The relationship between ownership type and firm performance

The relationship between ownership type

and firm performance

Andrej Jelenic (S1386220) Rijksuniversiteit Groningen Faculty of Economics and Business

Master Business Administration Finance

Supervisor: L. Dam

Table of Content

1. Introduction 4

2. Theoretical framework 5

3. Hypotheses 14

4. Data and Methodology 15

5. Results 21

6. Conclusion 29

7. References 31

1. Introduction

Ownership structure and firm performance has received much attention in the empirical literature for decades. Agency theory is used in previous research to try to explain the link between ownership concentration and firm performance. The earliest literature on this theory is by Berle and Means (1932) who were the first to emphasize the disadvantages of the separation of ownership and control in public corporations. Previous research on the effect of ownership concentration on firm performance shows mixed results.

In this paper we want to explain these mixed results by classifying the shareholders into groups. Some researchers claim that shareholder type is important with respect to the relationship between ownership and firm performance because every type has its own monitoring capabilities, objectives and incentives (Cubbin and Leech, 1983; Short, 1994). For example, the state as shareholder could pursue goals other than profit maximization which can harm the value of the firm. These political goals could involve the avoidance of job loss, producing environment friendly or producing goods and services that are economically not viable (Donahue, 1989; Hart, Shleifer and Vishny, 1997). Thus, we classify the largest shareholders into four types and research whether differences in the type of the largest shareholder of a company has a significant impact on the performance of the firm. These four types are individual/family, institutional investors, the state, and non-financial firms. Every group is similar to the groups used by La Porta, Lopez-de-Silanes and Shleifer (1999). These groups consist of the largest shareholder of a company because monitoring is costly and these costs can only be offset by large shareholders. The largest shareholder is therefore the most likely shareholder to have the incentive to monitor the management and to have an influence on firm performance.

Furthermore, some authors state that ownership concentration itself is endogenously determined (Demsetz and Lehn, 1985; Demsetz and Villalonga, 2001). This study will therefore also investigate the relation between ownership concentration and firm performance using the simultaneous equation model that is estimated with two-stage least squares to mitigate any endogeneity problems.

The data sample consists of listed non-financial German firms. The data is extracted from Amadeus and Datastream and covers the period from 2004 till the end of 2007. The three firm performance indicators are return on assets, Tobin’s Q and yearly stock returns. Germany is a suitable country for this study because it is one of the largest civil countries which is characterised by concentrated ownership (Franks and Mayer, 2001; La Porta et al., 1998). Therefore, there are many German firms with a large shareholder which we need to get a reliable sample.

The thesis is further organized as follows. First, we describe the theoretical framework the research is based on and address the relevant literature in the next section. Next, we discuss the data and the methodology. Finally, we discuss the results which are followed by the conclusion, references and appendix.

2. Theoretical Framework

Two theories are used to explain the link between concentrated ownership and the performance of a company: agency theory and the theory of expropriating minority shareholders.

Agency theory

Jensen and Meckling (1976) introduced agency theory, which states that conflicts between the owner and the managers can arise. The insulated positions the managers have in public corporations creates information asymmetry which disrupts the power balance between owner and manager. Moreover, the managers might take decisions that are harmful to shareholders when their goals are not aligned. While the shareholders demand a high performance of the company mostly in terms of return on investment, managers can have other goals. For example, they may engage in manager-specific investments to make it more difficult for the shareholders to replace the manager. This eventually leads to higher wages and perquisites for the entrenched managers (Shleifer and Vishny, 1989). Managers have the incentive to increase their company size even if this is reducing firm value (empire building) (Jensen, 1986). It will enlarge their status and it is more likely that they will receive higher compensations because then they have more resources under their control.

flow is, the less resources a manager has under its control which reduces his influence. Moreover, when a company has a low amount of free cash flow, it may need to borrow from the capital markets which results in additional monitoring from the debt holders (Jensen, 1986).

The managers should be monitored by the shareholders to avoid that they pursue other goals than value-maximization. The monitoring does not consists of only observing the managers, but it also consists of trying to influence the actions of the manager through restrictions, policies and rules (Jensen and Meckling 1976). It does only pay out for large investors to do the monitoring, because monitoring is costly and a part of these costs are fixed. For relatively small shareholders it is not worthwhile to monitor because the potential benefits (a higher performance of the company which results in a higher share price) do not outweigh the costs of monitoring. They also do not have enough votes to put enough pressure on the management to force change. A larger investor can offset these monitoring costs by the higher benefits the monitoring will produce. This means that only a relatively large shareholder generates sufficient incentives to monitor a company.

A large shareholder can monitor and act in three different ways according to the model of Shleifer and Vishny (1986). First of all, the shareholder can have informal negotiations with the management of the company. The effect of these negotiations depends on whether the shareholder has a controlling stake in the company. The second way is acquiring the firm when the shareholder is not satisfied with the performance of the company. With a larger stake, the potential benefits of the improvements in the company will be higher as well. When these potential benefits are higher than the costs of acquisition, it pays off for the shareholder to acquire the company and implement the necessary changes. Thirdly, when the large shareholder is not able to acquire the firm itself, it can try to find a third party outside the company and let them takeover the firm. The share prices before takeover will rise and both the large shareholder and the party who takes over the company will benefit from this construction (Shleifer and Vishny, 1986).

To summarize, because of the fixed costs it is too costly for the “smaller” shareholders in companies with dispersed ownership to monitor and we therefore expect a negative relationship between ownership concentration and firm performance.

not affiliated with the management have a significant effect on the probability of a hostile takeover. This finding confirms the view of Shleifer and Vishny (1986) concerning the three ways a large shareholder can act when monitoring, namely through informal negotiations, acquiring the firm or a takeover by a third party. According to Denis and Serrano (1996), firms where the hostile takeover fails, have a larger management turnover when ownership concentration is high. Kang and Shivdasani (1995) state that when the ownership of a firm consists of at least one large shareholder, it is more likely that a successor is chosen from outside the company.

Franks and Mayer (2001) find a relation between concentrated ownership and the disciplining of management in firms that perform poorly. The research done by Yafeh and Yosha (2003) shows that the presence of large shareholders also plays an important role in the spending behaviour of the company. They find that large shareholders actively monitor the managers and put pressure on them by imposing constraints on the level of expenditures in fields where managers can enjoy private benefits more easily. Sabherwal and Smith (2008) state that large shareholders in a concentrated ownership structure can be a substitute for other providers of information like for example financial analysts. This strengthens the view that blockholders monitor their company and provide the necessary information for the shareholders to protect their interests.

Expropriation of minority shareholders

Besides conflicts between owners and manager, conflicts among the owners themselves are possible (Shleifer and Vishny, 1997). In particular, large shareholders can expropriate minority shareholders in different ways and obtain control of voting rights that are well in excess of cash-flow rights (La Porta, Lopez-de-Silanes and Shleifer, 1999). These shareholders are not looking after the interest of all shareholders but they pursue their own interests. For example, they can expropriate minority shareholders through self-dealing. Self-dealing refers to the transfer of assets out of the firm to shareholders who have control (Ehrhardt and Nowak, 2003).

intermediate ownership level (Goergen, Manjon and Renneboog, 2008). Goergen et al. (2008) make this more easily to understand by the following example of a pyramid structure: “If

shareholder X owns 51% of the voting equity of firm Y which in turn owns 51% of the voting equity of firm Z, there is an uninterrupted control chain which gives shareholder X absolute majority control at each tier. Still, the cash flow rights of shareholder X in firm Z amount to only 26%” (Goergen et al. (2008, p. 8). This is a good example where the control rights exceed the cash-flow rights.

