The effect of ownership concentration on
payout decisions
Evidence from the UK and Germany
Abstract
This study investigates the effect of ownership concentration on cash dividends and (gross) share repurchases. Using a large panel of 736 listed firms in the UK and 355 listed firms in Germany I estimate panel data tobit, logit and least square regression models. I find that ownership concentration has a positive effect on cash dividends in the UK, which is caused by a positive effect of ownership concentration on the firms’ decision how much to pay in case the firm is a dividend payer. However, ownership concentration has no effect on cash dividends in Germany. In addition, I find that ownership concentration has no effect on (gross) share repurchases in the UK. However, ownership concentration has a negative effect on (gross) share repurchases in Germany, which is caused by a negative effect of ownership concentration on the decision whether to repurchase shares or not.
JEL classification: G35
Key words: Agency problems, payout policy, cash dividends, share repurchase, monitoring, ownership concentration, large shareholders
Joost Dekker 1633783
July 2010
1. Introduction
Dividend payments are one of the most thoroughly researched topics in modern finance, because the dividend behavior of firms has puzzled researchers for decades. Easterbrook (1984, p. 650) stated “businesses find dividends obvious”, whereas “economists find dividends mysterious”. Therefore, a vast literature already exists and many theories have been developed which try to explain (i) why firms pay dividends, (ii) which factors play a role in the decision to payout dividends and (iii) which factors determine the amount of dividends paid. However, as pointed out by Mancinelli and Ozkan (2006) the agency approach has not received much attention. Therefore, this paper takes an agency perspective for explaining payout decisions of firms1; more specifically I will focus on the effect of ownership concentration on payout decisions of firms in the United Kingdom (UK) and Germany over the period 2006 to 2008.
Firms have to make two key decisions regarding their dividend payout (i) whether to pay dividends or not and (ii) in case they do pay dividends they have to decide how much to pay. In addition, there are two ways in which firms can payout cash to stockholders (i) cash dividends (ii) share repurchases. In this context this paper deals with four main questions. Firstly, what is the effect of ownership concentration on the firms’ decision whether to pay cash dividends or not. Secondly, what is the effect of ownership concentration on the firms’ decision whether to repurchase shares or not. Thirdly, what is the effect of ownership concentration on the amount of cash dividends paid by dividend paying firms. Fourthly, what is the effect of ownership concentration on the amount of share repurchases by repurchasing firms.
Traditionally, corporate dividend policy makes the assumption that the managers objective is to maximize the value of the firm as a whole. However, separation of ownership and control give rise to agency problems, because of information asymmetries between managers and shareholders. These agency problems can take many forms, for example to quote from an article in ‘The New York Times’2:
1 Modigliani and Miller (1961) propose that the dividend policy of a firm is irrelevant when markets are efficient
and there are no taxes, bankruptcy cost and asymmetric information. However, there are several theories that explain why dividends exists. For example; a taxation, a signalling, and a behavioral explanation. I refer to Frankfurter and Wood (2002) for an overview of some of these theories and empirical tests.
“NEW YORK — Richard Parsons, chairman and chief executive of Time Warner, owns a
small vineyard in Tuscany that produces a Brunello di Montalcino selling for $80 a bottle, adorned with a crest of the Parsons family. Twice a year, he boards one of his company's four jets to visit his 20 acres in Italy. When he does, Time Warner shareholders pick up the bill”.3
Agency theory suggest that dividend payments to shareholders makes it more likely that the firm must return to the capital market for funds which will lead to monitoring by the capital market (Easterbrook 1984). In addition, the payment of dividends reduces the amount of free cash flow and thereby the probability that cash is invested below the cost of capital or wasted on organizational inefficiencies (Jensen 1986). The agency explanations of corporate dividend policy of Easterbrook (1984) and Jensen (1986) are both based on the implicit assumption that dividend policy can be used as a corporate governance mechanism, acting as a monitoring or disciplining device. Moreover, the extent to which the firms’ payout policy is effective in reducing the expected agency costs may also depend on its ownership structure.
Ownership concentration can have two opposite effects on cash dividends, depending on whether ownership concentration reduces or increases agency conflicts in the firm. Theoretically, concentrated ownership can play an important role in reducing agency problems between shareholders and managers, following Villalonga and Amit (2006) I will call these ‘agency problems type 1’. When ownership is concentrated it is relatively easy for individual shareholders to require information from managers and hence, information asymmetries are lower. As a result less funds are invested in low return projects or wasted on organizational inefficiencies, resulting in more cash available to return to the shareholders. In addition, large shareholders may have the power to pressure management to increase payouts. Hence, ownership concentration has a positive effect on cash dividends. If ownership concentration is relatively low, shareholdings are more diffuse and there might be significant information asymmetries hence, there are high agency problems. In this case even if the
3 There are many other examples of agency problems within firms. However, Yermack (2005) argues that
shareholders are aware that managers are not acting in their interests it may be hard to arrange for example a proxy fight. Therefore, ownership concentration can play an important role in reducing the agency conflicts between managers and shareholders. However, some researchers more recently argue that ownership concentration can give rise to agency problems between large and minority shareholders, following Villalonga and Amit (2006) I will call these ‘agency problems type 2’. Large shareholders for example can have private benefits and therefore reduce the amount of cash dividends, because these are shared with minority shareholders. Hence, ownership concentration has a negative effect on cash dividends. The aim of this paper is to study which effect dominates in the UK and Germany.
I find that ownership concentration in the UK has a positive effect on cash dividends, which is caused by a positive effect of ownership concentration on the decision how much to pay in case the firm is a cash dividend payer. I find no evidence for a dominant effect of ownership concentration on cash dividends in Germany. The results regarding (gross) share repurchases show no dominant effect of ownership concentration on (gross) share repurchases in the UK. However, ownership concentration has a negative effect on (gross) share repurchases in Germany. This is caused by a negative effect of ownership concentration on the firms’ decision whether to repurchase shares or not. The results regarding the effect of ownership concentration on share repurchases should be interpreted with care, because they are based on (gross) share repurchases.
This paper adds something to the existing literature in several ways. Firstly, to my best knowledge there are no studies that compare the relationship between ownership concentration, payout policy and agency conflicts in the UK and Germany. Secondly, I include ownership concentration data for more years, which makes that I do not have to assume that ownership concentration is stable, an assumption made in other studies. Thirdly, in contrast to several other studies I will not only focus on cash dividends, but also study the effect of ownership concentration on (gross) share repurchases.
2. Literature review and testable hypothesis
This literature review consists of four parts. The first part shows how agency problems, corporate governance and dividends are linked. The second part deals with the relationship between ownership concentration and dividends and argues that ownership concentration can have two opposite effects on payout decisions. The third part is about the two forms in which firms can distribute cash to shareholders and explains why I also include (gross) share repurchases in my analyses. Finally, I develop the hypotheses regarding the association between ownership concentration and dividend payment decisions.
