Thesis MSc International Financial Management
Does the origin of law have an effect on the relation between
ownership concentration and firm performance? Evidence from
Western Europe
“What is the effect of a country’s law origin on the relation between ownership
concentration and firm performance?”
University of Groningen
Faculty of Economics and Business Country Studies Master thesis IFM
Abstract
This thesis explores the relationship between ownership concentration and firm performance and the effect of the origin of law on this relationship. Ownership concentration is measured as the percentage of shares owned by the largest shareholder. Tobin’s Q is used as a proxy for firm performance. A multiple regression analysis is executed on the sample of 2056 financial, non-utility publicly held Western European companies. This thesis finds that the relation between ownership concentration and firm performance is significant for our sample of Western European companies. In this sample the relation between ownership concentration and firm performance is shown as a U-shaped relationship. With this U-shaped relation, firm performance is first decreasing and later increasing as ownership concentration increases. The analysis shows no significant influence of the origin of law on the relationship between ownership concentration and firm performance.
Keywords: Ownership concentration, Firm performance, Origin of law, Western Europe First supervisor: dr. ing. N. Brunia Author: H.J. Nieuwland
Co-assessor: prof. dr. C.L.M. Hermes S2729377
H.J.Nieuwland@student.rug.nl
Table of contents
1.! Introduction 3
2.! Literature review 7
2.1. Ownership concentration 7
2.2. “Optimal” level of ownership concentration 13
2.3. Origin of law effect 14
3.! Methods 18
3.1. Sample 18
3.2. Variables and data sources 22
3.3. Methods 27
4.! Results 29
4.1. Results 29
4.2. Robustness test 38
5.! Discussion and conclusion 42
6.! References 47
7.! Appendix A: Pearson correlation matrix of Robustness test 50 8.! Appendix B: Spearman rank correlation matrix of 51
Robustness test
1.! Introduction
In 1932, the well-known Berle-Means thesis already discussed the relation between ownership structure and firm performance (Kapopoulos and Lazaretou, 2007). Berle and Means (1932) find that controlling ownership could enhance firm performance, based on the assumption that in a company with dispersed ownership it becomes harder to actively monitor and influence managers (Kapopoulos and Lazaretou, 2007). After that, research on ownership concentration and firm performance becomes an extensive researched (Thomsen, Pedersen, and Kvist, 2006), yet contradictory topic (Hu et al, 2008), as researchers often find conflicting results (Wang and Shailer, 2015)). Just as Berle and Means (1932), several researchers find a positive relation between ownership concentration and firm performance (Claessens and Djankov, 1999; Kapopoulos and Lazaretou, 2007). However, other researchers find a negative effect between ownership concentration and firm performance (Wang and Shailer, 2015). Moreover, several researchers find an inverted U shape relationship between ownership concentration and firm performance (Ding, Zhang and Zhang, 2007; Morck et al, 1988l; McConnel and Servaes, 1990). While Hu et al (2008) find a U-shaped relationship between ownership concentration and firm performance. An overview of these articles can also be found in table 1.
between managers and shareholders (Wang and Shailer, 2015). These ways of monitoring and disciplining managers is not possible when the shareholders have a relatively small stake in the company, as the shareholders would not have the means and incentives to control the managers. An explanation of the negative effect of ownership concentration on firm performance might be conflicting interests between controlling and minority shareholders, for which controlling shareholders could pursue private interests at the expense of minority shareholders (Wang and Shailer, 2015). This is also called the ‘entrenchment’ effect and can hurt the overall firm performance (Ding, Zhang, and Zhang, 2007). In case of a U-shape relationship, the expropriation effect takes place at the first part of the relation which means that when ownership concentration increases, firm performance decreases (Hu et al, 2008). After this, when the U shaped relation arrives at the lowest point, the increased ownership concentration causes the monitoring effect which increases firm performance (Hu et al, 2008). In case of an inverted U-shape, the alignment effect continues and increases firm performance, until a certain point, for which the ‘entrenched’ effect takes place and increasing ownership concentration causes a decrease in firm performance. This entails that the influence of large shareholders can both have a positive and negative effect on firm performance (Claessens et al, 2002).
(2017) find that the relation between ownership concentration and firm performance is more positive in countries with weaker shareholder protection compared to countries with stronger shareholder protection. Contrary, Maury (2006) argues that in countries firm performance can be improved with ownership concentration as with higher shareholder protection, controlling ownership lowers agency problems. However, in countries with weak shareholder protection high control might give rise to conflicts between the controlling shareholders and the minority shareholders (Maury, 2006).
Corporate governance mechanisms such as ownership concentration and investor protection are important, mainly because they can diminish the agency conflict. Following Villonga (2019), there are two situations in which agency problems can arise:
(I) A conflict of interest between the shareholders and the managers: agency conflict I . (II) A conflict of interest between large and small shareholders: agency conflict II.