Franks and Mayer (2001) investigate the ownership structure of German companies and find evidence of the existence of controlling pyramids where the structure violates the rule of one share one vote. Another way to obtain voting rights in excess of cash-flow rights is through proxy votes. Banks in Germany can influence a company not only through direct ownership but also through proxy votes where other shareholders allow the bank to vote on behalf of them at shareholder meetings. Edwards and Nibler (2000) confirm that German banks do have an extra influence next to their direct ownership in other firms. They only find evidence for this in firms that are able to issue shares.

In short, the agency theory states that shareholders will only monitor when they have a high enough stake in the company to outweigh the cost of monitoring. This implies that there will be less monitoring in companies with a low ownership concentration which can result in lower firm performance. One the other hand, the theory on expropriation of the minority shareholder states that in companies with relative high ownership, large shareholders can pursue their own interests which can be harmful to minority shareholders and firm performance. The difference in performance between high and low concentrated ownership in firms may therefore be reduced or eliminated by the possibility of expropriation of the minority shareholders in firms with high concentrated ownership.

Ownership concentration and performance

variation in results concerning the relationship between ownership concentration and firm performance across countries. They ascribe the cause of these variations to differences in institutions and regulations. One study that finds significant effects regarding the relationship between ownership concentration and performance is the study of Edwards and Weichenrieder (1999). Their study shows a negative effect of ownership concentration on firm performance (ROA and market-to-book ratio) for several countries. They argue that the presence of a second largest shareholder lowers the ability of the largest shareholder to obtain private benefits because this second largest shareholder can act as a monitor of the monitor.

Other scholars who use simultaneous equation regressions also find conflicting results. Edwards and Nibler (2000) find that ownership concentration has a positive non-linear effect on the performance of a company while the studies of Demsetz and Villalonga (2001) and Himmelberg, Hubbard and Palia (1999) find no significant relation at all. Demsetz and Villalonga (2001) state that the studies which use single equations are biased because of their assumption that the ownership structure is exogenously determined. They refer to the work of Demsetz and Lehn (1985) which shows that ownership and performance can be used both as endogenous variables in a simultaneous model. This leads to the result that performance has a significant effect on ownership concentration instead of the other way around. Gedajlovic and Shapiro (1998) perform a study in 5 mayor countries around the world. They find a non-linear relationship between ownership concentration and performance which is negative at low levels of control but positive at high levels. There are also a small number of studies which use completely different measures of performance. Januszewski, Köke and Winter (2002) use productivity growth as an performance indicator. They estimate a production function with value added as dependent variable and use the residuals of the estimated regression equation as a measure of relative firm productivity. Januszewski et al. (2002) find that companies have a higher productivity growth when an ultimate owner is present. This result is not found when they investigate only the direct ownership of a firm. Köke and Renneboog (2005) perform this study with a much larger sample and find the same result. The above mentioned empirical evidence is inconclusive on the significant effects and the endogenous problem within the relationship between ownership concentration and firm performance.

degree of extracting rent. Others claim that the type of a shareholder is important because every type has its own monitoring capabilities, objectives and incentives (Cubbin and Leech, 1983; Short, 1994). Cubin and Leech (1983) make a distinction between internal (management) and external control where the external control can be subdivided into a single shareholder, an institutional investor and a firm in a (un)related industry. Edwards and Weichenrieder (1999) find that the actions taken by the shareholders depend on the balance between private benefits and the benefits associated with monitoring. When the private benefits are higher and more easily to acquire compared to the extra benefits through monitoring, it is more likely that the large shareholder will expropriate the minority shareholders instead of increasing the firm’s performance through monitoring. Every type of owner has different abilities in monitoring or enjoying private benefits and will therefore choose different paths. This research will divide the largest shareholders into four types with its own characteristics similar to the groups used by La Porta, Lopez-de-Silanes and Shleifer (1999). The groups are individual/family, institutional investors, the state, and non-financial firms. The next section will show empirical evidence on the relationship between these groups of owners and the performance of a firm.

Individuals and families

Numerous studies contradict the statement of Berle and Means (1932) that ownership concentration of a modern corporation is dispersed. La Porta et al. (1999) investigate how common widely held firms are in different countries around the world and find a complete other image. Their sample shows that in large firms only 36% is widely held and this percentage is lower for smaller firms. Using the 20% definition of control, 30% are family-controlled, 18% are state-controlled and the remaining 15% are divided between residual categories. This shows that families take a very important position in the ownership of companies. Families as largest shareholders are more common in countries with civil law (La Porta et al., 1999). The sample of Edwards and Nibler (2000) which shows the different types of largest shareholders in Germany for listed firms, has found that in 42% of the cases a family is the largest shareholder. Now the question is whether the type specific abilities, incentives and objectives of families as largest shareholder influence the performance of a company.

largest shareholder may use its influence to select a less capable family member as CEO instead of a ‘professional’ CEO supplied by the market, which can harm the performance of a firm. This family CEO may have the incentive to take decisions which are in the interest of him and his family instead of all the shareholders (James, 1999). Singell and Thornton (1997) confirm these expectations. They find that owners, who employ family in their business, maximize their own utility but earn less profit and therefore harm other non-family shareholders. Bennedsen, Nielsen, Perez-Gonzalez, and Wolfenzon (2007) also find that succession by a family CEO has a negative effect on the operating profitability of a firm and that these firms also underperform compared to non-family firms when Tobin’s q and accounting returns are used as performance measures. Cronqvist and Nilsson (2003) state that families are more likely to use control enhancing measures like dual-class shares and pyramid structures than any other type of owner categories. Their results show that value is most destroyed in terms of return on assets in firms which have families as controlling shareholder and are 50% less likely to be taken over in contrast to other companies. Morck, Stangeland and Yeung (2000) report that economic growth is lower in countries where capital is mainly controlled by a relative small group of wealthy families.

It may occur that in companies where an individual or family is the largest shareholder is also in the management of the firm. This implies that the principle who should monitor the agent and the agent could be the either same person or family related. This is an ideal situation for self dealing and we therefore expect that an individual or a family as largest shareholder has a negative impact on firm performance.

Institutional investor

Cable (1985) was one of the first who studied the influence of banks on firms. He finds that bank ownership and firm performance are positively related to each other in West-Germany. Gorton and Schmid (2000) their results are in line with Cable (1985), they find a positive relation between bank ownership and the market-to-book value of equity. Gorton and Schmid (2000) state that an additional commercial banking relationship with a firm next to direct ownership is not exploited by the banks to gain private benefits. Furthermore, they find no evidence of banks using the proxy votes for their private interest.

There are several studies that measure the effect of pension fund ownership on firms. Gillan and Starks (2000) study proposals sponsored by pension funds and find that these proposals obtain significantly more votes than any other type of investor. Del Guercio and Hawkins (1999) show that the proposals of pension funds lead to a higher number of governance actions like layoffs, restructuring and asset sales. Woidtke (2002) confirms the hypothesis that pension funds have an effect on firm performance. He finds a positive relation between private pension ownership and the relative firm value of a company measured by an industry adjusted Tobin-Q.

