Ownership structures on the European continent
- analyzing and explaining differences -
University of Groningen
Faculty of Management and Organization
MSC International Business & Management
July 2007
By
Erik Thomas Beverlo
S1576402
Preface
With this Master thesis I finish my study International Business and Management at the University of Groningen. Inspired by the work of LaPorta et al. (1999), this thesis analyzes the ownership structures of large European listed and non listed companies and explains why country differences exist. The research was conducted from February until July 2007.
During the study International Business and Management I became fascinated in the ownership structure literature because of its impact on corporate governance, politics and economy. As traditional ownership structure literature have mostly focused on wealthy nations, I became more and more interested in unravelling the ownership structures of companies in less wealthy nations, most notably the East European countries. As one of the first studies that compares ownership structures in Europe with an extensive dataset, it contributes substantially to the existing literature. I hope that in the near future more studies on this topic will be dedicated to the East European countries and other emerging nations.
I am very grateful to my supervisor dr. Niels Hermes for his valuable comments and insights during the research process. His advice and critical comments as well as our interesting discussions were an important contributor to the quality of my thesis. Furthermore I would like to thank my family, girlfriend and my friends for their mental support and understanding during the last months of my study.
Erik Thomas Beverlo
July
Ownership structures on the European continent
- analyzing and explaining differences - Abstract
By using an extensive dataset of 6,798 listed and non listed companies, this study analyzes ownership structures in Europe. Differences in ownership structures regarding shareholder concentration, the share of the ultimate owner, the type of the ultimate owner and pyramidal ownership were found between East and West Europe and between the individual countries. This study found that the state of the legal infrastructure determines for a large part the ownership structure differences between the countries. Listed firms tend to be owned more dispersed in countries with well developed legal infrastructures, whereas non listed firms tend to be owned more concentrated in well developed legal infrastructure countries. Furthermore, this study evidences that the effectiveness of laws lags behind on the extensiveness of laws in the East European countries which results in a camouflaged ownership structure picture for listed firms. Overall this study contributes significantly to the ownership structure literature by adding new valuable insights.
TABLE OF CONTENTS
1. Introduction 7
2. Review on previous empirical research 10
2.1 Shareholder concentration 10
2.2 Share of the ultimate owner 16
2.3 Type of the ultimate owner 18
2.4 Pyramidal ownership 22
3. Theory and hypotheses 24
3.1 Legal context 24
3.2 Firm specific variables 30
4. Data and descriptive analyses 32
4.1 Methodological issues of previous studies 32
4.2 Database 33
4.3 Ownership types and levels 34
4.4 Description of the ownership data 37
4.5 Shareholder concentration 37
4.6 Share of the ultimate owner 42
4.7 Type of the ultimate owner 44
4.8 Pyramidal ownership structures 55
5. Empirical methodology 60 5.1 Dependent variables 60 5.2 Independent variables 61 5.3 Multicollinearity 63 5.4 Regression models 64 6. Empirical results 66 6.1 Europe 66
6.2 West Europe and East Europe 72
6.3 West European listed firms 72
6.4 East European listed firms 73
6.5 West European non listed firms 74
6.6 East European non listed firms 76
6.7 Law Families 77
7. Discussion 78
7.1 Ownership concentration and pyramidal ownership 78
7.2 Type of the ultimate owner 82
8. Conclusion 84
Literature list 86
Appendix I 89
Appendix II 90
LIST OF TABLES AND FIGURES Figures
Figure 3.1: Direct, indirect and pyramidal ownership 28
Figure 4.1: Direct ownership 34
Figure 4.2: Total/indirect ownership 35
Figure 4.3: Pyramidal ownership 35
Figure 4.4: Pyramidal ownership 36
Figure 4.5: Carrefour; widely held ownership 45
Figure 4.6: Romaqua Group SA; Family ownership via pyramidal structure 45 Figure 4.7: ABB oil & gas Europe BV; Industrial company ownership 46
Figure 4.8: Patria OYJ; State ownership 47
Figure 4.9: Bashinformsvyaz; Financial company ownership 47 Figure 4.10: Avio SPA; Miscellaneous ownership 48
Tables
Table 2.1: Shareholder concentration in West Europe 11
Table 2.2: Percentage of widely held listed firms found by LaPorta et al. (1999) and Faccio
and Lang (2001) 13
Table 2.3: Shareholder concentration in East Europe 15
Table 2.4: Share of the ultimate owner as found by van der Elst (2004) and Faccio
and Lang (2001) 17
Table 2.5: Share of the ultimate owner; Pajuste 2002 17
Table 2.6: Type of the ultimate owner; LaPorta et al. (1999) 19 Table 2.7: Type of the ultimate owner; Faccio and Lang (2001) 20 Table 2.8: Pyramidal ownership; LaPorta et al. (1999) and Faccio and Lang (2001) 22
Table 3.1: ADR, rule of law and legal origin 26
Table 4.1: Dataset; number of listed and non listed companies per country 34
Table 4.2 Shareholder concentration in Europe 38
Table 4.3: Independent samples T-test; difference East and West Europe regarding shareholder
concentration 40
Table 4.4: Independent samples T-test; difference East and West Europe regarding the number of
shareholders 41
Table 4.5: Average share of the ultimate owner 43
Table 4.6: Ultimate ownership in West European non listed firms 49 Table 4.7: Ultimate ownership in East European non listed firms 50
Table 4.8: High vs. low ADR; non listed firms 51
Table 4.9: Ultimate ownership sub regions; non listed firms 52 Table 4.10: Ultimate ownership in West European listed firms 52 Table 4.11: Ultimate ownership in East European listed firms 53
Table 4.12: High vs. low ADR; listed firms 54
Table 4.13: Ultimate ownership sub regions; listed firms 55
Table 4.14: Pyramidal structures; West Europe 55
Table 4.15: Pyramidal structures; East Europe 56
Table 4.16: High vs. low ADR; Pyramidal ownership non listed and listed firms 57
Table 4.17: Pyramidal structures; sub regions 58
Table 6.13: Determinants of widely held ownership in listed firms; East Europe 74 Table 6.14: Determinants of the ultimate owner share; West Europe 75 Table 6.15: Determinants of widely held ownership in non listed firms; West Europe 75 Table 6.16: Determinants of the ultimate owner share in non listed firms; East Europe 76 Table 6.17: Determinants of widely held ownership in non listed firms; East Europe 76
Table 6.18: Summary results for Europe 77
Appendices
Table I.1: Descriptive statistic for the number of employees, number of subsidiaries and operational
revenue/ turnover; listed firms 89
