Managerial Incentive Pay and Payout Policy

TILBURG UNIVERSITY Managerial Incentive Pay and Payout Policy Is payout policy used to enrich management? Bram van Haren BSc. (982788) Master Thesis in Finance Dr. Fabio Feriozzi 12-11-2013

Abstract This thesis researches the relationship between the payout policy of the firm and the incentive compensation of executive management. The sample consists out of 286 firms from the STOXX Europe 600 index in a period of 2003-2007. I find that executive stock option ownership is positively related to dividend payout and share repurchases. This shows that there are gains to be made by management when increasing dividends and share repurchases in the presence of the stock option ownership. These findings are contradictory with earlier studies in this field. Several explanations can account for these results: compensation packages of European executives are linked to dividends (dividend protection) and taxation is on average lower in Europe lowering the barrier to pay dividends. Also legal origin and consequently corporate governance codes might be of influence. 1 P a g e

Table of Contents Abstract... 1 Introduction... 3 Theoretical Framework... 5 Payout Policy... 5 Agency Costs... 6 Free Cash Flow Hypothesis... 6 Managerial Compensation... 8 The Relationship between Payout Policy and Agency Costs... 9 Methodology... 11 Sample Selection... 11 Variables... 11 Managerial incentive variables... 11 Payout Variables... 13 Control Variables... 15 Empirical Results... 17 Taxes and Law... 19 Robustness... 23 Endogeneity... 23 Conclusion & Recommendation... 25 References... 26 Appendices... 29 2 P a g e

Introduction In modern history, firms have had a common goal, to make a profit on their activities. In other words, to make more money from the sales of their goods/services than it costs them to make. The profits are used for several purposes: reinvestment in the firm for innovation, paying for the perks of managers or providing a signal to investors by paying out (part of) the profit to investors. Dividends are established by the board of directors and then need to be approved by the annual shareholders meeting. The firm can choose in what form dividends are paid and how much dividends are paid. It is not obligatory for firms to pay dividends. However, there is research that points out changing dividends might signal the market about the health of the firms and future prospects. Roughly, profits can be considered as being free cash flows that are vital to the firm, this creates problems. Agency problems have been around since managers have assumed the task to look after the assets of the firm owned by investors. Managers are expected to act in the best interest of the investors in order to maximize their wealth. With a large pile of cash that is not needed or used by the company incentives can be created to use this money for their own purposes. (Jensen, 1986) Paying out (part of) the profit to investors might decrease this tendency. However, there are other ways to incentivize managers to act in the best interests of the investors, for example by giving managers stocks. Managerial ownership in the firm can help to align the goals of the managers with the goals of the owners of the firm. Many studies on agency costs have been conducted on the samples of American firms. Since America has the largest economy in the world it is an ideal sample of the population of firms on earth. However, the European Union is following close on America and Japan as the largest economy and gives a unique venue for conducting research on numerous financial questions. Von Eije and Megginson (2008) show that Europe is a unique sample to study financial problems because of several reasons, the most striking reason is the fact that Europe is a patchwork of different sovereign, but highly developed nations with each their own legal system, taxation system, corporate governance systems and culture. With these different yet closely linked systems it creates an interesting sample of firms to study different problems. Since the 1980s there has been a drastic increase in incentive based payment packages for management. (Hall and Liebman, 1997) Not only in the US but also in Europe the compensation was increased by stock options, share programs and bonuses. Since managers are in control of many decisions that have to be made within a firm, these incentive based payments might lead managers to decide in favor of shareholder wealth that will increase the value of the firm. Conversely, these incentives might make 3 P a g e

managers more aware of their stakes in the company and will make decisions that will ultimately lead to an increase of their own wealth due to managerial ownership. One of the decisions management is faced with is the investment choice. The firm can choose to payout funds to shareholders or to retain cash and use this to invest in profitable investment opportunities. The payout decision might not only be influenced by investment opportunities and agency costs but also by the managerial ownership of shares and stock options. In this thesis I want to combine the elements, payout policy and managerial incentives to research the relationship between both elements in the European market. My sample consists of the STOXX Europe 600 firms excluding financial firms and excluding firms that have no data on executive stock options or managerial ownership in the period 2003-2007. The final sample consists of 281 firms with each five years of data. I find that executive compensation in the form of stock options is positively related to dividend payout, repurchase payout and total payout. Managerial ownership in the form of shares has no or a weak relationship with the payout policy of the firm. Several explanations can be found for these results: Firstly, as Liljeblom and Pasternack (2006) show it might be that dividends play a larger role in the compensation packages of executive managers than was previously acknowledged. This might show the increase in dividends that stock options are causing. Secondly, repurchase payout is going up. This can be a result of the share dilution that stock options cause. In order to counter this, management repurchases shares in order to counter the dilution effects. Furthermore, profitability and free cash flow seem not to be important in the decision to payout dividends, most probably due the stickiness of dividends meaning that firms are not likely to alter their dividend policy. Profitability however is important for share repurchases showing that share repurchases are seen more as an extra method to payout cash to shareholders. (Dittmar, 2000) Lastly taxation and legal origin seem to have a negative effect on the payout policy of the firm. This is in line with the predictions that double taxation decreases dividend payout tendencies in firms. Brown et al. (2004) show the other side of the taxation problem. These findings are not in line with the findings of earlier studies of the relationship between executive pay and payout policy (Fenn and Liang, 2001; De Cesari and Ozkan, 2013). The rest of the thesis is organized as follows, the next section will give an elaborate overview of the relevant literature of the main topics in the thesis. Thirdly, the data selection process and the data are described. Section 4 shows the empirical results concerning the regression outcomes. Lastly, section 5 will give a conclusion and recommendations for further research. 4 P a g e

