 25% CEOs Vega is strongly related with R&D expenditure. Also, in my new regression I find that book leverage has negative coefficient with R& D, this result is not similar to the previous regression, in which the coefficient is positive.
Another regression, including capital expenditure as dependent variable, the coefficient of Vega is negative (-0.00026), however, the results are statistically insignificant (0.543). The result of this regression is different compared to the findings of the previous regression. Also Delta in the regression has negative coefficient unlike the previous regression and it is statistically significant (0.017). Moreover the control variables, sales growth and book leverage have negative coefficient with R&D expenditure, however, these results are statistically insignificant.
The last regression results, in which book leverage is taken as dependent variable, show that Vega has higher coefficient with book leverage compared to the previous regression, value of coefficient of new regression is (0.000073) and from previous regression is (0.000048). This result is also statistically significant (0.000). Analysing the delta part of the findings, I find that coefficient is negative and statistically significant similar to the previous regression. Looking at the control variables, I find that most of the coefficients are higher than the previous regression, such as Z-score has a negative coefficient (-0.0042) compared to the coefficient of
previous regression (-0.000048) is much higher. The main difference between the new regression with previous regression of book leverage could be the fact that the significance level of control variable. In the previous regression all the control variable, except executive age, are statistically significant but in the new regression there more control variable with statistically insignificant results.
Overall, If I compare the regressions of the 25% top Vega of the CEO’s with the previous regression of all the data, I can conclude that the coefficient of the Vega is much higher in the new regressions, also the values of control variables are stronger. Another change that is obvious in the new regressions is the increase in the value of the Adjust R squared, these values are much higher than previous regressions. The increase of Adjusted R squared is caused by the decrease in number of observations.
6. Conclusion
This thesis provides evidence on causal relationship between the executive stock option, ESO, sensitivity to stock price volatility and managerial decision-making. In this research I specifically focus on the investment policy and financial policy of the CEOs. Therefore, it is hypothesized that sensitivity of ESO to stock price volatility, Vega, is positively related to risky investment policy and financial policy. Besides Vega, also the role of the sensitivity of CEOs wealth to change in stock price is considered, or Delta. This particular variable is used as the control variable.
The first part of my research tests the impact of the ESO sensitivity to stock price volatility, Vega, on the research and development expenditure and capital expenditure and, as expected, the results shows a positive relation between these between R&D expenditure and Vega. On the other hand, Vega is negatively related to the capital expenditure in a firm. This result indicates that Vega of the CEOs has positive impact on their risk taking behaviour by incentivizing them to invest more in risky projects, R&D expenditures, as opposed to less risky projects, capital expenditure.
The second part of this thesis tests for the relation between Vega and the financial policy of CEOs, specifically looking into changes Vega of CEOs causes in book leverage. In my results I found a positive relation between Vega and book leverage, as hypothesized. This result indicates that sensitivity of ESO to stock return
volatility incentivizes CEOs to increase leverage, or more risky financial policy. Findings of my research add to prior literature, because the data analysed is from period of 2009-2013, this period is not covered by prior literature. During 2008 a huge collapse in the stock market took place, which had an impact on the economy and on the firms policies. By researching this particular period, I saw the changes in the impact of ESO sensitivity to stock return volatility on CEO decision making, if there are any. Comparing my findings with the results of the paper of Coles et al. (2006), I could not find any significant difference in the results. Therefore I conclude that there are no obvious changes in the impact of Vega on CEO policies caused by the collapse in the stock market of 2008.
An important issue to be mentioned regarding the limitations of this thesis is the problem of endogeneity among variables, which results in biased regression coefficients. This issue to the level addressed in my research analysis by adding several control variables, however, this might be unfortunately not enough. To further reduce the likelihood of biased results the data should be approached by several equation models, instead of only using the OLS regression (Coles et al, 2006).
Appendix A
In order to calculate the stock option’s value or sensitivity to stock price or stock return volatility, I use the formula, which is also implemented in Core & Guay (2002) paper, Black-Scholes (1973) model, also modified by Merton (1973) to account for dividend pay-outs. The measurements are based on the valuation of European call option.