Table I.2: Descriptive statistics for industry; listed firms 89 Table I.3: Descriptive statistic for the number of employees, number of subsidiaries and operational revenue/ turnover; non listed firms 89 Table I.4: Descriptive statistics for industry; non listed firms 89 Table II.:1 Correlation table independent and control variable(s) 90 Table III.1: Determinants of the ultimate owner share in listed firms; German legal origin 91 Table III.2: Determinants of widely held ownership in listed firms; German legal origin 91 Table III.3: Determinants of the ultimate owner share in listed firms; French legal origin 92 Table III.4: Determinants of widely held ownership in listed firms; French legal origin 92 Table III.5: Determinants of the ultimate owner share in listed firms; Scandinavian legal origin 93 Table III.6: Determinants of widely held ownership in listed firms; Scandinavian legal origin 93 Table III.7: Determinants of the ultimate owner share in listed firms; English legal origin 94 Table III.8: Determinants of widely held ownership in listed firms; English legal origin 94 Table III.9: Determinants of the ultimate owner share in non listed firms; German legal origin 94 Table III.10: Determinants of widely held ownership in non listed firms; German legal origin 95 Table III.11: Determinants of the ultimate owner share in non listed firms; French legal origin 96 Table III.12: Determinants of widely held ownership in non listed firms; French legal origin 96 Table III.13: Determinants of the share of the ultimate owner in non listed firms; Scandinavian
legal origin 97
Table III.14: Determinants of widely held ownership in non listed firms; Scandinavian legal
origin 97
1. INTRODUCTION
Corporate governance has been a popular subject since Berle and Means’ (1932) work, but gained most of its attention over the last two decades. The Organization for Economic Cooperation and Development (OECD, 2004) defines corporate governance as ‘a set of relationships between a companies’ management, its board, its shareholders and other stakeholders. It also provides the structure through which the objectives of the company are set, and the means of attaining those objectives and monitoring performance are determined (p.11)’. Corporate governance is thus a set of processes, customs, policies, laws and institutions by which a firm is directed, administered and controlled. This study focuses on the shareholder and control aspect of corporate governance in Europe. More specific, it focuses on differences in ownership structures of large listed and non listed companies on the European continent by using an extensive dataset. A substantial body of research in financial economic and strategic management literature links the type, pattern and amount of ownership to managerial behaviour and corporate performance (Gibbs, 1993; Hoskisson et al., 1994; Jensen and Warner, 1988). This indicates the necessity of knowledge of ownership structures in companies all over the world.
Most literature on ownership structures take the work of Berle and Means as starting point. In their book “The modern corporation and private property” (1932) they indicate that ownership structures in the United States (US) can be characterized as widely held, where control is concentrated in the hands of management and capital is dispersed among small investors. This image of the firm was soon labelled as the ‘Berle and Means’ firm. Contemporary studies to companies in the European countries and other wealthy nations show a more concentrated ownership structure compared to the US Berle and Means firm. (a.o. LaPorta et al., 1999; Faccio and Lang, 2001; Edwards and Fischer, 1994; Barca, 1995). Moreover, these studies indicated that the degree of ownership concentration differs between countries. Apart from shareholder concentration, also the type of the owners seem not to be homogeneous across nations (a.o. LaPorta et al. 1999; Claessens and Tzioumis, 2006; Faccio and Lang, 2001)
on the other hand, is argued to be characterized by narrow capital markets, large block holders (concentrated ownership) and cross-shareholdings between firms. Obviously, differences between US and European ownership structures exist, as also indicated above. But the view that an entire continent with many countries can be characterized into one model regarding ownership structures is questioned in this study. It is argued that we cannot use this division, as the differences in ownership structures have not been studied extensively and, maybe, understood well within Europe. The foregoing indicates a gap in the literature. This gap can be defined as the lack of study into ownership structure differences within Europe and especially in East Europe. Previous studies focussed mostly on West European companies (Faccio and Lang, 2001; van der Elst, 2004; La Porta et al., 1999). Claessens and Tzioumis (2006) did study (next to West European companies) ownership structures in East European companies, but used a limited dataset. Moreover, a good comparison with West European companies and an analysis of the findings was missing. Pajuste (2002) reported the ultimate ownership share of companies in nine East European countries, but did not document the type of the ultimate owner. Overall, studies into European ownership structures that include a substantial number of both East and West European companies seem to be missing.
It is important to study differences in ownership structures on the European continent in order to really be able to compare the corporate governance models of Europe with other regions and nations on the globe. The objective of this thesis is, thus, to analyze ownership structures within Europe. Furthermore, this study aims to explain the differences by using a set of external (e.g. legal rules) and internal (e.g. company specific) variables. This extensive study contributes to the existing literature in a number of ways. For the first time a large set of European listed and non listed companies is analyzed on different levels aiming to reveal ownership structure characteristics and differences. Furthermore, the effect of a set of explaining variables is analyzed for both listed and non listed firms on different levels which enables this study to explain ownership structure differences. As the focus of this study is on both wealthy and less wealthy nations, a new variable is used to better proxy the legal infrastructure in explaining ownership structure differences. Finally, due to a detailed comparison and analysis of European ownership structures this study contributes significantly to the understanding, but also adds new valuable insights. The main research question for this study is:
The ownership structure differences in Europe will be studied by analyzing shareholder concentration, the share of the ultimate owner, the type of the ultimate owner, and ownership via pyramidal structures. These analyses will be performed on three levels. The first is the regional level. Here the differences will be analyzed between East and West Europe. The second level is law family level. Here the focus is on differences between countries with different legal origins. The third level of interest are the country differences. It is expected to find differences in the before mentioned elements of ownership structures between East and West Europe and across the countries studied. This expectation is based on extensive studies to West European companies in which differences in ownership structures have been found (a.o. Faccio and Lang, 2001; van der Elst, 2004; La Porta et al., 1999). As indicated before, studies to East European ownership structures are limited. However the existing literature that is available for this region did found differences in East European ownership structures (e.g. Claessens and Tzioumis, 2006; Pajuste, 2002; Zhang and Beverlo, 2006). Subsequently, this study provides an analysis of ownership structures in Europe and assesses if differences exist between East and West Europe, countries with different law families and individual countries. Moreover, this study aims to explain the differences.