Theoretical Framework Payout Policy In 1961, Modigliani and Miller (MM) provided a theory that shows the dividend irrelevance proposition. In their article they describe, with three rather strict assumptions about markets and investors, that the payout policy of a firm is not important for the valuation of the stocks and firm value. According to MM, the only way to account for the fact that dividend paying stocks are trading at a premium is the irrationality of investors. In fact the mechanism shows clearly that dividends do not add value. An investor that is holding a share of $10 that pays 2$ dividend at the end of the year will be worth only $8 after the dividend has been paid. The value of the share is the same for an investor that holds an $8 share in a company that does not pay dividend. So the value of the firm is not contingent on the amount of dividend it pays. However, throughout history many firms have paid large amounts of dividend as an annual or semiannual activity. Many questions have been raised on why firms pay a dividend to shareholders if it does not create value. While firms are in fact established to create value and to pursue all activities that increase the value of the firm and therefore shareholder wealth. This puzzle has been addressed by Black (1976). The dividend puzzle is extensively covered in his paper. The dividend puzzle can be described as a puzzle with many pieces that all form small cohesive blocks but do not fit together in the big picture. Corporations might want to payout dividends to pay investors for the risk they take by investing their money into the company. Possibly firms want to reward investors by paying dividend and want to attract new investors with the prospect of receiving dividends. However, on the other side, there is the fact that retaining the money otherwise spend on dividends to pursue new profitable investment opportunities is a legitimate reason to not payout cash to investors. And might increase the value of the firm and consequently create a larger return for the same investors. In addition to this, double taxation might withhold firms to payout dividends and might rather choose to payout cash through share repurchases which tends to be less heavily taxed. Another reason for payout in the form of dividends might be the investor that demands it. When making the purchase decision for a specific stock does the investor select stocks for inclusion in its portfolio on the basis of dividend payments? It might be true that despite the theoretical explanations investors determine dividends to be a sign of strength. The signs that dividends might send to investors or analysts has been an argument since the 1960s. Dividends might signal the outside world expectations about firm performance in the future. A dividend increase might show that managers are optimistic about future operations. By conducting interviews 5 P a g e

with managers Lintner (1956) found that dividend payments are dependent on believes of management that there will be a sustainable level of earnings in the future. Empirical evidence has determined that this is indeed the case. (Healyand and Palepu, 1988) Furthermore, the punishment of the market on dividend decreases has led to dividend conservatism (Brav et al., 2005) making managers reluctant to increase dividend. More recent research has shown that the signaling theory of Lintner has lost some its strength. It cannot be the sole explanation for paying dividends. Empirical evidence does show that dividends tend to be sticky over time. Since the 1980s firms have switched increasingly to share repurchases (Ikenberrya et al., 1995; Fama and French, 2001). While dividends have long been the prevailing way to distribute cash among shareholders the 1980s and even more so the 1990s have shown that share repurchases are becoming a favorable form to distribute cash. Finance theory suggests that the decision to distribute cash through share repurchases is made by considering a number of factors such as: distribution, investment, capital structure, corporate control and compensation policies. (Dittmar, 2000) By analyzing all these possible factors that might influence the decision to repurchase shares Dittmar (2000) finds that share repurchases are not a substitute for dividend. They are rather a means of distributing extra cash and might arise in circumstances such as: adjusting their leverage ratio, to fend off takeovers and to counter the dilution effect of stock options. Furthermore, early empirical evidence suggests that share repurchases are also used to signal the market. (Vermaelen, 1981; Ofer and Thakor, 1987) Firms tend to repurchase shares when the value of the shares is perceived to be undervalued. By buying shares at a premium managers signal an undervaluation to the market with the intention of influencing the market. Additionally, research has been done in the actual information that is conveyed through share repurchase programs. Grullon et al. (2004) find that firms that have share repurchase programs reduce their systematic risk and cost of capital relative to firms that do not have a share repurchase program. This suggests that share repurchases signal positive news to the market. Also, share repurchases seem to be good news for firms that have higher agency costs, or are more likely to overinvest. This is in line with the free cash flow hypothesis that shows that a reduction in free cash flow might help to reduce agency costs. Agency Costs Free Cash Flow Hypothesis Ever since owners of a firm left daily management of the firm to agents that do not own a fraction of the firm, agency costs exist. The costs arise when managers do not act in the best interest of the 6 P a g e