Agency conflict I can be mitigated by large shareholders that have the power and the incentives to monitor managers and therefore, can make sure that the incentives of the managers are aligned with the shareholders (owners) of the company (Villonga, 2019). However, this gives rise to agency problem II. It has been discussed that focus of the traditional agency conflict between managers and owners (agency problem I) have now been shifted towards the conflict between controlling shareholders and minority shareholders (agency problem II) (Lepore et al, 2017; Villalonga, 2019), which often can be seen in companies with a high level of ownership concentration. This shift asks for research on the influence of ownership concentration.
extensively. We find that most studies are focused on one country, which make findings harder to generalize to other countries. Other studies are focused on an international sample, which could lead to heterogeneity in the sample. Wang and Shailer (2015) call for comparative studies in the area of ownership concentration and firm performance, as there exist different results in using national samples. Using a sample that contains several countries could give a more complete view. Additionally, Lepore et al (2017) call for an extensive Western European sample including common law countries, as common law countries have different investor protection and ownership structure than civil law countries. By focusing on Western European countries, we attempt to minimize the heterogeneity in macro-economic environment as these countries are from one economic zone. Additionally, studying Western Europe gives the unique opportunity to study the relation on ownership concentration and firm performance in one economic zone with different origins of law. This thesis attempts to answer the following research question:
(RQ) “What is the effect of a country’s law origin on the relation between ownership concentration and firm performance?”
methodology of this study will be discussed. After that the results of the analysis will be shown. We will conclude this thesis with a description of the main conclusions, discussion and contributions of this research.
2.! Literature review
In this part, an extensive literature review on the topic will be shown. The hypotheses are formulated based on the literature review.
2.1. Ownership concentration
(1999), a reason for this positive relation is that shareholders with large percentage of shares have the means and the incentives to properly monitor and discipline managers. Other reasons for a positive relation between ownership concentration and firm performance is that ownership concentration can be a “substitute for weak legal and institutional environments” (Wang and Shailer, 2015). Thus, ownership concentration could be an internal corporate governance mechanism in countries where there is weak investor protection (Lepore et al, 2017).
Table 1
Study Sample Measures Results
Morck et al (1988) Companies from the United States of America
Ownership concentration measured by shares held by management. Firm valuation measured by Tobin’s Q.
Inverted U shape between
management ownership and firm valuation
McConnell and Servaes (1990)
Companies from the United States of America
Ownership concentration by management ownership of common stock and firm performance by Tobin’s Q.
Inverted U shape between management ownership of common stock and firm performance.
Claessens and Djankov (1999).
Czech companies Ownership concentration percentage of shares owned by the top 5 investors. Firm performance measured by gross operating profit over net fixed assets plus inventory.
Concentrated ownership positively influences firm performance in Czech companies
Kapopoulos and Lazaretou (2007)
Greek companies Ownership concentration by percentage of shares largest shareholders (>5%) and management ownership. Firm performance by Tobin’s Q.
Concentrated ownership has a positive effect on firm performance in Greek companies.
Ding, Zhang and Zhang, (2007)
Chinese Enterprises Ownership concentration measured by the percentage shares held by the largest shareholder. Discretionary accruals are used as a proxy for earnings management.
Inverted U shaped relation between ownership concentration and earnings management in China.
Hu et al (2008) Japanese firms Ownership concentration measured by percentages of common stock owned by the 10 and 5 largest shareholders. Firm performance measured by ROA and Tobin’s Q.
U-shaped relation between ownership concentration and firm performance in Japan.
Wang & Shailer (2015)
Emerging markets* Ownership concentration as the percentage of shares held by the largest shareholder. Firm performance in accounting measures such as ROA and market measures such as Tobin’s Q.
Ownership concentration has a negative influence on firm performance in Emerging markets.
Lepore et al (2017) Italy, France, Germany and Spain
Ownership concentration is percentage shares held by the largest three shareholders. Firm performance by Tobin’s Q, profit margin and ROA.
Ownership concentration has a negative effect on firm performance in Western European countries.
As can be seen from the table Hu et al (2008) find a U-shape relation in their study on the relation between ownership concentration and firm performance in Japanese companies. This relation is consistent with the expropriation and monitoring effect. Hu et al (2008) state that they see a expropriation effect at low levels of ownership concentration and therefore a negative effect on firm performance in when there are low levels of ownership concentration. Additionally, they see increasing monitoring effect when the ownership concentration becomes larger and thus a positive effect of ownership concentration on firm performance when the ownership is highly concentrated. This relation shows firstly decreasing firm performance with concentrated ownership and later increasing firm performance with increasing concentrated ownership. This study also shows that in the Japanese sample, both highly dispersed and highly concentrated ownership can have a positive effect on firm performance (Hu et al, 2008).
inclusion of financial firms, modeling choices concerning performance and ownership measures and the inclusion of controls and the estimation method. Additionally, results could matter based on whether or not the article is featured in a refereed journal (Wang and Shailer, 2015).
We will follow the arguments of Lepore et al (2017) in our hypothesis. Lepore et al (2017) finds a negative effect of ownership concentration on firm performance consistent with the expropriation effect. This argument will be used in our hypothesis as this is a recent article that uses a similar sample as they also use Western European countries. Additionally, the firm performance measure is also similar as Tobin’s Q is one of the firm performance proxies that they use. However, Lepore et al (2017) use another measure for ownership concentration than is used in this thesis, namely the percentage of shares in the hands of the largest three shareholders. But as Wang and Shailer (2015) find that this will not matter, and therefore it is expected that using only the largest shareholder as proxy for ownership concentration will not influence our results. It is important to pick a study with a similar sample, estimation method and measures as there is much conflicting results on the research of the relationship between ownership concentration and firm performance. Wang and Shailer (2015) compared a great amount of research with each other and found that contrary findings on the relation between ownership concentration and firm performance are often the consequence of choices in sample, estimation method and measures. Therefore, it is important to compare our research which recent research that uses similar methods to our study. As Lepore et al (2017) concludes a negative effect of ownership concentration on firm performance.
From this, we arrive at the following hypothesis:
2.2. ‘Optimal’ level of ownership concentration
sample and what the ‘turning point’ of the (inverted) U-shape relationship between ownership concentration and firm performance is.