The link between agency problems, corporate governance and dividends
Jensen and Meckling (1976) were the first who specified agency problems within firms. They observe that if an owner sells stock to the public and his ownership is reduced below 100 percent, than the managers incentives increase to consume corporate resources for personal benefits4. “As the owner-manager’s fraction of the equity falls, his fractional claim on the outcomes falls and this will tend to encourage him to appropriate larger amounts of the corporate resources in the form of perquisites” (Jensen and Meckling 1976, p.313). In other words Jensen and Meckling (1976) suggest that managers of a firm have a tendency to expropriate cash from the firm when the separation between ownership and control increases. The reason is that the benefits for the managers are higher than the cost, because the costs are shared by varies shareholders. They further argue that perk consumption is affected by the difficulty of monitoring the managers actions. Separation of ownership and control thus results in diverging interests between managers and shareholders. As a result, control mechanisms are necessary to reduce the diverging interests between managers and shareholders of the firm.
Corporate governance is probably the widest control mechanism used for efficient utilization of corporate resources. It is aimed at reducing agency problems in firms and Denis and McConnell (2003, p. 2) define it as “the set of mechanisms that induce the self interested controllers of the company to make decisions that maximize the value of the firm to its owners”. Concentrated ownership is a key mechanism of corporate governance. Moreover, corporate governance mechanisms can be divided into internal and external mechanisms. Internal corporate governance can be obtained from the accounting information system, the
4 Jensen and Meckling (1976) also model agency problems between shareholders and debt holders, however I
auditing system, the organizational structure, and the remuneration structure. External corporate governance includes outside shareholders, debt holder monitoring, the takeover market, product market competition, and a countries legal system.
Since the paper of Jensen and Meckling (1976) many researchers try to link agency problems to the payment of dividends (Rozeff 1982, Easterbrook 1984, and Jensen 1986). Rozeff (1982) hypothesize that the dividend payout ratio is a function of agency problems among other variables like investment opportunities, risk and size. He uses two measures to capture the agency effects, the percentage of stock held by insiders and the number of shareholders. He hypothesizes an inverse relationship between the percentage of stocks held by insiders and payout. With disperse ownership the costs of monitoring managers become large for a single shareholder. Therefore, diffuse owners can utilize the discipline of the capital market to monitor the managers by forcing the payment of high dividends. He concludes that there is a strong relationship between dividend payout and the two agency proxies. Easterbrook (1984) argues that the payments of dividends force the firm to return to the capital market. The return to the capital market subjects the firm to the scrutiny of the capital market which reduces the possibility that funds are invested below the cost of capital. Therefore, monitoring of the capital market helps to ensure that managers act in the best interests of the shareholders. Several researchers expanded the model of Rozeff (1982). Lloyd, Jahera and Page (1985) included a third agency variable, the number of shares per shareholder. Their evidence confirmed the hypothesis of an agency influence on the dividend payout of firms. Therefore, the payout policy of a firm is one area of corporate decisions that cannot escape the influence of agency conflicts. Another argument, proposed by Jensen (1986) is that the payment of dividends reduces the amount of free cash flow in the firm. He argues that agency problems are more severe when there is excessive cash. The payment of dividends reduces the cash available for managers and therefore reduces the potential overinvestment problem and minimises agency problems between managers and shareholders.
The effect of ownership concentration on agency problems and dividends
legal protection. In countries where shareholders are relatively well protected (like in common law countries) concentrated ownership is less necessary, because small shareholders are well protected by law. In civil law countries shareholders are less protected and in response they take larger stakes in the firm to secure their interest. In a later paper, La Porta et al (1999) confirm that there are large differences in ownership concentration between countries. Although ownership is relatively dispersed in the US, in almost all other countries in the world ownership concentration is much more concentrated. In addition, ownership is relatively concentrated in Germany, but relatively dispersed in the UK.
Theoretically, ownership concentration can have two opposite effects on agency problems within firms. Firstly, ownership concentration may mitigate agency problems between managers and shareholders (agency problems type 1). According to Shleifer and Vishny (1986) agency problems between managers and shareholders, can be reduced by the presence of large shareholders, because these shareholders have greater incentives and more power to monitor management. Due to the monitoring of shareholders fewer resources are consumed in projects with a return below the cost of capital or wasted in organizational inefficiencies, implying that more cash flows can be distributed to shareholders as cash dividends. In addition, large shareholders may have the power to pressure management to increase payouts. Hence, a positive association between ownership concentration and cash dividends is expected in this case. However, one can also argue that ownership concentration has a negative impact on the amount of cash distributed to shareholders. Firstly, when concentrated ownership reduces or eliminates the agency problems, shareholders can be confident that free cash flows are not invested in projects with a return below the cost of capital or that cash is wasted on organizational inefficiencies. In this case it is not necessary to discipline management by paying out (costly) dividends. More significantly, large shareholders may prefer to retain cash rather than paying it out as dividends and sharing it with minority shareholders.
problems between large controlling shareholders and minority shareholders in the firm (agency problems type 2). Combining ownership and control allows concentrated shareholders to exchange profits for private benefits. Bebchuk and Kahan (1990, p. 1090) define private benefits as “any value captured by those controlling the company after the control contest and not shared among shareholders at large”. Johnson et al (2000, p. 22) call the expropriation of minority shareholders tunneling and define it as “the transfer of resources out of a company to its controlling shareholder”. Hence, large shareholders might prefer to retain cash rather than sharing it with minority shareholders and thus ownership concentration could also lead to lower dividends. To sum up, theory is ambiguous what the overall effect of ownership concentration is on agency problems in firms. As a result, ownership concentration can have a positive, but also a negative effect on dividends. Which of the two effects dominates remains an empirical question.
La Porta et al (2000) analyse the link between minority shareholder protection and dividend payouts. They discuss two models (i) the outcome model and (ii) the substitution model. Under the outcome model dividends are an outcome of an effective system of legal shareholder protection. The outcome model implies that dividends are paid, because minority shareholders pressure corporate insiders to disgorge cash. In the substitute model insiders interested in issuing equity in the future pay dividends to establish a reputation for good treatment of minority shareholders. The results of La Porta et al (2000) support the outcome model.
Cash dividends versus share repurchases
Until now I focused on the relation between cash dividends and ownership concentration. However, firms can also distribute cash to shareholders via share repurchases. Modigliani and Miller (1961) argue that dividends and share repurchases are perfect substitutes. With a fixed investment policy the residual cash can be distributed to shareholders either through cash dividends or by share repurchases. The main idea is that investment policy determines shareholder wealth, leverage and payout policy have no impact on the value of the firm. When a firm considers leverage and payout policy it is nothing more than slicing a fixed pie of cash flows into different pieces5. However, the argument of Modigliani and Miller (1961) is based on the assumption of perfect financial markets.