shareholders and use company funds to enrich themselves, either personally or professionally. The problem is sometimes hard to recognize and is often only noticeable when it is too late, and shareholders have lost millions of dollars. Several empirical events like ex-ceo of Tyco International Kozlowski, and Ex-CEO Bernhard Madoff of Madoff Investment Securities show the severity and the relative ease with which managers can use the money of investors for purposes that are not in line with the goal of the company and destroy value. The papers by Jensen (1986) and Stulz (1990) present a thorough explanation of the agency problems. The free cash flow hypothesis is one of the leading agency problems. The problem of free cash flow starts with an information asymmetry between managers and the shareholder of the firm. Since cash flow is not always sufficient to undertake all profitable investment projects management is not always able to invest. It is in the interest of management understate cash flow such that they can benefit from large cash flows in good periods and communicate the truth to shareholders in bad periods. This information asymmetry gives shareholders the unchecked ability to invest in projects that only serve the interest of the shareholders but also their own interest. In order to reduce free cash flow to a level that is not attractive for management to use for their perquisites. Debt is one of the measures that can reduce free cash flow. Interest payments are an extra cost to the firm and therefore reduce the amount that can be reinvested in positive net present value (NPV) projects. The optimal debt level is sufficiently high such that management is forced to act in the company s best interest however not so high that the company is transferring control to debt holders. Debt is a binding contract that should be the element that forces management to align their goals and behavior with the shareholders. Furthermore, the payout of free cash flow to shareholders is an alternative way to reduce free cash flow and eliminate (partly) the possibility for management to overinvest in perquisites that do not add to shareholder value. However, also with the payout policy there is a conflict of interest. Managers will be reluctant to payout large parts of the free cash flow since they will not benefit from it. Conversely, shareholders would like to gain as much as possible and in addition the shareholders want to minimize the funds available to the managers to invest in perquisites of negative NPV projects. The problem of corporate governance is similar to the agency problem however focuses also on the other stakeholders of the firm. It addresses the broad question: how can be ensured that the suppliers of capital get their capital back? (Shleifer et al., 1997) This question involves many aspects of the firms. Not only finance is included and not only shareholders are represented. All stakeholders in the firm are represented in the corporate governance theory. Furthermore, all aspects of the firm are represented in 7 P a g e

the theory. While above only the free cash flow problem has been addressed, the corporate governance problem also entails legal systems that influence the firm, the composition and presence of a two tier board system and the reward structures for management. All those aspects have an influence on the behavior of management and are meant to force them to behave in a way that merits all stakeholders of the firm. Managerial Compensation Over the years the payment structure of managers has changed drastically to accommodate the alignment of incentives between owners and managers. (Frydman et al., 2007; Bebchuck et al., 2005) From 1993 to 2003 the average CEO pay increased 146% and the average compensation for the top five executives rose in ten years 125%. (Bebchuck et al., 2005) It is argued that CEO and top executives compensation is related to firm size and performance. However, although firms have increased drastically in size between 1993 and 2003 the increase in size cannot account for the drastic increase in top management s compensation package only. Strikingly, the increase in compensation through share option plans was additional to the existing compensation package. And although public outrage would have occurred when the compensation rise consisted out of cash, the option plans convinced the public of the necessity. The necessity comes from the desire to align the incentives of management with the owners of the firm. Linking pay of managers to the performance of the firm is an attempt to make managers act in the best interest of the firm. However there is a second motive, as firms become larger management gets more powerful and is able to demand their compensation be increased to cope with the responsibilities. The managerial power approach can explain a large part of the increase in managerial compensation. (Bebchuck et al., 2007) Under the managerial power approach managerial compensation is not only a means against agency problems but also a cause of agency problems within the firm. The opposite approach, the arm s length approach or the optimal contracting approach, is not deemed to reflect reality. According to Bebchuck and Fried (2007) it is difficult for large firms to have regular contact with shareholders, since ownership is typically dispersed among many small shareholders. This gives management a large amount of power over their compensation package. The first approach views the compensation package as a way to incentivize managers. It is seen as a delicate process in which the board of directors in conjunction with the shareholders and market restrictions develop a compensation package that is in the best interest of the firm. However, the directors act not only in the interest of the firm, they also act out to enrich themselves, here the optimal 8 P a g e