The State

According to La Porta et al. (1999), in 18.3% of the investigated large publicly traded firms around the world, the state has an ownership stake of at least 20%. Furthermore, the state is in 8% of the German firms the largest shareholder (Edwards and Nibler, 2000).

The state as owner is a special case, because in this situation it is clearer than any other type of owner that the state could have other objectives than profit maximization. Theories about this subject state that the state could pursue political goals which can harm the value of the firm (Shleifer and Vishny, 1994). These political goals could involve the avoidance of job loss, producing environment friendly or producing goods and services that are economically not viable (Donahue, 1989; Anastassopoulos, 1981; Hart, Shleifer and Vishny, 1997).

Most research available on state ownership is about the differences between wholly owned firms by the government and private firms. Little research has been done on the effect of the state as largest shareholder. Dewenter and Maltesta (2001) study the difference between government-owned and private-owned firms and find that government-owned firms are significantly less profitable. These firms also tend to use more leverage and are more labour-intensive compared to private-owned firms. Boardman and Vining (1989) show that state owned firms perform far worse than private owned firms in terms of a long range of performance indicators including efficiency. Thomsen and Pedersen (2003) investigate large European firms and find that ownership concentration has a significant negative effect on the value of a firm when the largest owner is the state. It is reasonable to conclude that the state could pursue other political goals instead of only profit maximization which can harm firm performance. Hence, we expect that the State as largest shareholders has a negative impact on firm performance.

Non-financial firms

Allen and Philips (2000) give a few reasons why companies as largest shareholder can have a beneficial effect on the target firm. When a firm needs to raise capital by selling equity, it could be less costly to sell equity to an informed party (a corporation in the same industry) lowering the asymmetric information and therefore also the cost of external capital. Furthermore, corporate equity ownership can give access to important technology and other resources which will improve the value of the “target firm”. There are also negative effects which are explained by Bertrand, Mehta and Mullainathan (2002). They argue that if the companies have business ties, expropriating of the minority shareholders could be more easily through favourable prices or extra service. The study of Allen and Philips (2000) demonstrates that firms in which other corporations invest in through equity, show higher industry-adjusted cash flows and investment expenditures. This applies only for firms which have high R&D, high advertising expenses or endure relative high information asymmetries and is especially present if the “target” firm has an alliance or joint venture with the corporate equity owner. For the sample as a whole, no significant effect is found.

In short, the results on the research of non-financial firms and its effect on firm performance are not consistent. We expect that the characteristics of this group are too diverse to get any significant results. Thus, we expect non-financial firms as largest shareholder to have no significant impact on firm performance.

3. Hypotheses

First we will test the relationship between ownership concentration and firm performance. As mentioned before, the presence of large shareholders in companies with concentrated ownership has advantages and disadvantages. Monitoring could improve the performance of a company while self dealing behaviour of dominant shareholders harms firm performance. The above mentioned empirical evidence is inconclusive on the significant effects in the relationship between ownership concentration and firm performance. There is also no consensus on the endogeneity problem. We therefore state the following hypotheses:

We also form hypotheses for the specific group of owners. As stated before, every type of owner has it own specific abilities to monitor and its individual incentive of enjoying private benefits as largest shareholder. We expect that the group of individuals/families and the state have a negative effect on firm performance while institutional investors have a positive effect on the performance of companies. Furthermore, we anticipate that non-financial firms as largest shareholders do not have a significant impact on firm performance. The hypotheses are therefore as follows:

H3: An individual or a family as largest shareholder has a negative impact on firm performance. H4: Institutional investors as largest shareholder have a positive impact on firm performance. H5: The State as largest shareholder has a negative impact on firm performance.

H6: Non-financial firms as largest shareholders have no impact on firm performance

4. Data and methodology

In this study we investigate the relationship between owner type and firm performance. The sample consists of listed non-financial German firms. The data is extracted from Amadeus and Datastream and covers the period from 2004 till the end of 2007. Financial firms are excluded from the sample because of their special capital structure and unlisted firms are not included because of data availability. There is no specific shareholder information available in the Amadeus database for firms that are not listed.

Germany is a suitable country for this study because it is one of the largest civil countries which is characterised by concentrated ownership as result of weak shareholder protection by law (Franks and Mayer, 2001; La Porta et al., 1998). La Porta et al. (1998) state that minority shareholders are reluctant to hold a minority position in a company when they are inadequately protected by law. They fear expropriation by the larger shareholders. This lowers the demand for newly issued shares of companies and indirectly stimulates ownership concentration.

German firms have therefore a high number of large shareholders, which we need to get a reliable sample. Furthermore, we do not have to deal with any country effect by selecting only German firms. According to Franks and Mayer (2001), the most common identities of large shareholders in Germany are respectively another firm, families and institutional investors.

own 20% or more of the company are considered dispersed firms. There are several reasons why the 20% mark is chosen. First of all, a threshold is chosen because the shareholders have to own enough votes to be able to impose their own objectives and influence the decision making of the management. A largest shareholder with only a few percent ownership would not be able to influence the decision making of the firm and therefore not have the power to expropriate the minority shareholders. Secondly, the minimal ownership stake of 20% is selected in line with other studies. La Porta et al. (2002) argue that by putting the minimum ownership at a relative high figure, the incentive for expropriation of the minority shareholders will be constant and the effect of the ownership from the largest shareholder on performance can then be completely credited to the identity of the owner and its incentive. The data sample ranges from 2004 till the end of 2007. These are the most recent years where the data is available and Amadeus does not provide data on firms before these years.

Firm performance measures

The dependent variables represent the performance indicators of the firms. Three measurements are used in this study, the return on assets, Tobin’s Q ratio, and annual stock returns. These three variables are the most used in previous research on this subject as described in the above literature review. The exact definitions of these dependent variables can be found in the appendix.

The largest shareholder of the firms are identified in Amadeus and assigned to one of the four owner groups: individual/family, institutional investor, the state, and non-financial firms. Furthermore, control variables are included in the test, which could have a possible external influence on dependent variables. First of all, leverage as measured by debt/equity ratio is expected to have an influence on the performance of a company. The relatively more debt a firm has, the lower the available free cash flow, which reduces exposure to agency problems. The liquidity ratio (cash/current liabilities) also plays a role in curbing the potential overinvestment by managers (Lie, 2000). The yearly sales of the companies are used as the control variable for size. Moreover, betas are added to the regression equation to control for the firm specific risk. Lastly, sales growth will be a control variable as well because of the positive significant effect on performance in previous research.

characterized by pyramid structures and therefore it is very complicated to find the identity of the ultimate owner of the pyramid. Amadeus has not always shareholder information available of these smaller companies. The information provider also assigns each shareholder to one of its thirteen types but they are not that reliable, especially not when pyramid structures are involved. In some cases, the type ‘financial’ or ‘industrial company’ was assigned to companies, which were wholly owned by a family or individual. The whole pyramid should therefore be thoroughly investigated to be certain that the right type of shareholder will be registered. The database Amadeus has several sources to register a shareholder of a company in its archive. These sources can be annual reports, the website of a company, private communications or other information providers. The complete information is recorded in the database with a certain date. The archived data remains unchanged and keeps its validity until some other information is received or found.