2. REVIEW ON PREVIOUS EMPIRICAL RESEARCH
In this section, the findings of earlier studies regarding ownership structure differences in Europe will be discussed. To analyze ownership structure differences this study focuses on shareholder concentration, the share of the ultimate owner, the type of the ultimate owner and pyramidal ownership. This section forms the basis for the second part of this study which tries to explain these ownership structure differences.
2.1 Shareholder Concentration
Demsetz and Lehn (1985) argue that the more concentrated the ownership of a company is, the greater the benefits and costs for the largest (ultimate) owner are. They argue that if a company has one owner, all costs and benefits are borne by this owner and no externalities influence the owner’s decisions about attending to the tasks of ownership. On the other hand, dispersed ownership leads to a greater divergence between costs and benefits for a typical owner because he is not alone. Therefore, as Demsetz and Lehn argue, an owner in a company with dispersed ownership would ‘respond by neglecting some tasks of ownership’ (p.1156). Hence, in agreement with Demsetz and Lehn, the degree of shareholder concentration indicates a trade off between the risk to investors of concentrated investments ( i.e. all costs are borne by the same owner) and the control potential of the firm. According to Renneboog (1996) dispersed ownership can therefore be seen as a method for risk reduction as investments can be diversified, but it reduces also the active involvement of the shareholders in the companies’ decision making processes.
mainly attribute to weak boards with ample formal power, insurmountable take over barriers and anti-blockholder regulations that prevent owners from exerting control. Becht and Mayer (2001) indicate that studies investigating control of companies outside the US are limited due to, for a long time, a taken for granted picture of the Berle and Means firm.
However, contemporary studies to European ownership structures emerged. Most of the research focussed on West European countries and found in general that shareholder concentration is much higher in Europe than in the US. These studies not only revealed that differences exist between the US and the European countries, but also within Europe. Van der Elst (2004) for example found that UK firms have smaller blockholders than Belgian, German, French, Italian and Spanish firms. As small blockholders come with more small shareholders, one can state that UK firms are typically more dispersed than the other five studied countries. Claessens and Tzioumis (2006) analyzed the ownership concentration of 370 listed and 5,152 non listed European firms. They included the non listed firms in their analysis because in many countries these firms account for most of the economic activity. They found that the size of the blockholders in European countries is higher for non listed firms than for listed firms. 84% of the non listed companies in their entire sample was reported to have a blockholder of over 50,01% of the shares. In comparison, they found less than 30% of the listed companies to have such a blockholder. Table 2.1 presents the findings of Claessens and Tzioumis for the West European countries in their study.
Table 2.1: Shareholder concentration in West Europe
non listed listed
country >50,01% >50,01% 25,01%-50% <25% Belgium 85,12 59,46 21,62 18,92 Denmark 80,19 29,69 34,38 35,94 France 94,97 41,10 32,19 26,71 Germany 84,27 41,11 27,78 31,11 Greece 77,07 44,59 35,14 20,27 Italy 81,04 65,17 14,61 20,22 Netherlands 89,92 17,86 20,24 61,90 Norway 79,46 15,79 47,37 36,84 Portugal 89,09 45,71 20,00 34,29 Spain 85,03 28,75 32,50 38,75 Sweden 79,91 11,11 37,78 51,11 Switzerland 62,71 47,37 17,54 35,09 UK 80,83 5,84 12,34 81,82
was found to have the highest number of non listed companies with a majority blockholder. Almost 95% of the French non listed firms were reported to have a shareholder with stake of over 50,01%. As the distribution of the number of non listed companies in West Europe is mainly found to be in the group of blockholders with a control stake of over 50,01%, it can be deduced from the findings of Claessens and Tzioumis that this region can be characterized by high shareholder concentration.
Focussing on the listed companies, Claessens and Tzioumis found more differences in shareholder concentration among the West European companies. As expected, for being an Anglo-American country, UK firms were in only 6% of the cases found to have a majority blockholder. 12 % of the UK firms were reported to have a blockholder between 25,01% and 50%, whereas 81 % of the firms have no majority blockholder. This finding would support the Berle and Means view of the firm, as it indicates a rather dispersed ownership structure. Apart from the UK, also Sweden (11%), Norway (16%) and the Netherlands (18%) were found to have a relatively limited number of listed companies with a blockholder of over 50,01%. Dutch and Swedish firms were found to have no large blockholder in 62% and 51% of the cases respectively. Norwegian companies were mainly characterized by a blockholder between 25,01% and 50,01%, in 47% of the cases. Although the UK stands far apart in terms of ownership dispersion from the other West European countries studied by Claessens and Tzioumis, also Sweden, the Netherlands and to a lesser extend Norway show relatively dispersed ownership when being compared to the other West European countries. Particularly Italian (65%) and Belgian (59%) firms were reported to have a shareholder with more than 50,01% of the stakes. The other West European countries mostly circle around 30%-50% of the companies having a shareholder of over 50,01%. Claessens and Tzioumis clearly showed in their study that the continental European firms can be characterized by concentrated ownership. Besides this finding they also indicated that differences in shareholder concentration between West European countries exist.
Table 2.2: Percentage of widely held listed firms found by LaPorta et al. (1999) and Faccio and Lang (2001)
country
LaPorta et al.
(20 obs. per country)
Faccio and Lang
(5,232 obs. in total) Austria 5,00% 11,11% Belgium 5,00% 20,00% Denmark 40,00% NA Finland 35,00% 28,68% France 60,00% 14,00% Germany 50,00% 10,37% Greece 10,00% NA Ireland 65,00% 62,32% Italy 20,00% 12,98% Netherlands 30,00% NA Norway 25,00% 36,77% Portugal 10,00% 21,84% Spain 35,00% 26,42% Sweden 25,00% 39,18% Switzerland 60,00% 27,57% UK 100,00% 63,08%
Source: LaPorta et al. (1999) and Faccio and Lang (2001); widely held firms defined as a firm having no shareholder with more than 20% of the shares.