contracting approach fails. In order to stay in function it is likely that directors will not argue with CEOs over the package and will ultimately develop compensation packages that are not in the interest of the firm. Thus, the managerial compensation package is seen as a way for managers to use their power to enrich themselves. And to create a compensation package that through low interest loans, stock option plans and many more incentive plans will not work in the interest of the firm and the shareholders but mostly towards the wealth of the managers. Shleifer and Vishny (1989) show an additional way in which managers are able to abuse compensation packages to benefit their own wealth. Due to the entrenchment of managers it becomes easier for the managers to control a variety of tasks. In their article managers are seen as agents that invest over and beyond the level that is value maximizing. The overinvestment is for the entrenchment of the managers. Due to entrenchment it is difficult to replace them and it is easier for managers to influence their pay and to invest in perquisites. It can also influence the way they payout free cash flow. And use this to further entrench their position. The Relationship between Payout Policy and Agency Costs Previously I have argued that managerial compensation is not only a way to minimize agency costs, it is also part of the agency problems that can arise within the firm. Whether directors want to secure a place on the prestigious board of directors or whether executive managers want to control the firm such that they can enrich their own wealth with company funds. The agency costs of managerial compensation are not in the best interest of the shareholder. When the managerial power approach described above reflects reality and managers are able to determine the course of the firm. It is more than likely that this will not only hurt shareholder through share prices, it will also hurt shareholders through the payout policy. The decision whether to payout cash or not to payout cash will be affected by decisions to further enrich management by also adjusting the payout policy towards the gain of the leaders of the firm. According to Hu and Kumar (2004) managerial entrenchment can affect the payout policy of the firm. They find that CEOs that hold large amounts of stock influence the decision to payout dividends negatively. Furthermore total payout is affected positively which shows that there is a difference between the decision to payout dividend and the decision to repurchase shares. In addition to share holdings, the study also shows that executive stock options influence the decision to payout dividend. Lastly, the study shows that large amounts of cash compensation increase the likelihood of dividend payments. While they control for many firm specific characteristics the results do not change drastically 9 P a g e

it can therefore be said that the decision to payout funds to shareholders is influenced by the entrenchment of executives and by the compensation package that is provided. In line with this study is the study by Fenn and Liang (2001).They show that there is a relationship between the payout policy of the firm and the amount of managerial stock options that are paid out by the firm. They find that payout rates are higher with firms that have higher agency problems; this is in line with the story of excess cash. Furthermore, their study shows that share repurchases are more common when there are a large number of stock options. Lambert et al. (1989) investigate the relationship between dividends and executive stock option plans. Their results suggest that dividends are reduced relative to expected dividends when executive stock option plans are put into effect. They suggest that executives therefore can affect the corporate dividend policy with their personal incentives. Since their stock options are usually not dividend protected they prefer other repayments of excess cash from which they can benefit. A contrary study done by Brown et al. (2006) shows that after the tax cut in 2003 executive managers that had higher ownership in the company were more likely to increase dividend payout. There even seems to be a substitution effect between dividends and share repurchases. Similar studies have been conducted in Europe and also here no consensus has been reached on the subject. A European-wide study by De Cesari and Ozkan (2013) shows that the findings are very similar to Fenn and Liang (2001). The European findings show that large executive stock option holdings are associated with lower dividend payments. And the fraction of share repurchases of total payout increases with the increase of stock options. However again no consensus has been reached a different study although only focusing on Finland shows that the opposite can also happen when the stock option are dividend protected as 41% of the stock options are in Finland. This leads to an increase in dividend payment. (Liljeblom and Pasternack, 2006) A last anomaly that breaks the consensus is the study by Brown et al. (2004) which shows that a dividend tax cut will change the outcomes of the research. Large managerial ownership is then positively related with dividend payouts and shows that taxation and managerial ownership have strong influence on the payout policy. Although there is no consensus on the matter there is evidence that a relationship exists between executive pay and the payout policy of the firm. Both corporate governance literature and payout policy literature suggest links between the different topics. Empirically many different results have been found. Clearly more research is needed. 10 P a g e