Option value= [Se-dTN (Z)-Xe-rTN(Z- σ T(1/2))], Where,
Z = [ln(S/X)+T(r-d+σ2/2)] / σT(1/2),
(N) is cumulative probability function for the normal distribution , (S) is underlying stock price in on 31th of December for the period of 2009-2013, (X) is the exercise price of the option measured as the weighted average price of exercisable option, (σ) is the stock return volatility, this value is taken from the CRSP/ Compustat merged
database by the name of OPTVOL, (r) is risk free interest rate. This value is also taken from CRSP/Compustat merged database. (T) is time to maturity of the option in years , I use the expected life of the option used in fair value calculation, this value is available in CRSP/ Compustat merged database. (d) is the expected dividend rate over the life of the option, I have calculated this value with (ln(1+dividend rate)), dividend rate is calculated by dividing annual dividend per share by the year end stock price. The sensitivity with respect to a 0.01 change in stock-return volatility, Vega, is defined as (Core & Guay, 2002)
Ø [∂(option value) / ∂(stock volatility)] * 0.01 = e-dT Nʹ′(Z) S T(1/2) * (0.01), N’ = normal density function, this function is different from (N) used in the calculation of Delta.
The sensitivity with respect to a 1% change in stock price is defined, Delta, as (Core & Guay, 2002):
Ø [∂(option value) / ∂(price)] * (price/100) = e-dT N (Z) * (price/100)
Appendix B
In appendix B all the variables calculations with their description are listed below. Dependent variables
R&D = Research and Development expenditure scaled by assets
Capital expenditure = ((capital expenditure – Sale of property, plant & equipment)/Assets)
Book Leverage = (Long-Term Debt+ Debt in Accrued Expenses & Liabilities)/ Assets Independent variable
one percentage point change in the annualized standard deviation of stock returns.
Control variables
Delta = the change in the dollar value of CEO’ s stock and the stock options portfolio for one percentage point change in stock price.
Cash compensation = sum of the Bonus and Salary of the CEO Executive Age = age of the CEO
Market-to-book = (Total Assets - common equity+ price close at fiscal year*common shares outstanding)/ Total Assets
Return on Asset = operating income before depreciation/ Total Assets
Net property, plant and equipment = property, plant and equipment/ Total Assets Surplus cash = (Operating activities, NCF – Depreciation and Amortization + Research and development expenses)/ Total Assets
Firm size = Logarithm of sales Sales growth = Log (Salest /Salest-1)
Z-score = 3.3*operating income depreciation / Total Assets+ 1.2*(Current Asset – Current Liabilities)/ Total Assets + Sales / Total Assets + 0.6* price close at fiscal year *Common Shares Outstanding /(Long term debt+ Debt in Current Liabilities) + 1.4*Retained Earnings/ Total Assets, developed by Altman (1968)
Appendix C
1. Table of observations number and percentages categorized by type industry
Industry Observations Percentage Agriculture, Forestry and Fishing 8 0.2
Mining 170 4.7
Construction 76 2.2
Transportation, Communications, Electric, Gas and Sanitary service
288 8.0 Wholesale Trade 120 3.4 Retail Trade 337 9.4 Services 661 18.5 Public Administration 0 0 Total 3582 100
* Finance, Insurance and Real Estate, (6000- 6799) sic codes were already deleted from the data.
2. Observations divided by Years
7. Reference
Altman, E. I. (1968) Financial ratios, discriminant analysis and the prediction of corporate bankruptcy. The journal of finance, 23(4), pp. 589-609.
Balsam, S., & Miharjo, S. (2007) The effect of equity compensation on voluntary executive turnover. Journal of Accounting and Economics, 43(1), pp. 95-119. Barker III, V. L., & Mueller, G. C. (2002) CEO characteristics and firm R&D spending. Management Science, 48(6), pp.782-801.
Berger, P. G., E. Ofek, & Yermack, D. L. (1997) Managerial entrenchment and capital structure decisions, Journal of Finance, 52, pp.1411-1438.
Berk, J.and Demarzo, P. (2014) Corporate finance, 3rd ed. (Harlow: Pearson).
Black, F., & Scholes, M. (1973) The pricing of options and corporate liabilities. The
journal of political economy, pp.637-654.
Chahine, S., & Goergen, M. (2011) The two sides of CEO option grants at the IPO.
Journal of Corporate Finance, 17(4), pp.1116-1131.
Coles, J. L., Daniel, N. D., & Naveen, L. (2006) Managerial incentives and risk-taking. Journal of Financial Economics, 79(2), pp.431-468.
4% 25% 27% 23% 21%