However, historically cash dividends have been the predominant form of payout, but several studies provide evidence of large changes in the payout policy of firms. More specifically, share repurchases represent a significant portion of total corporate payout nowadays. Bagwell and Shoven (1989) show that firms distribute cash to shareholders more and more via share repurchases6 in the mid 1980’s. They show that in the period 1973-1977 aggregate share repurchases in percentage of aggregate earnings is only 3.37 percent, over the period 1983-1998 this is on average 31.10 percent. They argue that the increase in share repurchases indicates that managers have learned to substitute share repurchases for dividends
5
One of the reasons why dividends and share repurchases might not be perfect substitutes is mentioned by Asquith and Mullins (1986). They argue that dividends and stock repurchases play somewhat different roles in signalling information to shareholders.
6 There are generally four methods to repurchase stock, a firm can repurchase stock via; (i) a Dutch auction
in order to generate lower taxed capital gains for stockholders. Fama and French (2001) confirm that dividend policies of US industrial companies changed and that there is a rapid growth in share repurchases. Even if they control for changes in the characteristics of firms that are publicly traded there is a significant decline in the propensity to pay cash dividends. Moreover, they show that firms that repurchase shares are typically cash dividend payers. The evidence, however, does not imply a reduction in dividends, but only a decline in the number of dividend paying firms. Indeed DeAngelo, DeAngelo and Skinner (2004) show that the total amount of dividends paid by firms increased and that there is a high and increasing concentration in the paying of dividends. Von Eije and Megginson (2008) find that the fraction of dividend paying firms declines, while total real dividends increase and share repurchases surge for EU15 countries. However, the decline in the fraction of dividend paying firms in EU15 countries started much later and was more rapid than in the US.
Skinner (2008) among others, finds evidence that share repurchase and cash dividends are substitutes. If this is true, including share repurchases in the analysis can provide a more complete picture of the effect of ownership concentration on payout decisions. Therefore, I will not only focus on the effect of ownership concentration on cash dividends, but also on the effect of ownership concentration on (gross) share repurchases.
use a net measure of repurchasing activities or a gross measure. This in turn depends on the situation in which share repurchases are used. Weston and Siu (2003) argue that if you want to compare share repurchases to cash dividends from firms to individuals, net repurchases are relevant.
Hypotheses
Based on the literature, I now develop the hypothesis regarding the effect of ownership concentration on (i) the decision whether to pay cash dividends or not and (ii) in case a firm is a cash dividend payer on the decision how much to pay. More specifically, I develop the following hypothesis regarding the effect of ownership concentration on the decision whether to pay cash dividends or not:
H0: There is no association between ownership concentration and the firms’ decision whether to pay a cash dividend or not.
H1: There is an (positive or negative) association between ownership concentration and the firms’ decision whether to pay a cash dividend or not.
In addition, I develop the following hypothesis with respect to the effect of ownership concentration on the amount of cash dividends in case a firm is a dividend payer:
H0: There is no association between ownership concentration and the amount of cash dividends.
H1: There is an (positive or negative) association between ownership concentration and the amount of cash dividends.
3. Data
I conduct my empirical analysis using data from 2006 to 2008, data for the independent variables is lagged by one year and therefore collected from 2005 to 2007. I begin my investigation period in 2006, because the most important independent variable ownership concentration which is lagged by one year is not fully available before 2005. I end my investigation period in 2008, because this is the last year for which cash dividends are available for most firms included in my sample. The sample is constructed using three data sources, Amadeus, Thomson Reuters DataStream and Thomson Reuters Worldscope. The ownership data is from the Amadeus, the data of the other dependent and independent variables comes from Thomson Reuters DataStream/Worldscope unless indicated otherwise.
My initial sample includes all listed firms in the UK and Germany that are available in Amadeus. All firms with missing ISIN numbers and firms without recorded shareholders are dropped from the original sample. In addition, shareholder information must be available for the years 2005 to 2008. In contrast to Harada and Nguyen (2009) the shareholders information in this paper does not cover one cross section, but the years 2005 to 2008. Because Harada and Nguyen (2009) only use shareholder information from one cross section, they need to assume that ownership concentration is stable. They argue that this is a reasonable assumption based on Prowse (1992). Smit (2009) makes a similar assumption and argues that stable ownership is reasonable in Germany, because Gugler and Yurtoglu (2003) and Da Silva et al (2004) show that ownership concentration at least in the past is stable. However, Wojcik (2001) provides evidence that the ownership structure in Germany changed over the period 1997 through 2001. He shows that the level of ownership concentration fell significantly over this period. One of the advantages of collecting the ownership concentration from 2005 to 2008 is that I do not have to make the assumption that ownership concentration is stable. If shareholder information is only available for one, two or three years the firms are excluded from the sample. To calculate the dependent and independent variables information about cash dividends, total assets, age, market capitalization, book value of debt, stock price and net income must also be available7. If data for one of those variables is
7 For a few UK firms, but for many German firms (gross) share repurchases are not available. I would loose
missing, the firms are again excluded from the sample. Table 1 gives a detailed description of the sample construction for the UK in panel A and Germany in panel B respectively.
Table 1
The table shows the construction of the sample for the UK (Panel A) and Germany (Panel B).
Panel A: UK Panel B: Germany
Year 2006 Year 2007 Year 2008 Year 2006 Year 2007 Year 2008 Number of listed firms available
in Amadeus 1342 2028 2107 605 850 962
Missing ISIN number -12 -17 -12 -4 -11 -5
No recorded shareholders -9 -35 -50 -6 -11 -43
No shareholder information for 4
years -480 -1135 -1204 -155 -388 -474
One of control variables missing -105 -105 -105 -85 -85 -85
Number of firms in sample 736 736 736 355 355 355
The dependent variables are defined in two ways. Regarding the first and second research question I use two dummy variables. Firstly, cash dividend payer no cash dividend payer (CDPNCP) which is equal to 1 if the firm does pay a cash dividend and 0 otherwise. Secondly, share repurchaser no share repurchaser (SRNSR) which is equal to 1 if the firm repurchase shares and 0 otherwise. A common measure for the third and fourth research question is the ratio of the amount of cash dividends to earnings. However, a disadvantage of this ratio is that negative earnings lead to a negative payout ratio which does not make sense. Therefore, I will use another proxy for the firms’ payout ratio. Firstly, the cash dividend to asset ratio (CDTA) which is defined as the ratio of cash dividend (Worldscope item WC04551) to total assets (Worldscope item WC02999. Secondly, the share repurchase to asset ratio (SRTA) which is defined as the ratio of gross share repurchase to total assets.
repurchases in this paper8. The consequence of using (gross) share repurchases is that the results regarding the effect of ownership concentration on (gross) share repurchases should be interpreted with care.