2.2. Origin of law effect
The research question of this thesis is to find whether the origin of law has an influence on the relation between ownership concentration and firm performance. There are two reasons to expect that the origin of law has an influence on this relation. Firstly, the difference in ownership structure between companies from common law countries and companies from civil law countries. Companies from common law countries have a more dispersed ownership structure compared to companies from civil law countries (Wang and Shailer, 2015). Table 2 shows that common law countries have significantly lower ownership concentration than companies from civil law countries in our sample. This is also consistent with previous research (Wang and Shailer, 2015). We acknowledge that there is also a relation between the origin of law and the ownership structure (Wang and Shailer, 2015). However, this relation goes beyond the scope of this research. However, we acknowledge that this is an important effect to be investigated in future research.
Median ownership concentration Mean ownership concentration
Common Law 20,1% 24,2% Civil Law 35,98% 40,4% Total Sample 30,25% 37,06% T-value: -13,55 DF: 2067 P-Value, two-tailed: 0,0000
The second reason to expect differences in the relation of ownership concentration on firm performance between common and civil law countries is the difference in degree of investor protection between the two law origins (La Porta et al, 2000). This is an important difference as the efficiency of concentrated ownership can differ for companies from countries with strong legal protection of shareholders than in countries with a weaker legal protection of shareholders (Liu, Yang, and Zhan, 2012). Lepore et at (2017) find that the relationship between ownership concentration and firm performance is stronger in countries with weak investor protection and that ownership concentration can act as a substitute for investor protection in countries with weak investor protection. As in common law countries strong investor protection laws acts as an external corporate governance mechanism. Corporate governance mechanisms are important as they assure “that people who invest their capital can receive returns on this capital” (Lepore et al, 2017). Additionally, these mechanisms can act against the expropriation by large shareholders or managers (Lepore et al, 2017). Contrary, in civil law countries where investor protection laws are weaker, ownership concentration acts as an internal corporate governance mechanism and substitute for investor protection (Lepore et al, 2017).
(La Porta et al, 2000). In this research we will merge the German, Scandinavian and French civil law into one civil law legal origin following Anderson and Gupta (2009).
The root of the difference in investor protection between the two origins of laws comes from the way in which legal rules are made (La Porta et al, 2000). In common law countries, rules are made by judges that look at precedents and general principles. Therefore, in the case of expropriation of investors (self-dealing) that has not happened before, the judge will investigate whether this is an unfair activity of the insiders to the outsiders. The fear that these rulings will lead to the expansion of legal precedents will limit the expropriation of outsiders in common law countries (La Porta et al, 2000). Contrary, the law in civil law countries is based on legislatures, whereas a corporate insider can find a way to expropriate outside investors that is not legally forbidden and does not have to fear for judicial ruling that is disadvantageous (La Porta et al, 2000). This makes common law countries more protective to investors, as new cases are judged by general principles of fairness, whereas in civil law countries the clear rules can be circumvented by insiders. From this we can argue that in countries with better shareholder protection ownership can have a positive effect on firm performance as minority shareholders should be less afraid of expatriation because of the investor protection.
ownership concentration. On the other hand, common law countries would, according to this theory have more external corporate governance mechanisms (Lepore et al, 2017). As the ownership concentration is far more dispersed and there is thus no internal corporate governance mechanism in place. However, as the common law countries experience stronger investor protection there is an external governance mechanism in place. From this theory, it would be suspected that in civil law countries the ownership concentration would have stronger effects on firm performance. As ownership concentration is needed as internal corporate governance mechanism. From this theory, ownership concentration could have two effects on firm performance, either positive or negative.
From the literature review we find conflicting evidence concerning the influence of the origin of law on the relationship between ownership concentration and firm performance. However, we will follow the reasoning of Lepore et al (2017) as their sample is also focused on Western Europe. So, we expect the relation between ownership concentration and firm performance to be more significant in countries with weaker investor protection and higher level of ownership concentration (civil law countries) and state that in general ownership concentration has a negative influence on firm performance. But in countries with weaker investor protection it can have a positive effect as ownership concentration can act as a substitute for investor protection.
Therefore, we arrive at the following hypotheses:
Hypothesis 2a:The relationship between ownership concentration and firm performance is different from countries with different law systems.
3.! Methods 3.1. Sample
In this part, the sample that has been used to test the hypotheses and to answer the research question will be discussed. As mentioned before, the sample will consider Western European countries because of the common economic area, and comparability in accounting standards. Therefore, heterogeneity within the sample will be minimized. However, as we acknowledge that there still exists heterogeneity between the countries we also include country dummies in our robustness test. Previous research (Anderson and Gupta, 2009; Maury, 2006) use the commonly used database of Foccacio and Lang (2002). This database considers the following Western European countries in their research: Austria, Belgium, Finland, France, Germany, Ireland, Italy, Norway, Portugal, Spain, Sweden, Switzerland, and the United Kingdom. Foccacio and Lang (2002) consider Luxembourg, the Netherlands and Denmark also as Western European countries, but they exclude these countries from their sample as they experience a lack of data availability for these countries. However, we have found data on these three countries and will therefore add these countries to our sample. Table 3 shows how the countries are divided into common law countries and civil law countries.
Common law origin Civil law origin
Ireland, The United Kingdom Austria, Belgium, Denmark, Finland, France, Germany, Greece, Italy, Luxembourg, the Netherlands, Norway, Portugal, Spain, Sweden and Switzerland
Table 3: The sample countries divided into common law countries and civil law countries. The classification of common and civil law countries is based on La porta (1998).