As can be noted from table 2.2, both studies indicated differences in the number of widely held firms found per country. The findings, however, contradict each other in a number of ways. LaPorta et al. for example found many widely held corporations in France (60%), compared to only 14% found by Faccio and Lang. Furthermore LaPorta found higher numbers of widely held firms in Germany (50%) and Switzerland (60%), compared to 10% and 28% respectively of the findings by Faccio and Lang. These deviations might be resulting from the limited number of companies covered by LaPorta et al., as they only studied the 20 largest listed firms in a country in comparison to a far larger set of firms studied by Faccio and Lang. Comparing the results of Faccio and Lang with the Claessens and Tzioumis results, show less impressive differences. It must be noted however that the latter study used a different threshold, i.e. less than 25%. Finally, Zhang and Beverlo (2006) found that the average number of shareholders of companies differ across countries. Using the number of shareholders as a proxy for shareholder concentration, it can thus be argued that shareholder concentration differs across the West European countries. Examining the most important results of the before mentioned studies does not show a strict agreement in the findings.
That differences in ownership structures exist between East and West European countries seems plausible looking at the transition process the East European countries went through. This process initiated by the collapse of the communist block and the emergence of political and economical liberalization. For countries in transition, one of the most important policy aspects is the development of a financial system in order to achieve economic growth. Hermes and Lensink (2000) argue that stock markets have an important role in this, as stock markets only tend to supply finance to the most effective and profit generating projects. Many East European countries faced the emergence of a stock market with a process of privatization. According to Pajuste (2002), two types of privatization can be distinguished. The first one is the mandatory listing after mass privatization. This was primarily done in the Czech and Slovak Republics and Romania. The other type can be identified as a form of privatization which started with a small number of listed firms that increased over time. This was primarily done by Estonia, Hungary, Latvia, Slovenia and Poland. The effect of these both forms was different. The first type of privatization resulted in a high market capitalization from the outset, and then decreased to lower amounts due to delisting of (illiquid) firms. The second type of privatization resulted in a slow growth of the equity market capitalization. These major changes in the East European economic growth processes might have had an influence on the ownership concentration of East European listed firms. Also non listed firms might have different ownership structures compared to West European firms due to the transition process.
Table 2.3: Shareholder concentration in East Europe
non listed listed
country >50,01% >50,01% 25,01%-50% <25% Belarus NA NA NA NA Bulgaria NA NA NA NA Croatia 90,00 36,84 15,79 47,37 Czech Rep. 90,63 54,17 45,83 0,00 Hungary NA NA NA NA Poland 93,59 54,55 36,36 9,09 Romania 84,85 50,00 50,00 0,00 Russia 63,64 50,98 41,18 7,84 Serbia NA NA NA NA Slovak NA NA NA NA Slovenia NA NA NA NA Ukraine 66,90 NA NA NA
Source: Claessens and Tzioumis (2006), figures are the percentages of companies per country with a blockholder of over 50,01%, 25,01%-50% and less than 25%.
found in their sample of 7,684 East and West European firms, that the latter group of companies has relatively more shareholders than is the case in East European firms. The findings empirically supported the hypothesis that shareholder ownership tends to be more concentrated in East Europe and more dispersed in West Europe.
The foregoing showed that by no means shareholder concentration is similar within Europe. Based on the discussed studies one can thus argue that a difference in shareholder concentration exist within Europe and between East and West European countries.
2.2 Share of the Ultimate Owner
Table 2.4: Share of the ultimate owner as found by van der Elst (2004) and Faccio and Lang (2001)
country
van der Elst
1904 obs.
Faccio and Lang
4806 obs. Austria NA 53,53% Belgium 41,71% 40,09% Denmark NA NA Finland NA 37,43% France 51,98% 48,32% Germany 46,13% 54,50% Greece NA NA Ireland NA 21,55% Italy 48,14% 48,26% Netherlands NA NA Norway NA 31,47% Portugal NA 41,00% Spain 37,91% 44,24% Sweden NA 30,96% Switzerland NA 46,68% UK 18,26% 25,19%
Source: van der Elst (2004);Faccio and Lang (2001), average control rights share of the ultimate owner per country
Faccio and Lang found that the highest percentage of ultimate owner share in terms of control rights was in Germany (54,50%), followed by Austria (53,52%) and France (48,32%). The lowest ultimate owner share in terms of control rights was found in Ireland (21,55%) followed by the UK (25,13%), Sweden (30,96%) and Norway (31,47%). The total average share of the ultimate owner for the West European countries is 40,25%. When we compare tables 2.4 and 2.2, it can be observed that the countries with companies with higher average ultimate owner shares have also less widely held firms.
Focussing on East European companies, Pajuste (2002) reported the average ultimate owner shares of listed companies per country, based on the results of the ACE project (2001). The results are presented in table 2.5.
Table 2.5: Share of the ultimate owner; Pajuste 2002
highest average ultimate owner share in the findings of Pajuste, followed by Romania (53%). The lowest average share was found in Slovenia (27%). The total average share of the ultimate owner for the East European countries in the study of Pajuste was 46%. The problem however is that not every East European country is present in the study of Pajuste. This makes it difficult to compare the findings of Pajuste with the findings of the West European countries as documented by Faccio and Lang and van der Elst. Apart from the average share of the ultimate owner, Pajuste does not indicate shareholder concentration figures. Comparing the findings of Claessens and Tzioumis with the findings of Pajuste would suggest that there is a same relation between shareholder concentration and the average share of the ultimate owner as in West European countries, but there are only three countries to compare. The Czech firms for example, have on average an owner with 61% of the shares according to Pajuste. Claessens and Tzioumis indicate in their findings that the Czech Republic has no listed companies with an owner of less than 25% of the shares. Romania is characterized by the same relation of high ultimate owner share and concentrated ownership.
Two main points can be deduced from the previous discussion. Firstly, the average share of the ultimate owner varies per country and there tends to be a relation between the average share of the ultimate owner and shareholder concentration. Secondly, the average share of the ultimate owner tends to be the same or slightly higher for the East European firms, as found by Pajuste (2002), compared to the highest averages of the West European firms as found by Faccio and Lang (2001). The difficulty is however, that we cannot draw bold conclusions from these findings as previous studies failed to analyze the size of the ultimate owners for both East and West European countries in one study with one methodology.