There are 962 companies that meet the above set criteria. Companies with an incomplete set of variables were excluded from the sample. The final sample consists of 1501 firm-year observations where 1089 are companies which have at least one shareholder who owns at least 20% and 412 firms which have a dispersed ownership concentration. 634 have an individual or a family, 104 have institutional investors, 66 have the state and 285 have a non-financial company as largest shareholder. A considerable amount of firm-year observations were not included in the sample because there was at least one missing variable.

The simultaneous equations model will be estimated using two-stage least squares because of the endogenous problem in the relationship between ownership concentration and firm performance. The control variables are used as instrumental variable. The simultaneous regression equations are defined as follows:

t 4 3 2 1 0 Beta Sales Equity Debt ion Concentrat Ownership Indicator e Performanc =α +α +α +α +α +ε (1) t 4 3 2 1 0 SalesGrowth s Liabilitie Current Cash Indicator e Performanc ion Concentrat Ownership =α +α +α +α +α_Sales + ε (2)

the firms are dispersed and takes the value one when the company has a shareholder with a stake of at least 20%. Debt/equity ratio shows the leverage of the company, cash divided by current liabilities is the liquidity ratio and sales growth shows the yearly growth rates of the firm’s sales. Beta and the natural logarithm of sales are the two last control variables in the regression equation.

Because the dependent variable ownership concentration is a dummy, we should combine a probit/logit model with a simultaneous equation model. We do not use a probit/logit model for this research because combining a probit/logit model with 2SLS and panel techniques would require programming capabilities. This approach will lead to possible biased standard errors though the coefficients still will be consistent.

We use the control variables excluding the sales variable as instrument variables in the simultaneous equation model. The debt equity ratio is put into regression (1) because of its effect on firm performance. Berger and di Patti (2006) find that high leverage reduces agency costs and increases firm value by constraining managers to act more in the interests of shareholders. Furthermore, beta is included in this regression. Beta is also used as an instrumental variable since we expect it to have a relationship with firm performance but not with ownership concentration. We argue that the size of a company, measured in sales, influences both the dependent variables. Fama and French (1992) prove that one of the factors that capture the variation in stock returns is size. On the other hand, size could have an impact on the ownership concentration of a firm. When a firm is relatively small, the additional needed capital can be attained from a small group of investors which keeps the ownership concentration high. The remaining two control variables, the liquidity ratio and sales growth will be added to the second equations resulting in that both equations are over identified which makes the Two-stage least squares model the appropriate model to use. One limitation of this research is that we assume that the control variables are exogenous. It can lead to a wrong estimation of the model if this assumption is not correct.

dummy. For sake of completeness, the two reduced-form equations for the simultaneous system of (1) and (2) are as follows:

t 5 4 3 2 1 0 Beta Sales s Liabilitie Current Cash Equity Debt Indicator e

Performanc =α +α +α +α +α +α_Salesgrowth+ε (3)

t 5 4 3 2 1 0 Beta Sales s Liabilitie Current Cash Equity Debt ion Concentrat

Ownership =α +α +α +α +α +α_Salesgrowth+ε (4)

Because the gathered data changes over time, we do not only estimate a pooled regression but we also use the panel data to estimate regressions using the random effects model. The fixed effect model can not be used because there are almost no ownership concentration changes in the dummy variable over time which makes the results of the model unreliable.

There is no endogenous problem for the relationship between ownership type and firm performance, because it is not plausible that a certain performance attracts a certain type of investor. We can therefore estimate the regression equations using Ordinary Least Squares (OLS) in order to test if the owner types have a significant effect on the chosen performance indicators. The regression equations consist of the several owner type dummies and the above mentioned control variables. The regression equations are defined as follows:

t 6 5 4 3 2 1 0 Beta Sales s Liabilitie Current Cash Equity Debt DUM Indicator e

Performanc =α +α +α +α +α _{Sales growth}+α +α +ε (5)

0

α _{is the intercept,}

6 1−

α _{are the slopes of the variables. The performance indicator represents one}

Descriptive statistics

Table 1 shows the descriptive statistics of the entire sample per variable. The Jarque-Bera tests shows that there is some evidence for non-normality so the standard errors might be somewhat biased, but the estimates are still consistent.

Table 1

Descriptive statistics of the collected data

These statistics are based on 1501 observations

ROA (%) Tobin's Q SR (%) DE CC SGR (%) BETA SALES (in mill.)

Mean 5,24 1,63 16,04 1,47 2,03 20,00 0,39 € 2.890 Median 5,05 1,27 13,25 1,29 0,33 7,18 0,29 € 116 Maximum 56,25 22,44 234,47 28,92 138,14 938,23 2,50 € 150.000 Minimum -40,18 0,11 -99,48 -103,42 0,00 -99,94 -0,74 € 0,001 Jarque-Bera 1125 202328 548 1303169 746675 278606 565 188101 Probability 0 0 0 0 0 0 0 0 Observations 1501 1501 1501 1501 1501 1501 1501 1501

Note: SR represents the annual stock returns, DE is the debt/equity ratio, CC states the liquidity ratio and Sgr shows the growth rates of the firm’s sales in percentages. The variable BETA represents the firm specific risk and sales are

the yearly sales in millions of Euros.

The sales of the companies in the data sample vary between €1.000 and 150 billion Euros in a single year. Not only sales shows these large differences between the minimum and the maximum, but almost all variables show great divergence.

companies with concentrated ownership and dispersed ownership are not that large for all the three performance indicators. Whether all these differences are significant will be clear in the results section. A correlation matrix of all the variables can be found in table A1 in the appendix. None of the variables are strongly correlated with each other. One of the strongest correlations is the variable beta with sales with a correlation of 0.47.

Table 2

Descriptive statistics per identity group

Averages of variables per owner identity group

ROA (%) Tobin's Q SR (%) DE CC Sgr (%) BETA Sales (in mill.)

Family 6,55 1,76 16,00 1,33 2,55 18,29 0,35 € 1.620 Inst. Investors 3,36 1,71 22,47 1,07 2,67 35,71 0,34 € 3.880 State 3,21 1,32 18,06 3,12 0,63 10,09 0,34 € 5.370 Company 5,13 1,46 13,79 2,00 1,04 16,95 0,27 € 1.280 CON 5,67 1,65 16,17 1,59 2,05 19,11 0,33 € 1.970 DISP 4,12 1,57 15,71 1,16 1,95 22,36 0,55 € 5.320

Note: Family refers to the group of individuals/families, Inst. investors are institutional investors, Companies are non-financial companies. CON represents the “concentrated” firms and DISP the “Dispersed” firms. SR is the annual stock return, DE is the debt/equity ratio, CC states the liquidity ratio and Sgr shows the growth rates of the firm’s sales in percentages. BETA represents the firm specific risk

and the variable sales shows the average sales of a group in Euros.