Methodology In order to research the relationship between the payout policy and the amount of incentive payment by the firm I have used data from the Compustat database accessible through the WRDS-databank. Through Compustat I have accessed data regarding the dividend payments and repurchase information as well as firm characteristics used as control variables. The sample period used in this thesis covers information that corresponds to the years 2003 through 2007. This is a relatively quiet period with steady economic growth and no crises that can cause the data to present wide outliers. Sample Selection I used the STOXX Europe 600 for the construction of my sample. This is a broad index including large cap, mid cap and small cap firms from many different industries and most European countries are represented in this index. Via the Compustat Global database I obtained the constituent list for the sample period of this thesis, 2003-2007. From this list I removed the financial firms (by sic code, removing 6000-6999) since these firms tend to have a very different payout policy as well as managerial incentive structures. This yields a sample of 488 firms and 2295 data points. For this sample I construct the variables, which yields a final sample of 286 firms and 1430 data points, after the deletion of firms with missing data. Variables Managerial incentive variables While Execucomp exists for information regarding compensation packages for American-based companies, no such database exists openly for European firms. In order to collect the necessary data for my thesis I hand collected the information by going through all 2295 annual reports. Here I collected information on the number of stock options and the number of shares managers hold. These will be my primary measures for managerial stock incentive pay. There are several different ways of constructing these variables. Some studies use chief executive officer (CEO) stock incentive pay as their measure for managerial stock incentives (Mehran et al., 1998; De Cesari, 2013). However, since the CEO is not the only manager on the executive board that makes the decision regarding the payment of dividend it is important to include all executive managers. Leaving out regular executive management would mean that managers with a voice in the company are left out. Other studies take into account not only executive management but also inside board members. (Denis et al., 1997) While this does take into 11 P a g e

account the whole executive management team this also includes board members, which in general only have a supervisory function and have only minor influence on the decisions made by the executive staff. In this thesis I use information regarding the whole executive team of the company when the information is available. Fenn and Liang (2001) have a similar approach stating that the corporate financial policy reflects the interest of all executive officers and not only the CEO. One of the problems that I encountered during the collection of stock incentive pay for European firms was the fact that legislation regarding the existence of a two-tier system within the management of the firm is not widely accepted. While the United Kingdom has a one-tier board system where one board consists of both executive and non-executive managers, Germany has a strict two-tier board system that separates the management board with executive managers from the supervisory board. (Jungmann, 2006) This difference does not only exist between the UK and Germany, many countries adopt either a one-tier or a two-tier system, and some countries adopt a mix of the two systems or have their own system. This makes it harder to determine the actual executive managers. In the sample most firms had a separate or clear distinction between executive and non-executive managers. However, mostly the south European companies adopt a one-tier system where only the CEO is on the board. This means that for most companies of Spain, Portugal, Italy and Greece only the CEO is represented in the sample. For the northern European countries there is a clear distinction and the whole executive board is represented in the sample. Furthermore, many companies apply several different incentive schemes for their managers, all existing of the option to get shares in the future. Apart from the usual self acquired shares either through options or bought and the normal share options granted at the completion of certain growth objectives, companies have also introduced many forms of incentive schemes, such as long term incentive plans, deferred shares and many more. During the collection of the data I have separated the many forms of incentive schemes from the regular share option schemes and stocks owned. However not all companies are as clear in there reports and therefore some observations might include extra option programs. In addition to the careful selection of the incentive programs I have also looked at the managers separately and left out the managers with exceptional holding in the company they run. This can be for example founders of the company or managers with large investments in the company. This could give distorted information about the holdings of the executive staff en might create outliers in the data. 12 P a g e

Table 1 below shows that management ownership in the form of options and shares does not exceed 1% of the total shares outstanding. Management share ownership is 0.4% of common shares outstanding and management option ownership is 0.7% of common shares outstanding. This is significantly smaller than a similar study that has been done for the companies in the United States. Fenn and Liang (2001) find that on average 6% of the common shares is held by managers and management stock options represent 2.3% of common shares outstanding. A more recent European study finds figures that confirm that the figures tend to be lower in Europe compared to the US study. De Cesari and Ozkan (2013) find similar results for their sample. Figure one and two in the appendix show a highly skewed distribution of for the incentive pay to executive management. Most companies tend to grant executive management between zero and two percent of the common shares outstanding. Both figures do not show extreme values in the ownership of managers. This is possibly due to the fact that extreme ownership by managers was filtered out in the data collection process. It could also signal that the entrenchment theory by Morck et al. (1988) does not apply to the companies in the sample. And that extreme holdings in the companies are more exception than rule. Payout Variables The primary payout variables are dividend payout and repurchase payout. The variables are constructed by taking the cash dividend payment to common shares divided by the market value of equity and by taking share repurchases divided by the market value of equity. Both cash dividends and share repurchases have been obtained through Compustat Global. Furthermore the cash dividend to common shares has been hand checked to account for special dividends and extra information on the cash dividends. With the hand check I found out that the database did not have the complete and correct data for every firm. Consequently, I was able to find extra data. I did a similar procedure with the share repurchases, making sure all available data was there. Furthermore, I separated the repurchases for stock option plans and employee trusts from the regular share repurchases for cancellation and included only the latter. 13 P a g e