2.3 Type of the Ultimate Owner
The previous sub sections revealed the first differences in ownership structure between East and West European countries by indicating differences in shareholder concentration and the average share of the ultimate owner. This sub section analyzes the question who the owners of the East and West European companies actually are. The main interest is, of course, whether there are country differences regarding the type of the ultimate owner within Europe and between the East and West European countries.
indicating the presence of widely held firms. They argue furthermore that in some countries state ownership is present. The study of Franks and Mayer was one of the first in its kind to explore the type of the ultimate owner in European firms. However, as the scope of their study was limited to only France, Germany and the UK, most of the contemporary literature for ultimate ownership differences take the work of LaPorta et al. (1999) as point of reference. LaPorta et al. extended the work of Franks and Mayer by analyzing the ultimate ownership types of the 20 largest listed companies in 27 wealthy economies of which 16 West European countries. LaPorta et al. found that 36% of the firms in the studied countries are widely held, 30% is family owned and 18% is state owned. The remaining 15% is equally controlled by widely held financial companies, widely held corporations and miscellaneous owners. Focussing on the European countries in their work, it can be observed that large differences exist. The findings of LaPorta et al. are presented in table 2.6.
Table 2.6: Type of the ultimate owner; LaPorta et al. (1999)
country Widely held Family State
Widely held Financial widely held corporation Miscellaneous Austria 5,00% 15,00% 70,00% 0,00% 0,00% 10,00% Belgium 5,00% 50,00% 5,00% 30,00% 0,00% 10,00% Denmark 40,00% 35,00% 15,00% 0,00% 0,00% 10,00% Finland 35,00% 10,00% 35,00% 5,00% 5,00% 10,00% France 60,00% 20,00% 15,00% 5,00% 0,00% 0,00% Germany 50,00% 10,00% 25,00% 15,00% 0,00% 0,00% Greece 10,00% 50,00% 30,00% 10,00% 0,00% 0,00% Ireland 65,00% 10,00% 0,00% 0,00% 10,00% 15,00% Italy 20,00% 15,00% 40,00% 5,00% 10,00% 10,00% Netherlands 30,00% 20,00% 5,00% 0,00% 10,00% 35,00% Norway 25,00% 25,00% 35,00% 5,00% 0,00% 10,00% Portugal 10,00% 45,00% 25,00% 15,00% 0,00% 5,00% Spain 35,00% 15,00% 30,00% 10,00% 10,00% 0,00% Sweden 25,00% 45,00% 10,00% 15,00% 0,00% 5,00% Switzerland 60,00% 30,00% 0,00% 5,00% 0,00% 5,00% UK 100,00% 0,00% 0,00% 0,00% 0,00% 0,00%
Faccio and Lang (2001) focussed only on firms in West European countries and also found differences in ultimate owner type between the companies. In an extensive sample of 5,232 listed companies they found that nearly 37% of the West European firms are widely held and 44% of the firms family controlled. The findings are presented in table 2.7.
Table 2.7: Type of the ultimate owner; Faccio and Lang (2001)
country Widely held Family State
widely held corporation Widely held Financial Miscellaneous Austria 11,11% 52,86% 15,32% 0,00% 8,59% 11,11% Belgium 20,00% 51,54% 2,31% 0,77% 12,69% 12,69% Denmark NA NA NA NA NA NA Finland 28,68% 48,84% 15,76% 1,55% 0,65% 4,52% France 14,00% 64,82% 5,11% 3,79% 11,37% 0,91% Germany 10,37% 64,62% 6,30% 3,65% 9,07% 3,37% Greece NA NA NA NA NA NA Ireland 62,32% 24,63% 1,45% 2,17% 4,35% 5,07% Italy 12,98% 59,61% 10,34% 2,88% 12,26% 1,20% Netherlands NA NA NA NA NA NA Norway 36,77% 38,55% 13,09% 0,32% 4,46% 4,54% Portugal 21,84% 60,34% 5,75% 0,57% 4,60% 6,90% Spain 26,42% 55,79% 4.11% 1,64% 11,51% 0,47% Sweden 39,18% 46,94% 4,90% 0,00% 2,86% 5,71% Switzerland 27,57% 48,13% 7,32% 1,09% 9,35% 6,31% UK 63,08% 23,68% 0,08% 0,76% 8,94% 3,46%
Source: Faccio and Lang (2001); figures do not sum up to 100 because the table ‘cross holdings’ was omitted.
In some countries Faccio and Lang documented a significant proportion of firms to be owned by the state or by widely held financial firms. They indicated furthermore that widely held corporations control only few firms. Only 4% of the entire sample of 5,232 firms was documented as having the state as the ultimate owner at the 20% threshold. Most state ownership was found in Austria (15,32%), Finland (15,67%), Italy (10,34%) and Norway (13,09%). Widely held firms were, again, mostly found in the UK (63,08%) and Ireland (62,32%). Few widely held firms were observed in Germany (10,37%), Italy (12,98%) and France (14%). Family ownership was documented as the largest ultimate owner type in West Europe mostly prevalent in France (64,82%), Germany (64,62%) and Portugal (60,34%).
Claessens and Tzioumis (2006) documented that most of the 299 large listed West European firms1 they studied, had a family (52,7%) as being the ultimate owner. The second largest type identified are industrial companies (30,9%). Claessens and Tzioumis did however not analyze the differences of the type of the ultimate owner on country level. As can be noted in the foregoing, Franks and Mayer (1995), LaPorta et al. (1999), Faccio and Lang (2001) and
1 These figures have been recalculated from the study of Claessens and Tzioumis as they divided Europe in four
Claessens and Tzioumis (2006) all indicate the role of families as the ultimate owner in West European nations as very important. Moreover the first three studies significantly identified the presence of widely held firms in West European countries. LaPorta et al. and Faccio and Lang however, seem not to agree with Franks and Mayer and Claessens and Tzioumis that other companies make up for another substantial group of European ultimate owners, as they found very few widely held corporations to be such an owner. Faccio and Lang’s results regarding state ownership, furthermore, seem to conflict with the findings of LaPorta who found a more substantial amount of state ownership. This indicates that, except for family and widely held ownership, there is no fully accepted picture of the type of the ultimate owner in Western Europe.
Traditionally, studies into ownership structures focused on listed firms. This study however also analyzes the ownership structures of non listed companies because of their importance in many national economies. It would not be accurate to assume that listed companies have the same ultimate ownership type pattern as non listed companies. For being non listed, it is assumed for example that few widely held firms are among the companies studied. Claessens and Tzioumis (2006) would agree on this as they found only 11% of the non listed firms they studied to have an ultimate owner with less than 25% of the votes, which is in this study regarded as widely held. Most notably not the UK (18,8%) but Switzerland was found to have the highest percentage of non listed widely held firms (28,81%). However, at least 80% of the companies in the other countries where indicated to have a shareholder over 25% and thus an ultimate owner. They found that 64,19% of the 4,935 West European non listed companies was owned by an industrial firm. 22,9% of the companies was owned by a family, 6,4% was state owned and the remaining 6,5% was owned by a financial firm or miscellaneous2. These results indicate that for non listed firms, families and industrial companies are the most important owners.