5. Results

First, the results concerning the relationship between ownership concentration and performance are discussed. Table 3 shows the results of estimating the regression using pooled data and 2SLS for the full sample. The dummy ownership concentration does not have a significant effect on any of the three performance indicators. None of the control variables are significant as well. When we turn the variables and put the performance indicators as explanatory variables instead of dependent variables, we see that ROA, Tobin’s Q and the yearly stock returns all have a negative significant effect on ownership concentration. This implies that the higher the performance of a firm, the lower the ownership concentration of that company. Tobin’s Q and the yearly stock returns are significant on a level of 1% and return on assets on a level of 10%. We can not reject hypothesis H1 which states that ownership concentration has no significant effect on firm performance. However, we can reject the second hypothesis which states that firm performance has no significant effect on ownership concentration.

table A4 in the appendix. According to the Hausman test, ownership concentration is endogenous in the regression equation with ROA and the yearly stock returns as dependent variable. This implies that the two-stage least squares model is mandatory for these two dependent variables while the regression with Tobin’s Q as variable could also be estimated using OLS. However, table A2 in the appendix shows that the ownership concentration dummy has no significant effect on Tobin’s Q using OLS.

Table 3

Ownership concentration and firm performance

Results estimating pooled data regressions of simultaneous system of equation

(1) Performance Indicator = α0 + α1Ownership concentration + α2(Debt/Equity) + α3Beta + α4Sales + εt and (2) Ownership Concentration = α0 + α1Performance Indicator + α2(Cash/Current Liabilities) +

α_{3Salesgrowth + α4Sales + εt. using Two Stage Least Squares (2SLS). Sample consists} of 1501 firm year observations (486 companies) in the period 2004-2007

ROA TQ SR C -105.69 -3.13 -514.90 (285.89) (21.73) (1529.02) OC 165.36 12.27 883.37 (479.88) (36.47) (2566.49) DE -0.41 -0.03 -2.28 (1.35) (.1) (7.21) Beta 53.53 4.31 295.36 (151.05) (11.48) (807.85) Sales -1.57 -0.31 -11.90 (6.47) (.49) (34.58) OC OC OC C -0.89 3.31 *** 0.63 *** (1.11) (.65) (.18) PI (*100) -17.50 * -66.57 *** -1.91 *** (10.19) (17.24) (.4) CC (*100) 1.09 0.25 0.48 * (.88) (.31) (.28) Sgr (*100) 0.07 0.04 0.08 ** (.08) (.04) (.03) Sales 0.13 -0.08 *** 0.02 * (.09) (.02) (.01)

PI is the performance indicator (same as dep. variable) DE is the debt/equity ratio, CC states the liquidity ratio and Sgr is the sales growth

variables. BETA represents the firm specific risk and Sales is the natural logarithm of the yearly sales. ROA refers to return on assets,

The regressions are also estimated using OLS and 2SLS in combination with the random effects model. These results are reported respectively in table A2 and A3 in the appendix. When we look at the results of the OLS regression, we find a positive relationship between ownership concentration and two of the performance indicators (ROA and yearly stock returns). This is the same result as from the research of Berle and Means (1932) and (Cubbin and Leech, 1983) who find a positive association between ownership concentration and accounting profitability without considering the endogeneity of ownership structure. Table A3 shows a significant and positive relationship between ownership concentration and return on assets. The other results in this table support the results in table 3. We find again a negative relationship between two of the performance indicators (Tobin’s Q and yearly stock returns) and ownership concentration.

Table 4

The relationship between ownership type and firm performance (OLS)

For each type, we estimate the regression equation (5):

t 7 6 5 4 3 2 Beta Sales s Liabilitie Current Cash Equity Debt DUM Indicator e

Performanc =α +α +α +α +αSalesgrowth+α +α +ε

with OLS (pooled data) using as performance indicator ROA, Tobin’s Q and yearly stock returns as dependent variables. Sample consists of 1501 firm year observations/486 companies (full sample) in a period of 2004-2007.

ROA TQ SR ROA TQ SR ROA TQ SR ROA TQ SR

Intercept -11,25 *** 3,93 *** 1,57 -8,68 *** 4,06 *** 1,97 -9,51 *** 4,05 *** 3,14 -8,85 *** 4,05 *** 2,50 (2,32) (,29) (8,08) (2,29) (,28) (7,89) (2,3) (,29) (7,96) (2,29) (,28) (7,9) Family 2,86 *** 0,16 ** 1,14 (,58) (,07) (2,02) Inst. Investors -1,85 0,06 7,08 * (1,13) (,14) (3,9) State -3,36 ** -0,05 2,97 (1,42) (,18) (4,9) Company -0,17 -0,10 -0,22 (,75) (,09) (2,58) DE 0,04 0,01 0,12 0,04 0,01 0,12 0,05 0,01 0,11 0,04 0,01 0,11 (,05) (,01) (,17) (,05) (,01) (,17) (,05) (,01) (,17) (,05) (,01) (,17) CC (*100) 5,85 * 0,32 23,44 ** 6,23 * 0,33 23,31 ** 6,22 * 0,34 23,51 ** 6,15 * 0,32 23,53 ** (3,21) (,4) (11,17) (3,32) (,4) (11,16) (3,23) (,4) (11,2) (3,24) (,4) (11,17) SGR (*100) 0,44 0,06 4,05 *** 0,42 0,00 3,91 *** 0,38 0,05 4,04 *** 0,39 0,06 4,04 *** (,37) (,05) (1,27) (,37) (,05) (1,27) (,37) (,05) (1,27) (,37) (,05) (1,27) BETA 1,38 * 0,43 *** 16,10 *** 1,27 0,43 *** 16,22 *** 1,12 0,43 *** 16,24 *** 1,28 0,41 *** 16,02 *** (,82) (,1) (2,84) (,82) (,1) (2,84) (,83) (,1) (2,86) (,84) (,1) (2,89) SALES 0,78 *** -0,14 *** 0,34 0,71 *** -0,14 *** 0,31 0,76 *** -0,14 *** 0,27 0,72 *** -0,14 *** 0,32 (,13) (,02) (,44) (,13) (,02) (,44) (,13) (,02) (,45) (,13) (,02) (,45) R-squared 0,05 0,06 0,04 0,04 0,06 0,04 0,04 0,06 0,04 0,04 0,06 0,04

Note: ROA represents return on assets, TQ is the Tobin’s Q and SR is the yearly stock return. Family refers to the group of individuals/families, Inst. investors are institutional investors, Companies are non-financial companies, DE is the debt/equity ratio, CC states the liquidity ratio and Sgr is the sales growth variables.

BETA represents the firm specific risk and Sales is the natural logarithm of the yearly sales. *** indicates a significant level of 1%

Individual/family

Table 4 shows the results of the performed regressions on the relationship between ownership type and firm performance. Companies where the largest shareholder is a family/individual perform significant better in terms of return on asset and Tobin’s Q with a significance level of respectively 1% and 5%. This contradicts the hypothesis which states that a family or an individual as largest shareholder has a negative impact on firm performance. However, there are previous studies who also have found this positive association between family control and firm performance. Maury (2006) compares the performance (ROA and Tobin’s Q) of companies who are family controlled and non-family controlled and finds that firms which have active family control have a significant higher profitability compared to non-family firms. Villalonga and Amit (2006) find that ownership by families is positive for firm value, but only when the founder of the company is still active as Chairman or CEO. Numerous other studies find similar results (McConaughy et al., 1998, James, 1999). Anderson and Reeb (1998) state that this outcome could be explained by the fact that the family’s wealth is related to the firm’s welfare. Families often have a significant part of their wealth invested in the firms and therefore have the incentive to monitor the managers and to maximize firm performance. Families also could have a better oversight of the company they invest in compared to the other groups. Most families or individuals are involved with the founding of the company and know their company from inside out. In Germany, a member of the controlling family serves as CEO in 50% of the 20 largest publicly traded firms (La Porta et al., 1999). Another characteristic of this group is that they have a longer horizon, because in most of the cases they have a significant part of their wealth invested in the company. With this long horizon, they are willing to invest in profitable long-term investment while other shareholders with a shorter horizon would forgo on these investments (James, 1999).