The independent variables in this paper are collected over the period 2005 to 2007. These variables are lagged by one period in an attempt to reduce the potential problem of endogeneity.The most important independent variable I use is ownership concentration (OC). Ownership concentration is measured as the sum of the percentage of outstanding shares held by the five largest shareholders of the firm9. Amadeus shows the ownership percentage as a ‘direct’ or ‘total’ percentage. A direct percentage of ownership means that firm A owns a certain percentage of firm B, while total consists of direct and indirect percentages. For example, firm A owns 60 percent in B and B 100 percent of C which means that A owns 60 percent in C. To calculate the percentage held by the top five shareholders, I sum the direct or total percentages held by the five largest shareholders. In some cases adding up the percentages of the top five shareholders results in a percentage of ownership over 100 percent. When this occurs the ownership variable is set at 100. In some circumstances Amadeus does not show a percentage of shares held by a shareholder, because it is unknown how much the shareholder owns. If it is a direct percentage Amadeus shows ‘-’, if the shareholder holds an indirect percentage ‘n.a.’ is shown. In both cases I replaced the ‘-’ and ‘n.a.’ into a value of 0. Although this might result in a too low ownership concentration for firms with missing data, I believe that it is the best solution to the missing data problem in this case. Moreover, firms are required to register their stake in the firm in case they own more than 3 percent in the UK and Germany. Therefore, replacing the percentage by zero would not result in a large error. In addition, besides the numeric values Amadeus can also show the terms ‘Wholly owned (WO)’ which indicates a percentage of ownership for a particular shareholder over 98.00 percent, ‘Majority owned (MO)’ indicating that 50.01 percent or more is owned by a shareholder, ‘Jointly owned (JO)’ indicating 50.00 percent is owned by a shareholder, ‘Negligible (NG)’ indicating 0.01 percent or less is owned by a shareholder or CQP1 indicating 50.00 percent plus one share is owned by one shareholder. To be able to make calculations I replaced the
8
This is also the reason why I focus on cash dividends and (gross) share repurchases separately and not on a variable total dividend payment. Adding cash dividends and gross share repurchases could result in a too high
total dividend payment, because gross share repurchases also include shares repurchased for employee stock
ownership plans for example.
9 In case a firm has less than five recorded shareholders, ownership concentration is measured as the sum of
terms by the percentage belonging to the category. Amadeus also shows the type of a particular shareholder10. If one of the first five shareholders falls in the category ‘public’, ‘Unnamed private shareholders aggregated’ or ‘other unnamed shareholders aggregated’ the percentage is subtracted from the total percentage of ownership. The reason for this subtraction is that the percentage of ownership belonging to these categories are not owned by one shareholder.
In addition to the ownership concentration variable, I include seven control variables in the regressions. The control variables have either been observed in the literature to have an influence on payouts or can be seen as managerial monitoring mechanisms. The seven control variables I include are used in various studies and are; (i) Size, (ii) Age, (iii) Risk, (iv) Leverage (Lev), (v) Profitability (Profit), (vi) Current growth (CG), and (vii) Growth opportunities (GO). Firstly, Lloyd, Jahera and Page (1985) and Fama and French (2001) among others show that size plays a role in the payout policy of firms. There are three standard proxies for the firms’ size; (i) the sales of the firm, (ii) the book value of assets and (iii) the market value of the firms’ equity. In contrast to Harada and Nguyen (2009) who measure size by the natural log of total assets and Chae, Kim and Lee (2009) who measure size by the logarithm of sales, I will proxy the firms size by the logarithm of the market capitalization of the firm (Worldscope item WC08001). Based on the results of Fama and French (2001) a positive association between market capitalization of the firm and the dependent variables is expected. Age is included as the second control variable and measured as the logarithm of the number of years between the year of analysis and the year of incorporation in Amadeus. Smit (2009) mentions that Amadeus sometimes provides misleading data for the year of incorporation in Amadeus, as an example he takes Daimler AG. The year of incorporation in Amadeus is 1998 however, this is the year of the merger between Daimler Benz AG with Chrysler Corporation. To solve this problem I collect the year of incorporation for the years 2005 to 2008 for each firm and take the earliest year recorded in Amadeus as the year of incorporation. Grullon, Michaely, and Swaminathan (2002) who studied the maturity hypothesis relate changes in dividend policy to a firms’ life cycle. They find that firms increase payouts as they mature and their investment opportunity set shrinks. Therefore, I expect that there is a positive association between and the dependent
10 These categories are: Bank, Financial company, Insurance company, Industrial company, Mutual & Pension
variables. The third control variable is risk, Von Eije and Megginson (2008) define risk as the standard deviation of net income divided by the sales of the firm. Chay and Suh (2009) use the volatility of stock returns measured as the standard deviation of monthly stock returns over the two most recent years as a proxy for cash flow uncertainty. They argue that the volatility of stock returns reflect the degree of cash flow uncertainty, because stock prices fluctuate more when cash flows are unpredictable. I will use the volatility of stock returns as a proxy for risk as well. However, I measure it as the standard deviation of monthly stock returns based on the total return index (RI in Thomson Reuters DataStream) over the past year. Monthly stock returns are calculated using the standard deviation of continuously compounded returns, see formula 3.
= −1 ln Re t t P P turn (1)
Pt = stock price at moment t Pt-1= stock price at moment t-1
current growth I (CG) as the level of assets scaled by last years’ assets. The growth in assets is a direct measure of current investment and, if investment is persistent, it is also a proxy for expected investment. The second is growth opportunities (GO) of the firm, which can influence payout policy, because firms with more growth opportunities might retain more earnings to finance the projects with internal money. This is confirmed by Fama and French (2001) who show that growth opportunities have an influence on the payout policy of firms. More specifically they show that firms that have never paid dividends have the best growth opportunities. Mancinelli and Ozkan (2006) and Von Eije and Megginson (2008) among others proxy growth opportunities by the market to book ratio. I will use the market to book ratio as a proxy for growth opportunities as well. Following De Cesari (2009), I calculate the market to book ratio as the sum of a firms market capitalization (Worldscope item WC08001) and total debt (Worldscope item WC03255) divided by total assets (Worldscope item WC02999). Current growth and growth opportunities are expected to be negatively associated with the dependent variables. Appendix A summarizes the proxies for the dependent and independent variables.
Table 2 provides some descriptive statistics for the sample firms for the UK (n=736) in panel A and for Germany (n=355) in Panel B. The correlation tables in Appendix B show that multicollinearity between the variables included in the regressions is not very likely.
Table 2
The table shows descriptive statistics for the UK (Panel A) and Germany (Panel B). ‘OC’ is defined as the sum of the percentage of shares held by the five largest shareholders ‘Size’ is defined as the logarithm of the market capitalization of the firm. ‘Age’ is defined as logarithm of the number of years between year of incorporation in Amadeus and year of analysis. ‘Risk’ is defined as the standard deviation of monthly stock returns based on the total return index. ‘Lev’ is defined as the book value of debt divided by the book value of total assets. ‘Profit’ is defined as net income divided by book value of total assets. ‘CG’ is defined as the relative change in total assets. ‘GO’ is defined as the sum of market capitalization and book value of total debt divided by total assets.