European countries. The countries they studied were Croatia, the Czech Republic, Poland, Romania, Russia and Ukraine. However, Claessens and Tzioumis did not analyze the type of the ultimate owner per country, but for the whole region. By studying companies with more than € 50 million operational revenue/ turnover, they were only able to analyze 188 non listed and 26 listed companies. Despite the low number of companies analyzed, the results of their study to East European companies were interesting. They found that 58,51 % of the non listed and 57,69 % of the listed firms were state owned. Industrial firms were indicated as being the second largest ultimate owner type in this region, in 20,1% of the non listed firms and 26,9% of the listed firms. Family ownership was only substantial in non listed firms (19,7%) as compared to listed firms (7,6%). Except for family owned firms, listed and non listed firms do not differ much according to Claessens and Tzioumis, regarding the type of the ultimate owner. These results indicate thus a completely different view compared to West European firms.
2.4 Pyramidal Ownership
The final element in analyzing ownership structure differences in this study is the ownership via pyramidal structures. When a company owns a subject company via another firm, the subject company is said to be controlled via a pyramidal structure. By using pyramidal structures, an owner can exert control in excess of its cash flow rights thereby influencing the ownership structure of a company. In the next section the theory of pyramidal ownership structures and its effect on companies will be explained in more detail. For now we concentrate on the presence of these structures in Europe. Table 2.8 summarizes the findings of LaPorta et al. (1999) and Faccio and Lang (2001) for West European countries.
Table 2.8: Pyramidal ownership; LaPorta et al. (1999) and Faccio and Lang (2001)
It can be noted from table 2.8 that different results have been found regarding pyramidal ownership in both studies. We see for example that no pyramids were observed in the UK according to LaPorta et al. Faccio and Lang however found 21,13% of the UK companies to be owned via a pyramidal structure. Swedish firms were found in 53% of the cases to be owned via a pyramid by LaPorta et al., whereas Faccio and Lang found only 15,91% of the Swedish firms to be owned by such a structure. Furthermore, different results were observed in a.o. Belgium, Germany, Portugal and Austria. These differences are attributable to the different sample sizes used in both studies. The interesting point from the studies of LaPorta et al. and Faccio and Lang, however, is that they showed that country differences exist within West Europe regarding ownership via pyramidal structures. For East European firms, information on pyramidal ownership seems to be lacking.
3. THEORY AND HYPOTHESES
The previous section indicated that ownership structure differences exist between East and West European countries. It is therefore expected that differences will also be found in the subsequent study. As indicated in the introduction, the analysis will be done on regional, sub regional and country level. It is expected to find similarities and differences in shareholder concentration, ultimate owner share, ultimate owner type and ownership via pyramidal structures compared to the discussed literature. The dataset and descriptive findings of this study will be discussed and presented in section four. Apart from identifying and unravelling the ownership structures within Europe, it is aimed to explain the differences. Some studies that try to explain ownership structure differences examine the effect of investor protection on ownership concentration (LaPorta, 1997, 1999; Demsetz and Lehn, 1985). Others used industry and company specific characteristics to explain differences (Demsetz and Villalonga 2001; Bebchuk, 1999; Demsetz and Lehn, 1985).
With the help of earlier findings, this study attempts to explain ownership structure differences by using variables belonging to the legal context of a country and company specific variables. This way it uses both variables from inside and outside the company to characterize the differences in ownership structures. These independent variables will be used to explain ownership concentration (i.e. share of the ultimate owner), widely held ownership (i.e. type of the ultimate owner) and pyramidal ownership as being the dependent variables. Shareholder concentration, being also an element of analysis, will not be regarded as an additional dependent variable as it is solely used in the descriptive analysis in section four. Subsequently, a number of hypotheses will be constructed which will be tested in section five. Finally this section reports the methodological issues regarding ownership structures in different industries.
3.1 Legal Context
Relation to Ownership Concentration and Widely Held Ownership
Ireland), characterized by better minority protection, indeed faced more ownership dispersion than countries with weaker minority protection as is the case in most continental European nations. They suggest subsequently that the continental European countries err from a lack of regulation because minority investors are less protected (Aglietta and Rebérioux, 2005). The theory of LaPorta et al. is characterized as being the most accepted, which might be due to the dataset of large companies they used in combination with an international analysis.
The methodology of LaPorta et al. is a suitable approach to compare countries. They used the anti director right index to label countries based on their minority shareholder protection which corresponds to the written rules a country has regarding the rights and protection of minority investors. This index ranges from 0-6, with 6 being the best protection. The methodology for this index will be explained in more detail in section 5. LaPorta et al. subsequently formed two groups of countries, i.e. high and low protection. It can be noted that from the European countries, only the UK, Ireland, Spain and Norway were classified in the group with high minority protection. The other European countries were characterized by low minority protection. They found that most of the high investor protection countries have a substantial number of widely held firms.
LaPorta et al. however did not include the East European countries in their analysis. Djankov et al. (2006) revised the anti director index of LaPorta et al. and extended it by a large number of countries, including East European countries. They used the same methodology as LaPorta et al. It is expected that the theory of LaPorta et al. also holds for East European countries because minority investors in these countries are also expected to invest their funds in countries with good legal protection. Therefore one can hypothesize, using the data of Djankov et al., that the European countries with good anti director rights are characterized by relatively more widely held firms and companies with relatively more dispersed ownership in terms of the share of the ultimate owner.
H1a Countries with good anti director rights are characterized by relatively more widely held firms and companies with relatively more dispersed ownership.
H1b West European countries have relatively more widely held firms and companies with relatively more dispersed ownership than East European countries due to higher anti director
rights.