Selection criteria Search result in # of companies All active companies and companies with unknown situation 275,755,563 World region: Austria, Belgium, Denmark, Finland, France, Germany,
Greece, Italy, Luxembourg, the Netherlands, Norway, Portugal, Spain, Sweden, Switzerland, United Kingdom and Ireland
47,657,733
Excluding SIC codes: Financial industry (SIC 6000-6999), Regulated utilities (SIC 4900-4999)
32,497,404
Category of companies: very large, large and medium active firms with recent financial information
1,460,782
Publicly listed companies 4,552
Table 4: Search strategy for data from Orbis. With each step it is specified how many observations were selected.
Home country Before missing
values After ownership After financials
Own (Average) TQEQ (Average) TQEN (Average) Common Law 1,207 766 416 24.2% 2.2 1.6 Great Britain 1,132 723 397 24.3% 2.2 1.6 Ireland 75 43 19 23.5% 2.3 1.6 Civil Law 3,345 2,686 1,640 40.4% 2.1 1.5 The Netherlands 130 84 52 31.1% 2.3 1.6 Belgium 104 83 55 49.3% 2.2 1.5 France 637 470 319 51.1% 1.9 1.4 Germany 486 369 260 45.8% 2.2 1.6 Luxembourg 51 36 17 42.3% 1.6 1.3 Switzerland 151 114 99 34.2% 2.6 1.7 Austria 47 33 26 53.6% 1.6 1.3 Spain 140 108 53 30.3% 2.1 1.4 Portugal 46 37 22 42.4% 1.3 1 Greece 160 152 99 43.5% 1 1 Sweden 709 604 269 28.1% 2.9 1.9 Finland 137 123 97 27.8% 2 1.5 Norway 165 139 65 33.5% 2.1 1.5 Denmark 118 84 58 34.3% 2.8 1.9 Italy 264 250 149 46.7% 1.8 1.3 Total 4,552 3,452 2,056 37.06% 2.2 1.5
3.2. Variables and data sources
In the next section, the definitions, measures and data sources of the variables that are used in this research are described.
Independent variable
Dependent variable
2009). Thus, for the objective of this study, researching the relationship between ownership concentration and firm performance and the influence of the origin of law on this relationship a market performance measure such as Tobin’s Q will be more appropriate than accounting based performance measures. The construction of the dependent variable Tobin's Q for the enterprise (TQEN) and Tobin’s Q for equity (TQEQ) can be found in table 6.
Origin of law variable
Another variable that we use in this research is the origin of law of the country. This can be either common or civil law. Common law countries are associated with stronger investor protection than civil law countries (La Porta et al, 2000). Therefore, the origin of the law system can be used in order to make assumptions about the level of investor protection in the given country. Additionally, the origin of law can say something about the ownership structure of companies in that country. Common law countries are associated with more dispersed ownership whereas civil law countries are associated with more concentrated ownership (Wang and Shailer, 2015). This is also true for our sample, as shown in table 2 the ownership concentration average is significantly lower in common law countries than in civil law countries. The legal origin variable is a categorical variable that can either be common law or civil law. Therefore, a dummy variable will be created for this variable. This dummy variable will be 0 if the company is from a common law country (strong investor protection) and 1 if the company is from a civil law country (weaker investor protection).
Control variables
as it can show significant growth opportunities (Konijn, Kräussl and Lucas, 2011) and therefore it should be controlled for. Moreover, Wang and Shailer (2015) find that studies that include capital expenditure and capital intensity, find a lower relation between ownership concentration and firm performance. Therefore, we conclude that these variables have an effect on the relationship should be accounted for in the model. Additionally, we want to account for the potential heterogeneity between countries, therefore we add country dummies to the analysis in the robustness test section.
Symbol Variable name Definition Source Calculation
Own Ownership
concentration
The percentage of shares in the hands of the largest shareholder of the company. Using percentages
-9009
Own2 The square of
ownership concentration
Own2
MVE Market value of
equity Market capitalisation ASTK_MARKET_CAP
BVE Book value of
equity Total Shareholders equity (all stock) 314041 TQEQ Tobins Q equity Market to book ratio
of equity MVE/BVE
BVD Book value of debt Total liabilities and debt
314022
MVD Market value of
debt
Total liabilities and debt
BVD
TQEN Tobins Q
enterprise
market to book ratio of enterprise
(MVE+MVD)/(B VE+BVD) Law Legal origin Dummy: 1 = civil law
Law Own2_Law Interaction variable Squared ownership concentration and Law Own2 x Law
Size Firm size Log of total assets of
the firm 313077 log(total assets)
Age Firm age Firm age DATEINC 2019-(DATEINC)
Lev Leverage Ratio of total debt and
total assets Total debt: 314022 Total assets: 313077
Total debt/Total assets
CapInt Capital Intensity Ratio of total assets and operating revenue
Operating revenue: OPRE TotalAssets: 313077 Total assets/operating revenue CapExp Capital expenditure Expenditure of property plant and equipment as ratio of total assets PPE: 313068 Depreciation: 313019 TotalAssets: 313077 (PPE2018-PPE2017+Depreci ation)/TotalAssets
CountryDummy Home country Home country company
N-2 dummies for countries
CTRYISO
Table 6: The name, definition, source and calculation of the variables used in this thesis. All data is collected from Orbis.
3.3. Methods
concentration and firm performance use Ordinary Least Squares as estimation method. The model below shows the expected relationship. The definitions of the symbols that are being used in the models can be found in table 6 about variable construction. The i means that the variable is
measured at firm-level and the k means that the variable is measured at a country level.