Panel A: UK Panel B: Germany
Mean Median Max Min Std. Dev. Mean Median Max Min Std. Dev.
The most important independent variable in this paper is the firms’ ownership concentration (OC). Table 3 shows some statistics for this variable for the UK in panel A and for Germany in panel B for the years 2005 to 2008. Firstly, ownership concentration is lower in the UK than in Germany. In the UK the average ownership concentration from 2005 to 2008 is around 49 percent, this is around 64 percent for Germany. This is in line with La Porta (1999) who shows that ownership concentration in the UK is relatively low and in Germany relatively more concentrated. Furthermore, table 3 shows that the mean ownership concentration is relatively stable in each year for both the UK and Germany. This suggests that the assumption of stable ownership made for example by Harada and Nguyen (2009) and Smit (2009) is quite reasonable. However, if I look more closely to the change in ownership concentration for each firm, a different picture occurs. Table 3 also shows that ownership concentration is far from stable in the UK and Germany if I look at the level of the individual firm. As can be seen, the number of firms with a stable ownership concentration is around 36 percent in the UK and around 44 percent in Germany. This means that more than half of the firms in the UK and Germany have changes in ownership concentration compared to the previous year that are larger than 10 percent. This suggests that assuming that ownership concentration is stable is less reasonable.
Table 3
The table shows some descriptive statistics for the variable ‘OC’, which is defined as the sum of the percentage of shares held by the five largest shareholders. In addition, the table shows how many firms have a ‘stable OC’ and for how many firms ‘OC changes’. ‘Stable OC’ represent the percentage of firms with for which the change in ‘OC’ is not larger than plus or minus 10 percent compared to the previous year. ‘Change in OC’ represent the percentage of firms with a change in ‘OC’ larger 10 percent compared to the previous year. All numbers are percentages.
Panel A: UK Panel B: Germany
4. Methodology
This paper deals with four main questions (i) what is the effect of ownership concentration on the firms’ decision whether to pay cash dividends or not (ii) what is the effect of ownership concentration on the firms’ decision whether to repurchase shares or not (iii) in case the firm does pay cash dividends what is the effect of ownership concentration on the decision how much to pay and (iv) in case the firm does repurchases shares what is the effect of ownership concentration on the decision how much to repurchase.
I will use three regression techniques to answer these questions, while making use of the panel data structure of my sample. Firstly, I will start with a panel tobit regression to study the effect of ownership concentration on cash dividends and (gross) share repurchases. This tobit regression can be interpreted as showing the combined effect of ownership concentration on the decisions whether to pay cash dividends / repurchase shares or not and on the amount in case a firm does is a payer or a share repurchaser. All firms are included in these regressions and the regressions are estimated while controlling for firm characteristics that have a known influence on the dividend decisions based on literature.
I will use two other regression techniques to study the effect of ownership concentration individually on (i) the decision whether to pay cash dividends / repurchase shares or not and (ii) on the amount of cash dividends / share repurchases in case a firm does payout. Firstly, I will run panel logit regressions to study the effect of ownership concentration on the decision whether a firm pays a cash dividend / repurchase shares or not. Secondly, to study the effect of ownership concentration on the decision how much to pay / repurchase in case a firm is a cash dividend payer / share repurchaser I will run panel least square regressions. In these panel least square regressions only firms with a ratio of cash dividends to total assets higher than zero are included. All these regressions are estimated while controlling for firm characteristics that have a known influence on these decisions based on the literature. More specifically, I will run the following regression equation:
DEPi,t = α0 + β1 (Oc)i,t-1 + β2 (Size) i,t-1 + β3 (Age) i,t-1 + β4 (Risk)i,t-1 + β5 (Lev)i,t-1 + β6
In which:11
DEP = CDPNCDP, SRNSR, CDTA and SRTA.
CDPNCDP = Dummy variable taking a value of 1 in case the firm pays a cash dividend and 0 otherwise.
SRNSR = Dummy variable taking a value of 1 in case the firm repurchase shares and 0 otherwise.
CDTA = The amount of cash dividends divided by the total assets of the previous year.
SRTA = The amount of share repurchases divided by the total assets of the previous year.
β = The slope of the coefficient.
Oc = The sum of the percentage of shares held by the five largest shareholders.
Size = The logarithm of the market capitalization of the firm.
Age = The logarithm of the number of years between incorporation in Amadeus and year of
analysis.
Risk = The standard deviation of monthly stock returns based on the total return index.
Lev = The book value of debt divided by book value of total assets.
Profit = The net income divided by book value of total assets.
Cg = The relative change in total assets.
Go = The sum of market capitalization and book value of total debt divided by total assets.
5. Results
This section shows the results for the panel tobit, panel logit and panel least square regressions. The tables show the results for the effect of ownership concentration on both cash dividends and on (gross) share repurchases.
Panel tobit regression
Table 4 reveals the results of the panel tobit regression. Ownership concentration has a significant (at 5 percent level) positive effect on cash dividends in the UK. Ownership concentration has again a positive effect on cash dividend in Germany, but here the coefficient is insignificant. These findings are in contrast with Renneboog and Trojanowski (2005) who find a negative impact of voting power on payout in the UK and with Kahn (2006) who finds that in the UK dividend payout decreases with an increase in ownership concentration. Concerning the control variables, all the signs of the coefficients except for growth opportunities are in line with my expectations and are in general highly significant. Size, age and profit have the expected positive effect, risk, leverage and current growth have the expected negative effect on cash dividends.
11
Ownership concentration has a positive effect on (gross) share repurchases in the UK, but the coefficient is not significant. In contrast, ownership concentration has a negative and significant (at 1 percent level) effect on (gross) share repurchases in Germany. Regarding the control variables, size and profit have the expected positive sign, risk, current growth and growth opportunities have the predicted negative sign. However, in contrast to my expectations age has a negative effect on share repurchases in both the UK and Germany, although the coefficient is only significant in Germany.
The table further shows that there are more share repurchase observations in the UK than in Germany. In the UK 36 percent (541 of 2208) of the observations have a ratio of share repurchases to total assets larger than zero. In Germany this is 23 percent (177 of 760) of the observations. Share repurchases are thus of relatively lower importance in Germany. The lower number of repurchases in Germany can possibly be explained by the fact that share repurchases are allowed for only 12 years in Germany. Before 1998 share repurchases were prohibited by law, German firms can therefore be less used to repurchase shares.