Table 1 Sample distribution of payouts, managerial stock/option ownership and control variables The sample has 286 firms in the STOXX 600 index during 2003-2007. For all firms there are five years of data. Financial firms are excluded from the sample. Dividends are dividend payments to common shares. Repurchases are repurchases of common stock by the firm for the purpose of cancellation. Total payout is the sum of share repurchases and dividends. These variables are scaled by market value, which is the market capitalization at the end of each year. Management shares are the shares held by executive managers scaled by shares outstanding. Management options are the options held by executive managers scaled by shares outstanding. The control variables are Free Cash Flow, which is operating income before depreciation less capital expenditure. Profitability is operating income before depreciation. Free Cash Flow, Profitability, Operating Income Volatility are scaled by total Assets. The data are panel data with five observations for each company. Percentile Variables N Mean SD Min. 25th 50th 75th Max. Payout Variables Dividend Payout 1430 0,0669 0,1632 0,0000 0,0189 0,0322 0,4788 1,2115 Repurchase Payout 1430 0,0176 0,0482 0,0000 0,0000 0,0000 0,0122 0,3156 Total Payout 1430 0,0913 0,2071 0,0000 0,0237 0,0391 0,0656 1,3354 Repurchase share 1430 0,1618 0,2720 0,0000 0,0000 0,0000 0,2727 1,0000 Management Ownership Variables Management Shares 1430 0,0042 0,0144 0,0000 0,0000 0,0003 0,0014 0,0974 Mangement Options 1430 0,0070 0,0144 0,0000 0,0005 0,0024 0,0066 0,0914 Control Variables Free Cash Flow 1430 0,1014 0,0771-0,0839 0,0539 0,0925 0,1359 0,3599 Profitability 1430 0,1464 0,0734 0,0001 0,0992 0,1325 0,1821 0,4046 Log Assets 1430 8,7338 1,4904 5,9363 7,6028 8,5951 9,7149 12,0000 Retained Earnings Growth 1430 0,7851 1,9363 0,0000-0,0457 0,1808 0,7537 4,4660 Sales Growth 1430 0,0922 0,2062-0,4634-0,0007 0,0664 0,1481 1,1196 Operating Income Volatility 1430 0,0001 0,0002 0,0000 0,0000 0,0000 0,0001 0,0017 Debt to Assets 1430 0,2531 0,1544 0,0000 0,1372 0,2416 0,3468 0,7189 In the sample of this thesis dividend is the most used method to share the profit with shareholders, especially when looking at figure three and four it can be seen that dividend is used more although the magnitude of the payout is only small. Approximately 80% of the observations in the data payout between zero and six percent of their market value to shareholder. This exceeds share repurchases by a significant amount. Over 60% percent of the observations do not payout any funds through share repurchases. (See figure 4 in the Appendix) This is a remarkable outcome since it is said that repurchases are gaining on dividend. (Fama & French, 2001) For many of the companies in the sample share repurchases are not an annual way of payout to shareholders. Many companies use share repurchases as an extra way to payout profit and use dividends as the default way of sharing profits. The average total payout of the sample is approximately 10%, of this repurchases are only a small portion, 16.2%. An 14 P a g e

explanation for this low level of repurchases could be the inclusion of utility firms that tend to pay a high dividend. Furthermore, it could also signal that the advantages of share repurchases are not as large in this sample of companies as in other samples of other studies. Control Variables The payout policy is mostly a derivative of the free cash flow that is generated through the operations of the firm. Free cash flow can be used in a number of ways. It can be used to reinvest in new projects or it can be used by managers to enrich themselves through excessive spending of the free funds that have be created through the firm s operations. This last way of spending the free cash flow of the firm is the agency explanation of free cash flow. By distributing free cash flow to shareholders and other investors firms can counter the agency problems of managers trying to enrich themselves. Firms with large free cash flows will benefit greater by paying out excess cash and minimizing the chance of managers enriching themselves. Another explanation for the payout of free cash flow is the relatively expensiveness of holding cash. Firms with low marginal financing costs are likely to payout more cash since holding cash would be more expensive as attracting external funds. I control for free cash flow since this one of the sources of the ability of the firm to payout cash to investors. The free cash flow measure is built by earnings before interest taxes depreciation and amortization (EBITDA) minus capital expenditures scaled by total assets of the firm. (Fenn and Liang, 2001) By subtracting capital expenditures we take into account the investments made by the firm, and therefore creating the cash flow that is really kept within the firm and can be used to either reinvest or payout. I predict a positive relationship with the payout policy. Free cash flow theory suggests that firms with many investment opportunities tend to payout less to reinvest the cash into new profitable projects. As a proxy for possible investment opportunities I use sales growth. (Rozeff, 1982; Lehn and Poulsen, 1989)A negative relationship is predicted between payout policy and sales growth, since many investment opportunities would consume the free cash flows and leave little to distribute to shareholders. Furthermore, I control for profitability. Profitability is necessary in order to create any free cash flow that can be paid out to investors or used to reinvest. As a proxy for the profitability of the firm I use operating income before depreciation scaled by total assets. (E.g. Lemmon et al., 2008) If profitability is high then the chance to be able to payout more is also high. A positive relationship is predicted. Additionally, I include a measure for the size of the firm. Large firms tend to generate more free cash flow and are generally able to distribute more of their funds to shareholders. Furthermore, large firms 15 P a g e