The rule of law. However, there seems to be a problem with the approaches of LaPorta et al. (1999) and Djankov et al. (2006). By using the anti director right index, one only controls for the legal rules that are written down. Yet, the developedness of the regulatory infrastructure does not only consist of the written rules, but also on the extent to which these rules are followed. In many countries, deviations from the legal structure exist. Pajuste (2002) and Berglöf and Pajuste (2003) for example indicate that the enforcement of shareholder protection laws (effectiveness) lags behind on the quality of law (extensiveness) in the East European countries. But apart from the East European countries, this can also be the case in West European countries. To examine the relation of the effectiveness of the legal rules with ownership structures, the rule of law will be used. LaPorta et al. (1997) used the rule of law to evidence that better effectiveness of the legal system stimulates higher market capitalization. They used an index of average law and order tradition assessments of countries between 1982 and 1995. In this study the rule of law index of Kaufman et al. (2005) will be used to examine its effect on ownership structures. Kaufman et al. (2005) developed an index based on the perceptions of respondents in a country on the confidence they have in and abide by the rules of society (World Bank, 2007). This includes perceptions of the incidence of crime, the effectiveness and predictability of judiciary and the enforcements of contracts. This index ranges from -2.5 to +2.5. The average index number for the East European countries in this study is +0.03 as opposed to +1.49 of the West European countries under study, based on the data from 2005. (table 3.1).
Table 3.1: ADR, rule of law and legal origin
country ADR rule of law legal origin country ADR rule of law legal origin
Austria 2.5 + 1.87 German Belarus NA
Belgium 3.0 + 1.47 French Bulgaria 3.0 - 0.19 German Denmark 4.0 + 1.99 Scandinavian Croatia 2.5 + 0.00 German Finland 3.5 +1.96 Scandinavian Czech 4.0 + 0.70 German France 3.5 +1.35 French Hungary 2.0 + 0.70 German Germany 3.5 + 1.76 German Poland 2.0 + 0.32 German
Greece 2.0 + 0.66 French Romania 5.0 - 0.29 French
Ireland 5.0 + 1.63 English Russia 4.0 French Italy 2.0 + 0.51 French Serbia NA - 0.81
Netherlands 2.5 + 1.78 French Slovak 3.0 + 0.41 German Norway 3.5 + 1.99 Scandinavian Slovenia NA + 0.79
Portugal 2.5 + 1.10 French Ukraine 3.0 French Spain 5.0 + 1.13 French
Sweden 3.5 + 1.84 Scandinavian Switzerland 3.0 + 2.02 German
UK 5.0 + 1.69 English
average 3.4 + 1.49 3.2 + 0.03
This means that the perceptions regarding the effectiveness of laws are higher in West European countries than in East European countries. Because better investor protection is thought to decrease the degree of shareholder concentration (LaPorta et al., 1999) and better rule of law figures are found to stimulate market capitalization (LaPorta et al., 1997), it is expected that countries with better rule of law index figures have relatively more widely held firms and companies with relatively more dispersed ownership in terms of the share of the ultimate owner.
H2a Countries with higher rule of law indices are characterized by relatively more widely held firms and companies with relatively more dispersed ownership.
As West European countries have on average substantial higher rule of law indices, it is subsequently expected that West European countries have relatively more widely held firms and companies with relatively more dispersed ownership compared to East European countries.
H2b West European countries have relatively more widely held firms and companies with relatively more dispersed ownership than East European countries due to higher rule of law
indices.
Table 3.1 clearly illustrates the point of Pajuste (2002) and Berglöf and Pajuste (2003) in that the effectiveness of the East European regulatory systems is in general lower than its extensiveness. Also in the West European nations some countries score poorly.
Relation to Pyramiding
As indicated in section two, concentrated ownership results in larger benefits and costs for an owner (Demsetz and Lehn, 1985). It is therefore likely that an ultimate owner wants to increase its private benefits and decrease its costs. According to van der Elst (2004) shareholders will try to capture benefits when they are large. This theory is called the rent-protection theory and states that there is a connection between ownership structures and private benefits. Van der Elst furthermore argues that the ‘size of these benefits drives ownership patterns’ (p.425). According to LaPorta et al. (1997) the size of the benefits is limited by legal rules that protect investors. This subsequently implies that the size of the private benefits is determined by the extent of the legal infrastructure. These private benefits are mainly found in control rights, rather than in the cash flow rights. This is because the ultimate owner of the control rights can effectively control the firm as opposed to an owner with only cash flow rights. Moreover the controlling owner can exert control over a firm and leverage part of the costs to minority shareholders thereby decreasing its own costs and increasing its benefits. As can be noted, there is a difference in control rights and cash flow rights. This difference can be explained by figure 3.1.
Figure 3.1: Direct, indirect and pyramidal ownership
A B C 80 % 75 % D 20 % 25%
Greece, Italy, Netherlands, Portugal, Sweden and Switzerland) had on average a substantial larger amount of companies with pyramidal structures, nearly 30%. There is thus an indication that most of the West European firms show signs of pyramidal structures and that this might be attributed to the level of minority investor protection. As noted earlier, West European countries show on average slightly better anti director rights and far better enforcement of legal rules compared to the East European countries. This would assume that in East European countries more pyramidal structures exist.
H3 East European companies exert relatively more control via pyramids than West European companies due to lower anti director right and rule of law figures.
3.2 Firm Specific Variables Size of the firm
Apart from external variables, the degree of shareholder concentration might fairly well be attributed to internal variables. Demsetz and Lehn (1985) found that amongst other variables, size was negatively related to ownership concentration. They argue that this can be attributed to the higher market value of larger firms, and that it is subsequently more difficult for an owner to become a blockholder. This study uses two proxies for size namely the number of employees and the number of subsidiaries. It is hypothesized that larger numbers of both employees and subsidiaries result in more dispersed ownership.
H4a the larger the number of employees and the number of subsidiaries, the more dispersed the ownership structure is.
Turnover
Recall that the rent protection theory argues that shareholders try to capture the largest possible benefits and that the size of these benefits drives ownership patterns (van der Elst, 2004). On the one hand it might therefore be possible that the higher the turnover of a company is, the more concentrated the ownership is because the possibility for profits increases.
H4b The higher the turnover, the more concentrated the ownership structure is.
size. This would thus mean that higher turnover results in ownership dispersion. As there seems to be no other study using this variable, an expectation is difficult to make. Therefore a second hypothesis seems necessary.
H4c The higher the turnover, the more dispersed the ownership structure is.
Listed or non listed
In the second section it has been indicated that listed firms face relatively more dispersed ownership compared to non listed firms. This results form the fact that it is easier for minority shareholders to obtain shares in companies that are listed than in companies that are not. It is therefore expected that this will also be the case in this study. Using a dummy variable for being listed and being non listed, this study expects to find a significant result.