Table 4
This table shows the results of the random effects panel tobit regressions for both the UK and Germany, the dependent variable is defined as the amount of cash dividend in yeart divided by the total assets in yeart-1 in
column (1) and (2) and as the amount of (gross) share repurchases in yeart divided by total assets in yeart-1 in
column (3) and (4). All the independent variables are lagged by one year. ‘OC’ is defined as the sum of the percentage of shares held by the five largest shareholders ‘Size’ is defined as the logarithm of the market capitalization of the firm. ‘Age’ is defined as logarithm of the number of years between incorporation in Amadeus and year of analysis. ‘Risk’ is defined as the standard deviation of monthly stock returns based on the total return index. ‘Lev’ is defined as the book value of debt divided by the book value of total assets. ‘Profit’ is defined as net income divided by book value of total assets. ‘CG’ is defined as the relative change in total assets. ‘GO’ is defined as the sum of market capitalization and book value of total debt divided by total assets. The numbers in parentheses represent the p-values of the z-coefficients. Wald is the Wald chi-square of the equation for the 7 independent variables and prob indicates the significance of the equation.
Cash dividends (Gross) share repurchases
Panel logit regression
Table 5 shows the results for the panel logit regressions on decision whether to pay cash dividends or not and whether to repurchase shares or not. In the UK and in Germany ownership concentration has a positive effect on the decision to pay a cash dividend and on the decision to repurchase shares. However, the results are neither in the UK nor in Germany significant. The control variables size, age and profit have the expected positive sign and are highly significant. Risk, leverage, current growth and growth opportunities have the expected negative sign and are except for leverage significant.
Ownership concentration has a positive effect on the decision whether to repurchase shares or not in the UK and a negative effect on this decision in Germany. However, only the negative coefficient in Germany is significant (at 1 percent level). The control variables have again the expected signs and are overall significant. However, age has a negative and significant effect on the decision to repurchase shares in Germany, this is in contrast with my expectation. This result indicates that younger firms in Germany are more likely to repurchase shares than older firms.
Hence, the answer to the first research question is that ownership concentration has no effect on the decision whether to pay cash dividends or not neither in the UK nor in Germany. In addition, the answer to the second research question is that ownership concentration has no effect on the decision whether to repurchase shares or not in the UK. However, ownership concentration has a negative effect on this decision in Germany.
Table 5
This table shows the results of random effects panel logit regressions for both the UK and Germany. The dependent variable in column (1) and (2) is equal to 1 in case a firm does pay a cash dividend and 0 otherwise. The dependent variable in column (3) and (4) is equal to 1 in case the firm repurchase shares and 0 otherwise. All the independent variables are lagged by one year. ‘OC’ is defined as the sum of the percentage of shares held by the five largest shareholders. ‘Size’ is defined as the logarithm of the market capitalization of the firm. ‘Age’ is defined as logarithm of the number of years between incorporation in Amadeus and year of analysis. ‘Risk’ is defined as the standard deviation of monthly stock returns based on the total return index. ‘Lev’ is defined as the book value of debt divided by the book value of total assets. ‘Profit’ is defined as net income divided by book value of total assets. ‘CG’ is defined as the relative change in total assets. ‘GO’ is defined as the sum of market capitalization and book value of total debt divided by total assets. The numbers in parentheses represent the p-values of the z-coefficients. Wald is the Wald chi-square of the equation for the 7 independent variables and prob indicates the significance of the equation.
Cash dividend decision Share repurchase decision
Panel least square regression
Table 6 shows the results for the panel least square regressions on the decision regarding the amount of cash dividends or (gross) share repurchases for firms that payout cash. Ownership concentration has a positive effect on the amount of cash dividends in both the UK and Germany. However, only in the UK this effect is significant (at 10 percent level). Concerning the control variables, size has a negative effect on the amount of cash dividends in both the UK and Germany. However, the coefficient is significant only in Germany. Age has a positive effect on the amount of cash dividends in the UK, but a negative effect on the amount of cash dividends in Germany. Risk, leverage and profit have the expected signs. Current growth has the expected negative effect on the amount of cash dividends in the UK, the sign is positive but insignificant in Germany. In both countries growth opportunities have a positive sign which contrasts my expectations12.
Ownership concentration has a positive effect on the amount of (gross) share repurchases in the UK, but a negative effect on the amount of share repurchases in Germany. However, in both cases the coefficients are insignificant. Regarding the control variables size and age have a negative influence on the amount of share repurchases in both countries, but in both countries these coefficients are insignificant. Risk and leverage have the expected negative sign in both countries. Profit has a significant negative effect on the amount of share repurchases in Germany. Current growth has the expected negative sign in the UK and in Germany.
Hence, the answer to the third research question is that ownership concentration has a positive effect on the amount of cash dividends in the UK, but no effect on the amount of cash dividends in Germany. In addition, the answer to the third research question is that ownership concentration does not have an effect on the amount of share repurchases in the UK or in Germany.
The explanatory power of the panel least square (gross) share repurchase amount regression equations is smaller than the explanatory power of the panel least square cash dividend amount regressions in both the UK and Germany. However, again the repurchase equations are still significant according to the Wald test.
12 The substitute model of La Porta et al (2000) may provide a possible interpretation for this result. They argue
Table 6
This table shows the results of random effects panel least square regressions for both the UK and Germany if companies do payout. The dependent variable is defined as the amount of cash dividend in yeart divided by the
total assets in yeart-1 in column (1) and (2) and as the amount of (gross) share repurchases in yeart divided by
total assets in yeart-1 in column (3) and (4). All the independent variables are lagged by one year. ‘OC’ is defined
as the sum of the percentage of shares held by the five largest shareholders ‘Size’ is defined as the logarithm of the market capitalization of the firm. ‘Age’ is defined as logarithm of the number of years between incorporation in Amadeus and year of analysis. ‘Risk’ is defined as the standard deviation of monthly stock returns based on the total return index. ‘Lev’ is defined as the book value of debt divided by the book value of total assets. ‘Profit’ is defined as net income divided by book value of total assets. ‘CG’ is defined as the relative change in total assets. ‘GO’ are defined as the sum of market capitalization and book value of total debt divided by total assets. The numbers in parentheses represent the p-values of the z-coefficients using cluster robust standard errors for the individual firm. Wald is the Wald chi-square of the equation for the 7 independent variables and prob indicates the significance of the equation.
Cash dividend amount Share repurchase amount
Table 5 shows that the negative effect of ownership concentration on (gross) share repurchases in Germany (table 4) is driven by a negative effect of ownership concentration on the firms’ decision whether to repurchase shares or not. Table 6 shows that the positive effect of ownership concentration on cash dividends in the UK (table 4) is driven by a positive effect of ownership concentration on the amount of cash dividends in case the firm is a dividend payer.
Robustness check
To test the robustness of the above results I define another variable largest shareholder (LS). The variable largest shareholder represents the percentage of shares held by the largest shareholder in the firm, who is not of the owner type; ‘public’, ‘unnamed shareholders aggregated’ and ‘other unnamed shareholders aggregated’. Table 7 reveals the results for the panel tobit regression where in contrast to table 4 the largest shareholder is included as the test variable in stead of ownership concentration.