have less information asymmetry and can therefore attract external financing at a less expensive rate. (Beck et al., 2005) As a proxy for firm size I use the log of total assets. In addition to size of the firm, I also include a control variable regarding debt levels. Firms with large debt levels are less likely to distribute profits and will rather use the profit to repay their debt. As a measure of debt I use debt-to-assets ratio. Distribution can also be done either partly of completely from retained earnings. To account for changes in retained earnings that might trigger the firm to payout cash, I include retained earnings growth. Excessive growth of retained earnings might play a role in the payout policy of the firm. The variable is expected to be positively related to the payout variables. Lastly, I include a control variable for uncertainty. Uncertainty about cash flows can withhold a firm to distribute cash to shareholders. Greater uncertainty might lead to retention of cash to able to deal with unexpected drops in income of rises in costs. As a proxy for uncertainty I use the volatility of operating income scaled by assets. (e.g. Lemmon et al., 2008) 16 P a g e

Empirical Results The empirical results shown below test the relationships between different elements of the payout policy and the incentive compensation to executives. In the regressions I control for several firms specific characteristics. A preliminary test of unconditional correlations is shown in table 2 below. Strikingly, the four main variables are all significantly and positively correlated. The correlations show that management stock option ownership has the highest correlation with dividend payout. Although all correlations are significant this is the largest correlation. Compared to other studies this is surprising since it is often stock options in compensation packages are not dividend protected. Management shares are less correlated with the payout policy of the firm however still positively and significantly. Again this is surprising since earlier studies have shown that management share ownership has little or no influence on the payout policy of the firm. Table 2 Correlation matrix of payout variables with managerial ownership The sample has 286 firms in the STOXX 600 index during 2003-2007. For all firms there are five years of data. Financial firms are excluded from the sample. Dividends are dividend payments to common shares. Repurchases are repurchases of common stock by the firm for the purpose of cancellation. Total payout is the sum of share repurchases and dividends. These variables are scaled by market value, which is the market capitalization at the end of each year. Management shares are the shares held by executive managers scaled by shares outstanding. Management options are the options held by executive managers scaled by shares outstanding. The data are panel data with five observations for each company. P-values are shown below the correlation coefficients. Variable Dividend Payout Repurchase Payout Management Shares Management Options (n=1430) Dividend Payout 1,0000 Repurchase Payout 0,2432 *** 1,0000 0,000 Management Shares 0,2335 *** 0,0855 *** 1,0000 0,000 0,000 Management Options 0,3633 *** 0,1211 *** 0,3166 *** 1,0000 0,000 0,000 0,000 ***, **, * signal 1%, 5% and 10% statitical significance respectively 17 P a g e

Table three shows the basic regression model of this thesis. The management ownership variables are tested on different elements of the payout policy of the firm. Firm specific characteristics are included to control for factors that might also influence the payout policy. The first striking result is that management share ownership does not seem to influence the payout policy of the firm, although they are almost significant at the 10% level. There appears to be no link between management ownership and the urge to payout more. The second striking result in the regression is the positive and significant outcome for management stock options. The first entry in the second row shows that an increase in management stock options increases the dividend payout by 3,56% when management option ownership increases with 1%. There is also a slight increase in the share repurchases when management stock options increase. The very significant outcome of the total payout entry show that the total payout increases significantly when the management stock options increase. Table 3 OLS estimates of determinants of payout policy The sample has 286 firms in the STOXX 600 index during 2003-2007. For all firms there are five years of data. Financial firms are excluded from the sample. The first entry is the regression coefficient. The second entry is the p-value. Independent Variables Dividend Payout Repurchase Payout Total Payout Repurchase Share Management Ownership Variables Management Shares 1,876 0,223 2,009 2,695 * 0,122 0,119 0,111 0,070 Mangement Options 3,564 *** 0,348 ** 4,069 *** -0,310 0,001 0,039 0,001 0,681 Control Variables Free Cash Flow Measure -0,112-0,014-0,147 0,108 0,175 0,544 0,153 0,504 Profitability -0,020 0,084 *** 0,111 0,580 *** 0,846 0,004 0,388 0,001 Retained Earnings Growth -0,006 ** -0,001 * -0,006 ** -0,004 0,020 0,065 0,044 0,329 Log Assets 0,007 *** 0,001 0,008 *** 0,007 ** 0,000 0,220 0,000 0,023 Sales Growth -0,079 *** -0,022 *** -0,101 *** -0,137 *** 0,000 0,000 0,000 0,000 Operating Income Volatility -53,257 *** -6,269-63,198 *** 13,710 0,001 0,233 0,002 0,699 Debt to Assets -0,018-0,006-0,034-0,126 * 0,617 0,546 0,446 0,056 R² 0,3100 0,1665 0,2931 0,3215 N 1430 1430 1430 1430 Firm Fixed Effects Yes Yes Yes Yes Year Fixed Effects Yes Yes Yes Yes ***, **, * signal 1%, 5% and 10% statitical significance respectively 18 P a g e