Institutional investors

(2002) although the other two performance indicators show no significant results. Other research also shows that institutional investors as shareholders are beneficial to the firm. Gorton and Schmid (2000) find that bank ownership, significantly improves firm performance compared to non-bank shareholders. Hartzell and Starks (2003) find that institutional ownership is positively related to the size of the performance based part of the compensation of executives which is an important factor in mitigating the agency problems between managers and shareholders. However, when we analyse the data in different ways to test the robustness of the result, we find opposite results. Table A5 in the appendix, shows the results of estimating the regression equations with the random effects model. We see similar results as in table 4. The institutional investor dummy only has a significant positive effect on the yearly stock returns. Table A6 and A7 show the results of analysing the data with a smaller sample (firms with concentrated ownership only. The positive relationship between institutional ownership and yearly stock returns is not significant anymore. Now, we see a significant negative result between institutional ownership and return on assets in both table A6 and A7. One explanation for this could be that the solid image institutional investors in general have attract others shareholders to invest in the company while in reality institutional ownership has a significant negative effect on return on assets. Gillan and Starks (2000) study the effects of proposals at shareholder meetings and find that the proposals from institutional investors in general receive more support than proposals from any other group. Duggal and Millar (1999) research active institutional investors and do not find any evidence of superior monitoring capabilities found by earlier scholars. Chaganti and Damanpour (1991) argue that institutional investors emphasize on short-term financial results because the performance measurement of institutional managers is assessed relatively frequent. This statement is confirmed by the research of Baker (1998) who finds that fund managers are affected by their performance evaluation in terms of the managers’ attitude to risk, motivation and time horizon. She concludes that the quarterly performance measurement of the managers stimulates the managers to take a more short-term view.

The State

earlier research. Dewenter and Maltesta (2001) study the difference between government-owned and private-owned firms and find that government-owned firms are significantly less profitable. Boardman and Vining (1989) show that state owned firms perform far worse than private owned firms in terms of a long range of performance indicators including efficiency. As argued before, the state as largest shareholder could use its voting power to pursue its own political goals that might harm firm performance. For the other two performance indicators, Tobin’s Q and the stock returns, the state ownership dummy is not significant. Thus, the question still remains whether the state has only profit maximization as a goal or pursues other goals like avoidance of job loss, producing environment friendly or producing goods and services that are economically not viable.

Non-financial firms

Lastly, we look at the results of non-financial companies as owner group. Table 4 shows that the dummy is not significant for any of the performance indicators. Non-financial firms as largest shareholder have no impact on the performance of a firm thus we can not reject hypothesis 6. We argue that this result is caused by the broad range of companies this group contains. It makes sense that such a broad group does not have uniform monitor capabilities or a higher incentive for expropriating minority shareholders. It is therefore logical that we do not find any significant results. Furthermore, with the numerous pyramid structures present in Germany, it is very complicated to find and label the ultimate owner which makes it more difficult to get a group with homogenous characteristics.

performance indicator and negatively significant in all regressions with Tobin’s Q as performance indicator.

Robustness

The analyses are performed in different ways to test the robustness of the results. The tables with the results can be found in the appendix. Table A6 shows the results of estimating the regression equation using OLS with a smaller sample (firms with concentrated ownership only), table A5 shows the results of estimating the regression equations with the random effects model using OLS (full sample) and table A7 shows the results of the same analysis as table A5, but this time performed only with firms that have concentrated ownership.

The positive relationship between when an individual or family is the largest shareholder and return on assets is the most robust results. All four tables display a significant positive relationship with return on assets on a significance level of 1%. The results of the institutional investors show mixed results. When we use the full sample, the relationship between the group of institutional investors as largest shareholder with the yearly stock return is positively significant. But with the smaller sample (only firms with concentrated ownership), the group only shows a significant negative relationship with return on assets.

The state as largest shareholder has a significant negative influence on the return on assets of a company when the analysis is done without the random effects model. However, the coefficient is still negative but not significant when the analysis is done with the random effects. Concerning the group of non-financial firms, we find very consistent results. In all four tables, we find no evidence of a significant relationship between non-financial firms as largest shareholder and any of the three performance indicators.

6. Conclusion

effect on firm performance. This is in line with previous literature which considers the endogeneity of ownership concentration.

Concerning the relationship between ownership type and firm performance, we find that companies with family/individual as largest shareholder show a significant higher performance (ROA and Tobin’s Q). As argued before, families often have a significant part of their wealth invested in the firms and therefore have the incentive to monitor the managers and to maximize firm performance. Families also have better oversight of the company they invest in compared to the other groups which improves their monitoring capabilities. Companies with institutional investors as largest shareholder show a significant higher performance for the stock return as performance indicator, but show a significant lower performance for return on assets when we test the robustness of the data with a smaller sample. It appears that the solid image institutional investors in general have, attracts other shareholders to invest in the company while in reality institutional ownership has a significant negative effect on return on assets. The negative relationship between institutional ownership and firm performance can be explained by the research of Baker (1998). She finds that the quarterly performance measurement of the institutional managers stimulates the managers to take a more short-term view which could harm the performance of the firm they invest in. Companies with the state as largest shareholder show a significant lower performance (ROA). This is in line with earlier research and the theories that the state uses its voting power to pursue its own political goals that harm firm performance. We find no significant results concerning the group of non-financial firms.

We can conclude that minority shareholders are best of in companies where an individual/family is the largest shareholder. It is difficult to point out if this is caused by their better monitoring capabilities or by the fact that the agency problem disappears when the owner and manager is one person or from the same family as it is regular the case in these firms.

data sample therefore covers no more than four years. The 4 groups that are chosen to type the shareholder can be seen as a limitation as well. The fact that the groups are too broad and general might have resulted in conflicting results, like the results of institutional investors. Future studies on this subject should classify the largest shareholders in to smaller groups which increases the heterogeneity of the group and will probably give more clear results. For example, the group of institutional investors could be further subdivided into pension funds, banks, and insurance companies to get more reliable results.

7. References

Allen, J.W. and G.M. Philips, 2000. Corporate equity ownership, strategic alliances, and

product market relationship, Vol. 55, no. 6, pp. 2719-2815

Anderson, R.C. and D.M. Reeb, 2003. Founding-family ownership and firm performance:

Evidence from the S&P 500, The Journal of Finance, Vol. 58, no. 3, pp. 1301-1328

Anastassopoulos, J.C., 1981, The French Experience: Conflicts with government, Redwood Burn Limited, Great Britain

Baker, M., 1998. Fund manager’ attitudes to risk and time horizons: the effect of performance

benchmarking, The European Journal of Finance, Vol. 4, pp. 257-278

Becht, M., 1997. Strong Blockholders, Weak owners and the need of european

mandatory disclosure, In: European Corporate Governance Network, the Separation of

Ownership and Control: A Survey of 7 European Countries. Preliminary Report to the European Commission, Brussels.