(1)!Model:
PERF = ß0 + ß1 Owni+ ß2(Owni)² + ß3 Lawk + ß4 Owni*Lawk + ß5 (Owni)2*Lawk + ß6 Agei
+ ß7 Sizei + ß8Levi + ß9 Capexi + ß10 Capinti + ß11 CountryDummyk + eik
Model 1 tests the effect of ownership concentration (measured as the percentage of shares held by the top 1 largest shareholder) on firm performance as a multivariate relationship including the control variables, the squared term of ownership concentration, the law variable and the interaction variables. The interaction variables of the law times the ownership concentration measures whether there is a significant difference in the relation between common and civil law countries.
Figure 1: The moderating effect of origin of law on the relation between ownership concentration and firm performance
Figure 1 shows the moderation effect that is being researched in this thesis. For the firm performance we have two measures: Tobin’s Q for enterprise and Tobin’s Q for equity (Robustness test). For ownership concentration we use the measure of percentage of shares in the hands of the largest shareholder. The moderator variable for legal origin will be either classified as common law or civil law. We acknowledge that, based on our literature review, there is also a relation between the origin of law and the ownership concentration as we have seen that ownership is more concentrated in civil law countries compared to common law countries (Wang and Shailer, 2015). However, this relation goes beyond the scope of this research and therefore we will not research it in this thesis, but acknowledge that this is an important subject for future research.
4. Results
In this section the results of the analysis are shown. Next to the main analysis, also the robustness test is shown in this part of the thesis.
4.1. Results
the sample in order to decrease the firm heterogeneity within our sample. Financial firms are prone to other regulations and can therefore influence the sample (El Ghoul et al, 2017)). Additionally, utility firms are excluded as their firm performance can be influenced by government regulation (Claessens et al, 2002). The control variables are carefully selected and are based on previous research in order to minimize the endogeneity and omitted variable bias (Konijn, Kräussl and Lucas, 2011). However, it is important to note that we cannot rule out the existence of an endogeneity or omitted variable bias completely. And therefore the results of this research should be observed carefully.
Descriptive statistics
case for companies with minority shareholders as well as companies without minority shareholders. The latter is dependent on debt financing which, next to the raise of cost of capital also raises the monitoring from banks to the company.
Variable No. of observations Minimum Maximum Mean Median SD
Own 2056 0.0001 1 0.371 0.31 0.229 Own2 2056 0.00000001 1 0.191 0.097 0.213 TQEN 2056 0.66 4.04 1.54 1.2 0.88 TQEQ 2056 0.36 7 2.15 1.5 1.79 Civil Law 2056 0 1 0.8 1 0.4 Size 2056 8.8 16.2 12.21 12.03 2.11 Age 2056 8 145 46.67 31 38.8 Lev 2056 0.14 0.82 0.51 0.52 0.19 CapInt 2056 0.47 12.7 2.15 1.16 2.9 CapExp 2056 -0.006 0.13 0.035 0.021 0.038 CountryDummy 2056 1 17 9.9 12 5.6 Own_Law 2056 0 0.97 0.322 0.28 0.26 Own2_Law 2056 0 1 0.173 0.079 0.214
Table 7: Descriptive statistics of the year 2018. The table shows the average, minimum, maximum, mean, median and standard deviation of the variables used in this thesis. The construction of these variables can be seen in table 6.
Correlation matrix
any potential multicollinearity as multicollinearity is a violation of one of the assumptions of OLS. This is important as OLS will be used as estimation method in the analysis (Brooks, 2014). Additionally, multicollinearity can cause inappropriate conclusions of significance tests which can make it hard to base conclusions on the findings (Brooks, 2014). Two types of correlation analysis are used: Pearson correlation and Spearman correlation. Contrary to the Pearson correlation, Spearman correlation takes into account any potential outliers. As the spearman correlation ranks the variables and can therefore say something about the correlation between the variables without being misled by any possible outliers. For both correlations matrix, Tobin's Q for the enterprise will be used as the dependent variable as this will be the main performance measure used for the regression analysis.
TQEN Own Law Age18 Size18 Lev18 Capint18 Capex18 TQEN 1.000 Own -0.112* 1.000 Law -0.036 0.284* 1.000 Age -0.147* 0.155* 0.196* 1.000 Size -0.131* 0.042 0.084* 0.319* 1.000 Lev -0.195* 0.067* 0.168* 0.118* 0.286* 1.000 Capint 0.002 -0.028 -0.074* -0.137* -0.125* -0.324* 1.000 Capex 0.015 0.001 -0.049* 0.090* 0.186* 0.046* -0.099* 1.000
Table 8: Pearson correlation matrix of the variables used in the analysis to find whether there is any potential multicollinearity. This correlation matrix shows the correlation between the variables including the Tobin’s Q of the enterprise as performance measure. The data used is data from the financial year of 2018. The table also shows the significance level, *= significance at a 5% interval level.
mentioned, the threshold used is > 0.8 to suspect extreme multicollinearity (Grewal et al, 2004). And > 0.5 is the threshold to suspect any multicollinearity (Grewal et al, 2004). As shown in the Pearson correlation matrix below, there is no correlation found above the 0.5 threshold. Therefore, we can conclude that there is no multicollinearity found amongst the variables. The interaction variables are not included as they would highly correlate with each other as they are based on other variables in the correlation matrix. From the correlation matrix, we see that all variables are correlated with TQEN at least at the minimum significant level of 5%.