The results of these regressions are in line with the results of table 4. The largest shareholder has a positive and significant (at 10 percent level) effect on cash dividends in the UK. The largest shareholder has also a positive effect on cash dividends in Germany, but the coefficient is insignificant. In addition, ownership concentration has no effect on share repurchases in the UK, but a significant negative effect on share repurchases in Germany. I find similar results as I found in table 5 and 613 if I run the regressions with respect to (i) the decision whether to pay cash dividends / repurchase shares or not and (ii) the decision on the amount of cash dividends / share repurchases. The panel logit and panel least square regressions show that the positive effect of the largest shareholder on cash dividends in the UK is caused by the positive effect of the largest shareholder on the amount of cash dividends. In addition, the negative effect of the largest shareholder on (gross) share repurchases in Germany is caused by a negative effect of the largest shareholder on the decision whether to repurchase shares or not.
These results for Germany are in line with Goergen, Renneboog and Da Silva (2005) and Smit (2009) who find that the decision on cash dividend payments is not influenced by the presence of a large shareholder.
13 To preserve space I have not tabulated the results of the panel logit and panel least square regression with the
Table 7
This table shows the results of random effects panel tobit regressions for both the UK and Germany. The dependent variable is defined as the amount of cash dividend in yeart divided by the total assets in yeart-1 in
column (1) and (2) and as the amount of (gross) share repurchases in yeart divided by total assets in yeart-1 in
column (3) and (4). All the independent variables are lagged by one year. ‘LS’ is defined as the percentage of shares held by the largest shareholder. ‘Size’ is defined as the logarithm of the market capitalization of the firm. ‘Age’ is defined as logarithm of the number of years between incorporation in Amadeus and year of analysis. ‘Risk’ is defined as the standard deviation of monthly stock returns based on the total return index. ‘Lev’ is defined as the book value of debt divided by the book value of total assets. ‘Profit’ is defined as net income divided by book value of total assets. ‘CG’ is defined as the relative change in total assets. ‘GO’ are defined as the sum of market capitalization and book value of total debt divided by total assets. The numbers in parentheses represent the p-values of the z-coefficients. Wald is the Wald chi-square of the equation for the 7 independent variables and prob indicates the significance of the equation.
Cash dividend Share repurchase
6. Conclusion
Using panel data of over a 1000 listed firms I examine the effect of ownership concentration on payout decisions of firms in the UK and Germany over the period 2006 to 2008. Theory suggests that ownership concentration can have both a positive and negative effect on cash dividends and I expect that the same relations will hold for share repurchases. The aim of this paper is to examine which of the effects (the positive or negative) dominate in the UK and in Germany.
Firstly, I examine the effect of ownership concentration on (i) the decision to pay cash dividends or not and in case the firm does pay cash dividends (ii) on the amount of cash dividends. Based on the results I conclude that in the UK the positive effect of ownership concentration dominates with respect to cash dividends. This effect is mainly caused by a positive effect of ownership concentration on the decision with respect to the amount of cash dividends paid. However, I find no dominating effect in Germany. Ownership concentration has no effect on cash dividends in Germany. These results therefore do not support the view that large controlling shareholders expropriate other shareholders, but are at least in the UK more in line with the view that large shareholders use their power to increase cash dividends or that agency problems between managers and shareholder are reduced. These findings are in contrast with Kahn (2006) and with Renneboog and Trojanowski (2005) who find that in the UK dividend payout decreases with an increase in ownership concentration.
Secondly, I consider the effect of ownership concentration on (i) the decision whether to repurchase shares or not and in case the firm does repurchase shares (ii) on the amount of share repurchases. Skinner (2008) argues that share repurchases and cash dividends are substitutes. Therefore, including share repurchases in the analyses can provide a more complete view of ownership concentration on payout decisions. My results show no dominant effect of ownership concentration on (gross) share repurchases in the UK. For Germany I find that ownership concentration has a negative effect on (gross) share repurchases. This effect is mainly caused by a negative effect of ownership concentration on the decision whether to repurchase shares or not. Although share repurchases can complete the view of the effect of ownership concentration on payout decisions, the results should be interpreted with care, mainly because the analysis is based on (gross) share repurchases and not on net repurchases.
Suggestions for further research
7. Reference list
Asquith, P., Mullins, D.W., 1986, Signalling with dividends, stock repurchases, and equity issues, Financial Management, pp. 27–44.
Bagwell, L., Shoven, J., 1989, Cash Distributions to Shareholders, Journal of Economic Perspectives 3, pp. 129–140.
Barclay M.J., Holderness C.G., 1989, Private benefits from control of public corporations, Journal of Financial Economics 25, pp. 371-395.
Bebchuk, L.A., Kahan, M., 1990, A framework for analyzing legal policy towards proxy contests, California Law Review 78, pp. 1071-1136.
Berle, A.A., Means, G.C., 1932, The modern corporation and private property, New York Harcourt Brace.
Chae, J., Kim, S., and Lee, E.J., 2009, How corporate governance affects payout policy under agency problems and external financing constraints, Journal of Banking and Finance 33, pp. 2093-2101.
Chay, J.B., Suh, J., 2009, Payout policy and cash-flow uncertainty, Journal of Financial Economics 93, pp. 88-107.
Da Silva, L.C., Goergen, M., and Renneboog, L., 2004, Dividend policy and corporate governance, Oxford University Press.
DeAngelo, H., DeAngelo, L., and Skinner, D., 2004, Are dividends disappearing? Dividend concentration and the consolidation of earnings, Journal of Financial Economics 72, pp. 425– 456.
Denis, D.K., McConnell, J.J., 2003, International corporate governance, The Journal of Financial and Quantitative Analysis 38, pp. 1-36.
Easterbrook, F. H., 1984, Two Agency-Cost Explanations of Dividends, American Economic Review 74, pp. 650-659.
Fama, E.F., French, K.R., 2001, Disappearing dividends: Changing firm characteristics or lower propensity to pay? Journal of Financial Economics 60, pp. 3-43.
Frankfurter G.M., Wood B.G., 2002, Dividend policy theories and their empirical tests, International Review of Financial Analysis 11, pp. 111-138.
Goergen, M., Renneboog, L., and Da Silva, L.C., 2005, When do German firms change their dividends? Journal of Corporate Finance 11, pp. 375-399.
Grullon, G., Michaely, R., 2002, Dividends, share repurchases, and the substitution hypothesis, The Journal of Finance 57, pp. 1649-1684.
Grullon, G., Michaely, R., and Swaminathan, B., 2002, Are Dividend Changes a Sign of Firm Maturity?, Journal of Business 75, pp. 387–424.
Gugler, K., Yurtoglu, B.B., 2003, Corporate governance and dividend payout policy in Germany, European Economic Review 47, pp. 731-758.