Becht, M. and E. Boehmer, 1999. Transparency of ownership and control in Germany, Working Paper

Bennedsen, M., K. Nielsen, F. Perez-Gonzalez and D. Wolfenzon, 2007. Inside the Family Firm:

the Role of Families in Succession Decisions and Performance, The Quarterly Journal of economics, Vol. 122, no. 2, pp. 647-691

Berger, A.N. and di Patti, E.B., 2006, “Capital Structure and Firm Performance: A New

Approach to Testing Agency Theory and an Application to the Banking Industry”, Journal of Banking and Finance, vol. 30, no. 4, pp. 1065-1102

Bertrand, M., P. Mehta and S. Mullainathan, 2002. Ferreting out tunneling: An Application to

Indian business groups, Vol. 117, no. 1, pp. 121-148

Boardman, A.E. and A.R. Vining, 1989. Ownership and performance in competitive

environments: A comparison of the performance of private, mixed and state-owned enterprises, Journal of Law and Economics, Vol. 32, pp. 1-33

Cable, J., 1985. Capital market information and industrial performance: The role of west

German banks, The Economic Journal, Vol. 95, pp. 118-132

Chaganti, R. and F. Damanpour, Institutional ownersjip, capital strcture, and firm performance, Strategic Management Journal, Vol. 12, pp. 479-491

Cho, M., 1998. Ownership structure, investment and corporate value: an empirical analysis, Journal of Financial Economics, Vol. 46, pp. 103-121

Chung, R., M. Firth and J. Kim, 2002. Institutional monitoring and opportunistic earnings

management, Journal of Corporate Finance, Vol. 8, pp. 29-48

Cornett, M.M., A.J. Marcus, A. Saunders and H. Tehranian, 2007. The impact of institutional

ownership on corporate operating performance, Journal of Banking and finance, Vol. 31, pp. 1771-1794

Cronqvist H. and M. Nilsson, 2003. Agency cost of controlling minority shareholders, vol. 38, no. 4, pp. 695-719

Cubbin, J. and D. Leech, 1983. The effect of shareholding dispersion on the degree of control in

British companies: Theory and measurement, Vol. 93, pp. 351-369

Del Guercio, D. and J. Hawkins, 1999. The motivation and impact of pension fund activism, Journal of Financial economics, Vol. 52, pp. 293-340

Demsetz, H., 1983. The structure of ownership and the Theory of the Firm, Journal of Law and Economics, Vol. 26, no. 2, pp. 375-390

Demsetz, H. and K. Lehn, 1985. The structure of corporate ownership: Causes and

Consequences, Journal of Political economy, vol. 93, no. 6, pp. 1155-1177

Demsetz H. and B. Villalonga, 2001. Ownership and corporate performance, Elsevier, Vol. 7, no. 3, pp. 209-233

Denis, D.J. and J.M. Serrano, 1996. Active investors and management turnover following

Dewenter, K.L. and P.H. Malatesta, 2001. State-Owned and Privately Owned Firms: An

Empirical Analysis of Profitability, Leverage, and Labor Intensity, The Amercian Economic Review, Vol. 91, no. 1, pp. 320-334

Donahue, J.D., 1989. The Privatization Decision: Public Ends, Private Means, Basic Books, ISBN 0-465-06358-6

Duggal, R. and J.A. Millar, 1999. Institutional ownership and firm performance: The case of

bidder returns, Journal of Corporate Finance, Vol. 5, pp. 103-117

Edwards, J. and M. Nibler, 2000. Corporate governance in Germany: the role of banks and

ownership concentration, Economic Policy, vol. 15, no. 31, pp. 239-267

Edwards, J.S.S. and A.J. Weichenrieder, 1999. Ownership concentration and share valuation:

Evidence from Germany, Cambridge Working paper 8285

Ehrhardt, O. and E. Nowak, 2003. The Effect of IPOs on German Family-Owned Firms:

Governance Changes, Ownership Structure, and Performance, Journal of small business management, Vol. 41, no. 2, pp. 222-232

Fama, E.F. and K.R. French, 1992, The cross-section of expected stock returns, The Journal of Finance, Vol. 47, no. 2, pp. 427-465

Franks, J. and C. Mayer, 2001. Ownership and control of German corporations, Review of Financial Studies, Vol. 14, no. 4, pp. 943-977

Gedajlovic, R. and D.M. Shapiro, 1998. Management and Ownership Effects: Evidence from

Five Countries, Strategic Management Journal, Vol. 19, no. 6, pp. 533-553

Gillan S.L. and L.T. Starks, 2000. Corporate governance proposals and

shareholder activism: the role of institutional investors, Journal of Financial Economics, Vol. 57, pp. 275-305

Goergen, M., M.C. Manjon and L. Renneboog, 2008, Is the German system of corporate

governance converging towards the Anglo-American model?, Vol. 12, no. 1, pp 37-71

Gorton, G. and F.A. Schmid, 2000. Universal banking and the performance of German firms, Journal of Financial Economics, Vol. 58, pp. 29-80

Hart, O., A. Shleifer and R.W. Vishny, 1997. The Proper Scope of Government: Theory and an

Application to Prisons, The Quarterly Journal of Economics, Vol. 112, no. 4, pp. 1127-1161

Himmelberg, C.P., G. Hubbard and D. Palia, 1999. Understanding the determinants of

managerial ownership and the link between ownership and performance, Vol. 53, no. 3, pp. 353-384

Holderness, C.G. and D.P. Sheehan, 1988. The role of majority shareholders in publicly held

corporations: an exploratory analysis, vol. 20, issue 1-2, pp. 317-346

James, H.S., 1999. Owner as manager, Extended horizons and the family firm, Economics of Business, Vol. 6, no. 1, pp. 41-55

Januszewski, S.I., J. Köke and J.K. Winter, 2002. Product market competition, corporate

governance and firm performance: an empirical analysis for Germany, Research in Economics, Vol. 56, no. 3, pp. 299-332

Jensen, M.C., 1986. Agency Costs of Free Cash Flow, Corporate Finance, and Takeovers, The American Economic Review, vol.76, no. 2, pp. 323-329

Jensen, M.C. and W.H Meckling, 1976. Theory of the Firm: Managerial Behavior, Agency Costs

and Ownership Structure, Journal of Financial Economics, Vol. 3, pp. 305–360

Kang, J. and A. Shivdasani, 1995. Firm performance, corporate governance, and top executive

turnover in Japan, vol. 38, pp. 29-59

Köke, J. and L. Renneboog, 2005. Do Corporate Control and Product Market Competition Lead

to Stronger Productivity Growth? Evidence from Market‐Oriented and Blockholder‐Based Governance Regimes, The Journal of Law and Economics, Vol. 48,

La Porta, R., F. Lopez de Silanes and A. Shleifer, 1998. Corporate ownership around the world, Harvard Institute of Economic Research paper no. 1840

La Porta, R., F. Lopez de Silanes, A. Shleifer and R. Vishny, 1998. Law and Finance, Journal of Political Economy, vol. 106, no. 6, 1113-1155