TQEN Own Law Age Size Lev Capint Capex
TQEN 1.000 Own -0.136* 1.000 Law -0.025 0.306* 1.000 Age -0.145* 0.195* 0.265* 1.000 Size -0.054* 0.046* 0.089* 0.317* 1.000 Lev -0.084* 0.081* 0.168* 0.139* 0.273* 1.000 Capint -0.116* -0.042 -0.063* -0.164* 0.036 -0.259* 1.000 Capex 0.084* 0.025 -0.012 0.134* 0.270* 0.067* -0.13* 1.000
Table 9: Spearman rank correlation matrix as an additional check for any multicollinearity issues. The correlation value is shown between brackets is the p-value shown.
no multicollinearity between the variables as no values are above the threshold. The same data is used in the Spearman rank correlation matrix for the Pearson correlation matrix.
Multivariate regression model
Variables Model (1) Model (2) Model (3) Model (4) Own -0.318*** (0.000) -0.363*** (0.000) -0.992*** (0.002) -0.531 (0.476) Own2 0.694** (0.037) 0.0141 (0.987) Civil Law 0.102** (0.040) 0.115** (0.022) 0.177 (0.200) Own_Law -0.539 (0.511) Own2_Law 0.7802 (0.404) Age -0.002*** (0.000) -0.003*** (0.000) -0.003*** (0.000) -0.003*** (0.000) Size -0.023** (0.019) -0.023** (0.019) -0.024** (0.016) -0.023** (0.020) Leverage -0.867*** (0.000) -0.895*** (0.000) -0.888*** (0.000) -0.891*** (0.000) Capex 0.824 (0.105) 0.901* (0.078) 0.925* (0.070 0.938* (0.066) Capint -0.025*** (0.000) -0.024 (0.001) -0.025*** (0.000) -0.026 (0.000) Cons 2.517*** (0.000) 2.471*** (0.000) 2.57*** (0.000) 2.513*** (0.000) R- Square 0.0695 0.0714 0.0734 0.0738 Adjusted R-Square 0.0668 0.0682 0.0697 0.0692 # of observations 2056 2056 2056 2056
Table 10: Regression analysis using Tobin’s Q of enterprise as firm performance measures. The data used in this analysis is from 2018. The significance level is also shown in the table: ***= significant at a 99% confidence level, **= significant at a 95% confidence level, *= significant at a 90% confidence level
our analysis. Figure 2, shows the curve of a U-shaped relationship as we find in model (3). It is clear to see that there is a u shape effect in the relationship between ownership concentration and firm performance. For which the negative effect is greater than the positive effect. Because as the graph finds the lowest point it only has a little positive effect in the end. Calculating the optimal point in the graph will be in this case the ‘turning point’ from which ownership concentration percentage the effect becomes positive again. This point is calculated by dividing the coefficient of the squared term of ownership concentration by the ownership concentration coefficient. This results in a ‘turning point’ of 70% (0.694/0.992) for ownership concentration in civil law countries. Which means that until the point of 70% of shares in the hands of the largest shareholder the relation between ownership concentration and firm performance is negative, after this threshold the relation becomes positive.
Figure 2: The U shape relation between ownership concentration and firm performance. Where ownership concentration is on the x-axis and firm performance on the y-axis.
4.2. Robustness test
In order to test the robustness of the analysis. We conduct the same analysis with another proxy for firm performance. In this robustness test the firm performance measure of Tobin’s Q of enterprise is substituted by Tobin's Q for a company’s equity. The construction of this measure can be seen in table 6. Similar to the tests in the result section, the Pearson and Spearman correlation matrix are also shown for these models, these can be found in Appendix A. The correlation matrices do not show any suspects for multicollinearity in the models. The same models are tested as in the regression analysis using Tobin's Q for enterprise. The only change is that leverage is not included in the models as this variable did not give a significant value in any of the models.
Variables Model (1) Model (2) Model (3) Model (4) Own -0.641*** (0.000) -0.717*** (0.000) -2.0816*** (0.001) -0.3655 (0.813) Own2 1.506** (0.029) -0.604 (0.734) Law 0.167 (0.103) 0.197* (0.057) 0.483* (0.091) Own_Law -2.0614 (0.226) Own2_Law 2.488 (0.199) Age -0.005*** (0.000) -0.006*** (0.000) -0.006 (0.000) -0.006*** (0.000) Size -0.051*** (0.010) -0.0518*** (0.009) -0.053*** (0.007) -0.051*** (0.010) Leverage Capex 1.809* (0.086) 1.94* (0.067) 1.992* (0.060) 2.023* (0.056) Capint -0.042*** (0.002) -0.0412*** (0.003) -0.0437*** (0.002) -0.044*** (0.002) Cons 3.292*** (0.000) 3.205*** (0.000) 3.424*** (0.000) 3.17*** (0.000) R- Square 0.0327 0.0340 0.0362 0.0370 Adjusted R-Square 0.0304 0.0312 0.0329 0.0328 # of observations 2056 2056 2056 2056
Table 11: Regression analysis using Tobin’s Q of equity as firm performance measures. The significance level is also shown in the table: ***= significant at a 99% confidence level, **= significant at a 95% confidence level, *= significant at a 90% confidence level
country influences the results. Table 12 shows the multivariate regression analysis including country dummies, and as can be seen from the table the results do not differ greatly. The same effects are observed: (I) a U-shape relation between ownership concentration and firm performance and (II) no significant effect of the origin of law on this relationship.