Harada, K., Nguyen, P., 2009, Ownership concentration, agency conflicts, and dividend policy in Japan, SSRN Working Paper Series.
Jensen, M., 1986, Agency Costs of Free Cash-flow, Corporate finance and Takeovers, American Economic Review 76, pp. 323-329.
Jiraporn, P., 2006, Share repurchases, shareholder rights, and corporate governance provisions, North American Journal of Economics and Finance 17, pp. 35-47.
Johnson, S., La Porta, R., Lopez-de-Silanes, F., and Shleifer, A., 2000, Tunneling, American Economic Review 90, pp. 22-27.
Khan, T., 2006, Company Dividends and Ownership Structure: Evidence from UK Panel Data, Economic Journal 116, pp. 172-189.
La Porta, R., Lopez-de-Silanes, F., and Shleifer, A., 1999, Corporate ownership around the world, Journal of Finance 54, pp. 471-517.
La Porta, R., Lopez-de-Silanes, F., Shleifer, A., and Vishny, R., 1998, Law and finance, Journal of Political Economy 106, pp. 1113-1155.
La Porta, R., Lopez-de-Silanes, F., Shleifer, A., and Vishny, R., 2000, Agency problems and dividend policies around the world, Journal of Finance 55, pp. 1–33.
Lloyd, W., Jahera, J., and Page, D., 1985, Agency Costs and Dividend Payout Ratios, Quarterly Journal of Business & Economics 24 (3), pp. 19-29.
Mancinelli, L., Ozkan, A., 2006, Ownership structure and dividend policy: Evidence from Italian firms, The European Journal of Finance 12, pp. 265-282.
Maury, C., A. Pajuste, 2002, Controlling Shareholders, Agency Problems, and Dividend Policy in Finland, Working Paper, Stockholm School of Economics.
Mikkelson, W., Regassa, H., 1991, Premiums paid in block transactions, Managerial and Decision Economics 12, pp. 511-517.
Mitton, T., 2005, Corporate governance and dividend policy in emerging markets, Emerging Markets Review 5, pp. 409-426.
Pedersen, T., Thomsen, S., 1997, European patterns of corporate ownership: A twelve-country study, Journal of International Business Studies 28, pp. 759–778.
Prowse, S., 1992, The structure of corporate ownership in Japan, Journal of Finance 47, pp. 1121–1140.
Renneboog, L., Trojanowski, G., 2005, Control structures and payout policy, ECGI Working Paper Series in Finance.
Rozeff, M., 1982, Growth, Beta and Agency Costs as Determinants of Dividend Payout Ratios, Journal of Financial Research 5, pp. 249-259.
Skinner, D., 2008, The evolving relation between earnings, dividends, and stock repurchases, Journal of Financial Economics 87, pp. 582–609.
Shleifer, A., Vishny, R., 1986, Large Shareholders and Corporate Control, Journal of Political Economy 94, pp. 461-488.
Shleifer, A., Vishny, R., 1997, A survey of corporate governance, Journal of Finance 52, pp. 737–783.
Smit, M., 2009, The relationship between ownership and dividend policy. Evidence from Germany, Master thesis University of Groningen, Groningen.
Stephens, C., Weisbach, M., 1998, Actual share reacquisitions in open market repurchases programs, Journal of Finance 53, pp. 313–333.
Von Eije, H., Megginson, W.L., 2008, Dividends and share repurchases in the European Union. Journal of Financial Economics 89, pp. 347-374.
Weston, J.F., Siu J.A., 2003, Changing motives for share repurchases, University of California, Los Angeles, CA, working paper.
Wojcik, D., 2001, Change in the German model of corporate governance: evidence from block holdings 1997-2001, SSRN Working Paper Series.
Appendix A: Variable definition and data sources
Variable Variable definition Measurement period Source
CDPNCDP Equal to 1 if firm pays a cash dividend, 2006-2008 WC04551
and 0 otherwise.
SRNSR Equal to 1 if firm does repurchase shares, 2006-2008 WC Sale of com and pfd stk CFStmt
and 0 otherwise.
CDTA The amount of cash dividend divided by 2006-2008 WC04551t / WC02999t-1
total assets of the previous year.
SRTA The amount of (gross) share repurchases in £ 2006-2008 WC Sale of com and pfd stk CFStmt
and € divided by total assets of the previous year. / WC02999t-1
OC Logarithm of the sum of the percentage of shares 2005-2007 Amadeus
held by the five largest shareholders.
Size Logarithm of market capitalization of the firm. 2005-2007 WC08001
Age Year of analysis minus the year of incorporation 2005-2007 Amadeus
as in recorded in Amadeus.
Risk Standard deviation of monthly stock returns 2005-2007 Datastream (RI)
based on the total return index of the particular firm.
Lev Book value of debt divided by book value of 2005-2007 WC03255 / WC02999
total assets.
Profit Net income divided by total assets. 2005-2007 WC01250/WC02999
CG Relative change in total assets. 2004-2007 (WC02999t - WC02999t-1) / WC02999t-1
Appendix B: Correlation tables for the UK and Germany
Table 1
Correlations among independent variables in the UK. ‘Size’ is defined as the logarithm of the market capitalization of the firm. ‘Age’ is defined as logarithm of the number of years between incorporation in Amadeus and year of analysis. ‘Risk’ is defined as the standard deviation of monthly stock returns based on the total return index. ‘Lev’ is defined as the book value of debt divided by the book value of total assets. ‘Profit’ is defined as net income divided by book value of total assets. ‘CG’ is defined as the relative change in total assets. ‘GO’ is defined as the sum of market capitalization and book value of total debt divided by total assets. ‘OC’ is defined as the sum of the percentage of shares held by the five largest shareholders.
Size Age Risk Lev Profit CG GO OC
Size 1.00 Age 0.10 1.00 Risk -0.32 -0.17 1.00 Lev 0.21 0.06 0.01 1.00 Profit 0.25 0.05 -0.19 0.04 1.00 CG 0.02 -0.17 0.12 -0.01 0.01 1.00 GO 0.23 -0.18 0.11 -0.06 0.11 0.07 1.00 OC -0.20 -0.05 0.17 -0.01 0.02 -0.04 0.01 1.00 Table 2
Correlations among independent variables Germany. Correlations among independent variables in the UK. ‘Size’ is defined as the logarithm of the market capitalization of the firm. ‘Age’ is defined as logarithm of the number of years between incorporation in Amadeus and year of analysis. ‘Risk’ is defined as the standard deviation of monthly stock returns based on the total return index. ‘Lev’ is defined as the book value of debt divided by the book value of total assets. ‘Profit’ is defined as net income divided by book value of total assets. ‘CG’ is defined as the relative change in total assets. ‘GO’ is defined as the sum of market capitalization and book value of total debt divided by total assets. ‘OC’ is defined as the sum of the percentage of shares held by the five largest shareholders.
Size Age Risk Lev Profit CG GO OC