La Porta, R., F. Lopez de Silanes, A. Shleifer and R. Vishny, 2002. Investor protection and

corporate valuation, The Journal of Finance, Vol. 57, no. 3, pp. 1147-1170

McConaughy, D.L., Walker, M.C., Henderson, G.V. and C.S. Mishra, 1998. Founding family

controlled firms: efficiency and value, Review of Financial Economics, Vol. 7, pp. 1 –19

Morck, R., D. Stangeland and B. Yeung, 2000. Inherited Wealth, Corporate

Control, and Economic Growth: The Canadian Disease, Research working paper #6814

Pedersen, T. and S. Thomsen, 2003. Ownership Structure and Value of the Largest European

Sabherwal, S. and S. Smith, 2008. Concentrated shareholders as substitutes for outside analysts, Corporate Governance: An International Review, Vol. 16, Issue 6, pp. 562-577

Santos, J.A.C. and A.S. Rumble, 2005. The American keiretsu and universal banks: Investing,

voting and sitting on nonfinancials’ corporate boards, Journal of Financial economics, Vol. 80, issue 2, pp. 419-454

Shleifer, A. and R. Vishny, 1986, Large shareholders and corporate control, The Journal of political economy, vol. 94, no. 3-1, pp. 461-488

Shleifer, A. and R. Vishny, 1994. Politicians and Firms, The Quarterly Journal of Economics, Vol. 109, no. 4, pp. 995-1025

Shleifer, A. and R. Vishny, 1997. A survey of corporate governance, The Journal of Finance, vol. 52, no. 2, pp. 737-783

Singell, L.D. and J. Thornton, 1997. Nepotism, Discrimination, and the Persistence of

Utility-Maximizing, Owner-Operated Firms, Southern Economic Journal, Vol. 63, no. 4, pp. 904-919

Shivdasani, A., 1993. Board composition, ownership structure, and hostile takeovers, Journal of Accounting and Economics, vol. 16, no. 1-3, pp. 167-198

Short, H., 1994. Ownership, Control, Financial structure and the performance of firms, Journal of economic surveys, Vol. 8, no. 3, pp. 203-249

Thomsen S. and T. Pedersen, 2000. Ownership structure and economic performance in the

largest European companies, Strategic Management Journal, Vol. 21, pp. 689-705

Villalonga, B. and R. Amit, 2006. How do family ownership, control and management affect firm

value?, Journal of Financial Economics, Vol. 80, pp. 385-417

Walsh, J.P. and J.K. Seward, 1990. On the efficiency of internal and external corporate control

mechanisms, Academy of management Review, Vol. 15, no. 3, pp. 421-458

Woidtke, T., 2002. Agents watching agents?: evidence from pension fund ownership and firm

value, Journal of Financial economics, Vol. 63, pp. 99-131

8. Appendix

Definitions of dependent variables:

assets Current assets Fixed Taxation before (Loss) Profit assets on Return + = Assets Total debt Total tion capitaliza Market ratio Q s Tobin' = +

The market capitalization represents a market value while total debt and assets are book values.

) ln( ) ln( returns stock Yearly = Xt − Xt−1

Where X represents the stock prices, corrected for dividend payments at the end of year t.

Table A1

Correlation matrix

1581 variable observations in a period of 2004-2007

TOBIN'S Q STOCKR DE CC SGR BETA SALES

ROA -0,01 0,31 0,06 0,02 0,00 0,13 0,23 TOBIN'S Q 0,00 -0,01 0,01 0,00 -0,06 -0,09 STOCKR 0,03 0,04 0,00 0,19 0,12 DE -0,01 -0,01 0,01 0,06 CC 0,00 -0,06 -0,15 SGR -0,01 -0,03 BETA 0,47

Note: Stockr is the variable for yearly stock returns, DE is the debt/equity ratio, CC states the Liquidity Ratio and Sgr shows the growth rates of the firm’s sales. BETA represents the

Table A2

Ownership concentration and firm performance (OLS)

Results estimating the pooled regression equation

(1) Performance Indicator = α0 + α1Ownership concentration + α2(Debt/Equity) + α3Beta + α4Sales + εt and (2) Ownership Concentration = α0 + α1Performance Indicator + α2(Cash/Current Liabilities) +

α_{3Salesgrowth + α4Sales + εt.using OLS. Sample consists of 1501 firm year} observations/486 companies in the period 2004-2007.

ROA TQ SR C -8,59 *** 4,10 *** 8,11 (2,21) (,27) (7,66) OC 2,19 *** 0,13 4,45 * (,66) (,08) (2,3) DE 0,04 0,01 0,12 (,05) (,01) (,17) Beta 2,20 *** 0,49 *** 18,87 *** (,84) (,1) (2,92) Sales 0,61 *** -0,15 *** -0,15 (,12) (,02) (,43) OC OC OC C 0,93 *** 0,89 *** 0,90 *** (,09) (,09) (,09) PI (*100) 0,29 *** 0,43 0,01 (,1) (,83) (,03) CC (*100) -0,05 -0,04 -0,04 (,13) (,13) (,13) Sgr (*100) -0,01 -0,01 -0,01 (,01) (,01) (,01) Sales -0,01 ** -0,01 * -0,01 ** (,01) (,01) (,01)

PI is the performance indicator (same as dep. variable) DE is the debt/equity ratio, CC states the liquidity ratio and Sgr is the sales growth variables. BETA represents the firm specific

risk and Sales is the natural logarithm of the yearly sales. ROA refers to return on assets, TQ is Tobin’s Q and

Table A3

Ownership concentration and firm performance: 2SLS with random effects

Results estimating pooled data regressions of simultaneous system of equation

(1) Performance Indicator = α0 + α1Ownership concentration + α2(Debt/Equity) + α3Beta + α4Sales + εt and (2) Ownership Concentration = α0 + α1Performance Indicator + α2(Cash/Current Liabilities) + α_{3Salesgrowth + α4Sales + εt with random effects model using 2SLS. Sample consists of 1501 firm}

year observations/486 companies in the period 2004-2007.

ROA TQ SR C -266,15 ** 43,27 * -607,38 (107,47) (25,85) (395,73) OC 434,63 ** -63,32 1039,30 (181,5) (41,94) (665,26) DE -0,75 ** 0,03 * -2,50 (,34) (,01) (1,74) Beta 136,50 ** -18,74 343,41 (56,37) (12,7) (208,71) Sales -5,15 ** 0,61 -14,00 (2,45) (,51) (8,97) OC OC OC C 0,60 *** 2,20 *** 1,02 *** (,18) (,84) (,18) PI (*100) -0,17 -0,39 * -0,21 * (,26) (,23) (,11) CC (*100) 0,02 0,00 0,10 (,03) (,) (,07) Sgr (*100) 0,01 0,00 0,01 (,02) (,) (,01) Sales 0,01 -0,04 * -0,01 (,01) (,03) (,01)

PI is the performance indicator (same as dep. variable) DE is the debt/equity ratio, CC states the liquidity ratio and Sgr is the sales growth variables. BETA represents the firm specific

risk and Sales is the natural logarithm of the yearly sales. ROA refers to return on assets, TQ is Tobin’s Q and