(0.987) (0.842) (0.894) (0.914) DE 0.195** (0.019) 0.21** (0.014) 0.204** (0.018) 0.204** (0.018) LU -0.157 (0.460) -0.14 (0.513) -0.1435 (0.503) -0.144 (0.502) CH 0.3765*** (0.001) 0.393*** (0.000) 0.39*** (0.000) 0.398 (0.000) AT 0.014 (0.934) 0.03 (0.868) 0.032 (0.857) 0.03 (0.865) ES 0.037 (0.781) 0.054 (0.689) 0.034 (0.800) 0.042 (0.757) PT -0.249 (0.186) -0.2329 (0.219) -0.235 (0.215) -0.232 (0.220) GR -0.417*** (0.000) -0.402*** (0.000) -0.401*** (0.000) -0.399*** (0.000) SE 0.448 (0.000) 0.465*** (0.000) 0.459*** (0.000) 0.468 (0.000)*** FI 0.060 (0.575) 0.0768 (0.486) 0.061*** (0.579) 0.071 (0.521) NO 0.067 (0.580) 0.084 (0.499) 0.082 (0.508) 0.088 (0.476) DK 0.495*** (0.000) 0.511*** (0.000) 0.496*** (0.000) 0.502*** (0.000) Cons 2.286*** (0.000) 2.41*** (0.000) 2.49*** (0.000) 2.502*** (0.000) R- Square 0.1226 0.1228 0.1240 0.1250 Adjusted R-Square 0.1136 0.1133 0.1141 0.1142 # of observations 2056 2056 2056 2056
Table 12: Regression analysis using Tobin’s Q of enterprise as firm performance measures. The
5. Discussion and conclusion
In this thesis, the relationship between ownership concentration and firm performance and the effect of the origin of law on this relation was researched. In the analysis, a sample of 2056 companies excluding financial and utility companies in the year of 2018 were studied. The findings showed a U-shaped relationship between ownership concentration and firm performance. Hence, indicating that as ownership concentration is increasing, firm performance is firstly decreasing, later increasing. This is similar to the findings of Hu et al (2008) when studying the relation between ownership concentration and firm performance in Japan. Hu et al (2008) ascribe the U shaped relation to the existence of expropriation effect when there is low ownership concentration and monitoring effect when there exist higher levels of ownership concentration. Whereas, expropriation effect decreases firm performance because of the expropriation of minority shareholders and the monitoring effect increasing firm performance as the concentrated owners can effectively monitor the managers (Hu et al, 2008). Additionally, these results show that highly dispersed and highly concentrated ownership concentration can both be beneficial to firm performance (Hu et al, 2008). Moreover, the U-shaped relation suggests diversifying corporate governance trends, and that no one corporate structure fits all companies (Hu et al, 2008). Concerning the first hypothesis and Lepore et al (2017) we see that, firstly, ownership concentration has a negative effect on firm performance. However, up until a certain point, which is 70% in our sample, the effect of ownership concentration on firm performance becomes positive. Resulting in a U-shaped relationship. Therefore, the first hypothesis cannot be accepted, as our analysis shows a U shape relation and not only a negative relation.
companies from civil law countries of Western Europe and Japan are both characterized by concentrated ownership.
type of investor (Ding, Zhang, Zhang, 2007). Therefore, we suggest that future research focuses on the type of investor and firm and the effect of these types on the relation between ownership concentration and firm performance.
This research adds to previous research by including a sample of Western European countries, whereas most previous research is focused on a sample of companies from the United States. This thesis includes several western European countries answering to the demand for using an expanded Western European sample and including common law countries (Lepore et al, 2017), using an international sample of developed countries (Wang and Shailer, 2015) and for more comparative country studies on this subject (Aguilera et al, 2016)). Moreover, most previous research is focused on inside ownership (Hu et al, 2008), whereas this research focuses on all ownership. Additionally, adding the squared term of ownership concentration adds to the previous research by investigating the existence of a nonlinear relationship between ownership concentration and firm performance which is formerly mostly done on national samples as discussed before. It also adds to the understanding of ownership structures and corporate governance mechanisms in common and civil law countries (Lepore et al, 2017).
should include the relation between the origin of law and ownership concentration in their analysis. This thesis acknowledged that there exist a potential relation. However, including this relation goes beyond the scope of this thesis.
6. References
Aguilera, R. V., & Crespi-Cladera, R. (2016). Global corporate governance: On the relevance of firms’ ownership structure. Journal of World Business, 51(1), 50-57.
Anderson, A., & Gupta, P. P. (2009). A cross-country comparison of corporate governance and firm performance: do financial structure and the legal system matter?. Journal of Contemporary Accounting & Economics, 5(2), 61-79.
Berle, A. A. and Means, G. C. (1932) The ModernCorporation and Private Property. New York:Harcourt, Brace and World.
Brooks, C. (2019). Introductory econometrics for finance. Cambridge university press.
Claessens, S., & Djankov, S. (1999). Ownership concentration and corporate performance in the Czech Republic. Journal of comparative economics, 27(3), 498-513.
Claessens, S., Djankov, S., Fan, J. P., & Lang, L. H. (2002). Disentangling the incentive and entrenchment effects of large shareholdings. The journal of finance, 57(6), 2741-2771.
Ding, Y., Zhang, H., & Zhang, J. (2007). Private vs state ownership and earnings management: evidence from Chinese listed companies. Corporate Governance: An International Review, 15(2), 223-238
El Ghoul, S., Guedhami, O., & Kim, Y. (2017). Country-level institutions, firm value, and the role of corporate social responsibility initiatives. Journal of International Business Studies, 48(3), 360-385.