The results above show that dividend payouts increase with management share options. This might show that management stock options in the sample are largely dividend protected or are linked to the dividends that the firm pays out. Since management stock options are a large part of the incentive compensation for the firms in the sample it is likely that these influence the payout policy the most. The increase in repurchase payout can be explained by the dilution effect of the stock options. Due to this effect there is a small amount of shares repurchased to keep the share value intact. The firm specific characteristics do not show the signs that are to be expected according to earlier studies and theories. However, it can be explained. For dividend payout, profitability and free cash flow do not seem to be important. This can be explained by the stickiness of dividend. Lowering a dividend or failing to pay a dividend can be seen by investors as a bad sign. Therefore, management might choose to pay a dividend since the loss of failing to pay a dividend is larger than paying a dividend irrespective of the loss that might occur. Profitability is statistically significant for share repurchase payout. Although dividends tend to be sticky and not easily adjustable without signaling investors about the state of the firm, share repurchases are less sticky. They might be perceived by companies as a measure to distribute cash whenever necessary. In the data, share repurchases are not widely used and this may show that share repurchases do not substitute for dividend in the long run but are used as an extra measure to payout cash to shareholders. (Dittmar, 2000) This will most probably only be done when profitability is high, leaving enough cash. Three firm specific characteristics do show the predicted signs and are statistically and economically significant. Log assets a proxy for firm size shows that larger firms are likely to payout more. This is expected by many studies. Furthermore, sales a proxy for investment opportunities shows that when investment opportunities are abundant less cash is paid out to shareholders in order to benefit from the positive NPV projects. And finally uncertainty about the cash flows is strongly significant and shows that uncertainty about future cash flows decreases dividend payments. All results are measured with firm fixed and year fixed effects to control for within firm or within year anomalies. Taxes and Law Many studies in the field of payout policy and agency costs are being conducted on the American population of companies. Other parts of the world are often compared to American studies since the American economy is so large that it has to be a miniature world. However although there is numerous data about American companies, America cannot be held to be a miniature world. The sample of this thesis consists of companies that originate from Europe or even the European Union. Although Europe is 19 P a g e

a modern society with many similar practices in the field of business there are still some differences that influence policies within firms and governmental regulations that control firms. Earlier studies have shown that law and finance have a direct influence on each other. Law is a fundamental difference between America and Europe. While America employs a single law system, common law. Europe is a combination of four different kinds of law systems. Common law, which is employed in the United Kingdom and Ireland; French law, which is employed in France, The Netherlands, Spain and several other countries; German law, which is used in Germany and Austria and Switzerland and Scandinavian law which is used in the Nordic countries. These four law system are very different in their protection of shareholders and in the rules that are set for corporate governance. (LaPorta et al., 1998) Furthermore, and in a similar way attached to law is the taxation system. While this is closely related to the legal origin of the law system that is employed in each country, it does differ among nations. (Deloitte tax information worldwide) According to information from Deloitte there are many different dividend tax levels. In table 4 I have included both dividend taxation levels and law dummies to control for the special circumstances in Europe. Table 4 shows that dividend taxation does influence the payout policy of the firm. Higher dividend taxation leads to lower dividend payout, and lower total payout. It even leads to lower share repurchases. This is in line with the study of Brown et al. (2004) that show that the 2003 tax cut in the US increases dividend and replaces share repurchases. By including taxation and law dummies management share ownership becomes slightly statistically significant. By inclusion of taxation and law variables there are not many differences. Firm size, investment opportunities and cash flow uncertainty are still important determinants of payout policy as well as the management compensation package. The last three variables in the table are law dummies. Scandinavian law is excluded. We can see that in comparison to Scandinavian law countries all other law areas pay fewer dividends and have a lower total payout. This might be explained by the fact that Scandinavian law is relatively good to shareholders compared to the other legal origins. LaPorta et al. (1998) show that Scandinavian legal origin has the second best shareholder protection, in combination with relatively low average dividend taxation this can positively influence the payout policy. Common law origin and German law origin areas tend to payout less in comparison to Scandinavian law origin areas. French law origin areas have insignificant results and although they are not shareholder friendly according to LaPorta et al. they can be considered to payout the same as the Nordic systems. 20 P a g e