Faccio, M., & Lang, L. (2002). The ultimate ownership of western european corporations. Journal of Financial Economics, 65(3), 365-395. doi:10.1016/S0304-405X(02)00146-0
Grewal, R., Cote, J. A., & Baumgartner, H. (2004). Multicollinearity and measurement error in structural equation models: Implications for theory testing. Marketing science, 23(4), 519-529.
Hu, Yabei, and Shigemi Izumida. (2008). Ownership concentration and corporate performance: A causal analysis with Japanese panel data. Corporate Governance: An International Review 16(4), 342-358.
Konijn, S. J., Kräussl, R., & Lucas, A. (2011). Blockholder dispersion and firm value. Journal of Corporate Finance, 17(5), 1330-1339.
La Porta, R., Lopez-de-Silanes, F., Shleifer, A., & Vishny, R. (2000). Investor protection and corporate governance. Journal of financial economics, 58(1-2), 3-27.
Lepore, L., Paolone, F., Pisano, S., & Alvino, F. (2017). A cross-country comparison of the relationship between ownership concentration and firm performance: does judicial system efficiency matter?. Corporate Governance: The international journal of business in society, 17(2), 321-340.
Li, J., & Qian, C. (2013). Principal! principal conflicts under weak institutions: A study of corporate takeovers in China. Strategic Management Journal, 34(4), 498-508.
Liu, W., Yang, H., & Zhang, G. (2012). Does family business excel in firm performance? An institution-based view. Asia Pacific Journal of Management, 29(4), 965-987.
doi:10.1007/s10490-010-9216-6
Maury, B. (2006). Family ownership and firm performance: Empirical evidence from Western European corporations. Journal of corporate finance, 12(2), 321-341.
McConnell, J. J., & Servaes, H. (1990). Additional evidence on equity ownership and corporate value. Journal of Financial economics, 27(2), 595-612.
Morck, R., Nakamura, M., & Shivdasani, A. (2000). Banks, ownership structure, and firm value in Japan. The Journal of Business, 73(4), 539-567.
Morck, Randall, Andrei Shleifer, and Robert Vishny, (1988). Management ownership and market valuation: An empirical analysis, Journal of Financial Economics 20, 293-315.
Porta, R. L., Lopez-de-Silanes, F., Shleifer, A., & Vishny, R. W. (1998). Law and finance. Journal of political economy, 106(6), 1113-1155.
Thomsen, S., Pedersen, T., & Kvist, H. K. (2006). Blockholder ownership: Effects on firm value in market and control based governance systems. Journal of Corporate finance, 12(2), 246-269.
Villalonga, B. (2019). on ownership structure and corporate performance: Looking back and looking forward. Journal of Corporate Finance, 58, 64-67.
Wang, K., & Shailer, G. (2015). Ownership concentration and firm performance in emerging markets: a meta! analysis. Journal of Economic Surveys, 29(2), 199-229.
Xu, X., & Wang, Y. (1999). Ownership structure and corporate governance in Chinese stock companies. China economic review, 10(1), 75-98.
7. Appendix A: Pearson correlation matrix including TQEQ
TQEQ Own Law Age Size Lev Capint Capex
TQEQ 1.000 Own -0.101* 1.000 Law -0.014 0.284* 1.000 Age -0.134* 0.155* 0.196* 1.000 Size -0.084* 0.042 0.084* 0.319* 1.000 Lev 0.020 0.067* 0.168* 0.118* 0.286* 1.000 Capint -0.046* -0.028 -0.074* -0.137* -0.125* -0.324* 1.000 Capex 0.023 0.001 -0.049* 0.090* 0.186* 0.046* -0.099* 1.000
8. Appendix B: Spearman rank correlation matrix including TQEQ
TQEQ Own Law Age Size Lev Capint Capex
TQEQ 1.000 Own -0.128 (0.000) 1.000 Law -0.017 (0.440) 0.306 (0.000) 1.000 Age -0.140 (0.000) 0.195 (0.000) 0.265 (0.000) 1.000 Size -0.0350 (0.1126) 0.046 (0.0391) 0.089 (0.0001) 0.317 (0.000) 1.000 Lev 0.029 (0.188) 0.081 (0.002) 0.168 (0.000) 0.139 (0.000) 0.273 (0.000) 1.000 Capint -0.142 (0.000) -0.042 (0.058) -0.063 (0.004) -0.164 (0.000) 0.036 (0.1042) -0.259 (0.000) 1.000 Capex 0.086 (0.001) 0.025 (0.2608) -0.012 (0.5812) 0.134 (0.000) 0.270 (0.000) 0.067 (0.002) -0.13 (0.000) 1.000
9. Appendix C: Descriptive statistics of the country dummies
Variable No. of observations Minimum Maximum Mean Median SD
GB 2056 0 1 0.193 0 0.395 NL 2056 0 1 0.253 0 0.157 BE 2056 0 1 0.268 0 0 .161 FR 2056 0 1 0.155 0 0.362 DE 2056 0 1 0.126 0 0.332 LU 2056 0 1 0.008 0 0.091 CH 2056 0 1 0.048 0 0.214 AT 2056 0 1 0.013 0 0.112 ES 2056 0 1 0.026 0 0.159 PT 2056 0 1 0.011 0 0.103 GR 2056 0 1 0.048 0 0.214 SE 2056 0 1 0.131 0 0.337 FI 2056 0 1 0.047 0 0.212 NO 2056 0 1 0.032 0 0.175 DK 2056 0 1 .028 0 0.16