H4d Listed firms are significantly more dispersed than non listed firms.
Industry
4. DATA AND DESCRIPTIVE ANALYSES
In the previous sections it has been shown that differences in ownership structure exist between East and West European countries and that legal and firm specific factors are thought to be the reason for these differences. This section presents the dataset and descriptive findings of this study. Firstly the methodological issues of earlier studies will be discussed which results in the discussion of the methodology and dataset used in this study. Secondly the descriptive findings of shareholder concentration, the share of the ultimate owner, the type of the ultimate owner and the number of pyramids are presented on regional, law family and country level. During these stages there will be comparisons with the earlier studies as presented in section two. Finally, there will be a discussion about the differences and similarities in European ownership structures found.
4.1 Methodological Issues of Previous Studies
The problem with the study of Claessens and Tzioumis (2006) is that they failed to take into account widely held ownership as an ultimate owner type. They only analyzed the type of the ultimate owner when there actually was an ultimate owner (i.e. more than 50% of the shares). This might have biased the average percentages of the other ultimate owner types. Furthermore there is a data issue in the study of Claessens and Tzioumis as the number of East European listed firms studied is limited. As indicated in section two, they analyzed only companies with more than €50 million operational revenue/ turnover. This study will make use of a larger set of East European companies by using a lower threshold for revenue/ turnover (i.e. €25 million).
4.2 Database
The dataset used in this thesis is retrieved from the database Amadeus of bureau van Dijk (2006). This company is specialized in providing company information to its customers in various forms. Their database Amadeus is an information source for about 9 million public and private companies in 38 European countries. Bureau van Dijk retrieves its information from over 30 specialist regional information providers which limits the possibility of bias.
Table 4.1: Dataset; number of listed and non listed companies per country
country listed non listed total country listed non listed total
Austria 10 55 65 Belarus 0 25 25 Belgium 21 102 123 Bulgaria 6 33 39 Denmark 25 111 136 Croatia 14 21 35 Finland 37 43 80 Czech 7 44 51 France 152 554 706 Hungary 7 15 22 Germany 126 1011 1137 Poland 32 179 211 Greece 20 18 38 Romania 11 64 75 Ireland 13 28 41 Russia 59 62 121 Italy 63 277 340 Serbia 6 15 21 Netherlands 64 171 235 Slovak 3 18 21 Norway 5 25 30 Ukraine 11 125 136 Portugal 15 20 35 Spain 57 353 410 Sweden 68 195 263 Switzerland 76 151 227 UK 365 1810 2175 average 1117 4924 6041 156 601 757
It can be noted that the majority of the companies in this study have a West European origin. This indicates that there are simply more West European companies with more than €25 million operational revenue/turnover and more than 1000 employees than East European firms. The ratio of the listed firms to the total firms in this study is for both regions more or less similar, i.e. 18% for the West European countries and 21% for the East European countries. Furthermore it can be noted that in each region some countries are more present than others. The UK and Germany together account for 55% of the West European firms and Poland and the Ukraine together account for 46% of the East European firms in the sample. This can arguably be attributed to the size of the mentioned countries.
4.3 Ownership Types and Levels
Before presenting the findings regarding ownership structures in Europe, it is necessary to shortly discuss the ownership data of Amadeus. The database roughly distinguishes two types of ownership, direct and total (indirect). A company is owned directly when another company owns it on a direct basis. This is presented in figure 4.1.
Figure 4.1: Direct ownership
A B
75%
as no other owner can have a higher stake. The second type of ownership used by Amadeus, is the total or indirect ownership type. It is possible that a source reports to bureau van Dijk that a firm has both a direct and an indirect ownership link without specifying the path through which the indirect link is held. This is presented in figure 4.2.
Figure 4.2: Total/indirect ownership
A B
75%
10%
Amadeus subsequently makes a summation of the indirect and direct percentages and indicates it as a total ownership link. In figure 4.2 this means that A directly owns B with a 75% stake, indirectly owns B with a 10% stake via an unknown company and totally owns B with a 85% stake. Cumulating the ownership percentages between two companies is possible in this case as Amadeus tracks the control relationships rather than the cash-flow relationships.
The focus in this thesis is on the ultimate owner of a company. The database of Amadeus controls for this by calculating the link between the subject company and the controlling company. This can simply be the direct link or the summated total indirect link between two companies as indicated above, but it can also be the total control percentage from the subject company to the ultimate owner that is on top in a pyramidal structure. Two forms of pyramidal structures are possible here and will be shortly discussed as they have a slightly different methodology in calculating the ultimate owner percentage. The first form is presented in figure 4.3.
Figure 4.3: Pyramidal ownership
A B
75% C
100%
this case, cumulating control st lf has no indirect shares in B. Figure 4.4 shows the other form
company B as being the ultimate owner of C with a 40% stake, which looks on first sight h
as being t
akes is not possible as C itse .
Figure 4.4: Pyramidal ownership
Source: adapted from bureau van Dijk (2006)
Figure 4.4 presents the case where company C is owned by company A with a 30% stake and by company B with a 40% stake. Because company A effectively controls company B with 80% of the shares, it controls company C with an ‘extra’ 40% stake. In this case the subject company (C) is thus owned by two minority owners, but by the same majority owner at a higher level (A). Therefore Amadeus computes the calculated total percentage of control rights between A and C which is 70% (30+40). This method is considered just as it enables to identify the correct ultimate owner. If we do not pay attention to these structures one would identify
igher than the 30% of company A. In this study, thus, the 70% stake of company A is used.
Due to the methodological rules in the Amadeus database, one can limit the bias of identifying the wrong ultimate owners resulting in inaccurate study of ownership structures. The database of Amadeus also enables to distinguish between normal ownership relations and pyramidal structures. Total ownership for example points in Amadeus to a holding via other companies and is therefore, in this study regarded, as a pyramidal structure. There is however one exception. Amadeus indicates companies owned by an owner with a 100% stake also
otally owned. They reason that this is the case because there logically can not be another owner. Subsequently these companies are not labelled as a pyramid in this study.
Finally, Amadeus distinguishes between the immediate, domestic and global ultimate owner. One would expect that the global and domestic ultimate owner figures would deal with the nationality of the owner. This was however not consistent in each situation. To decide which ownership figure was used, the following procedure was performed. The figures of the global ultimate owner were perceived as being the highest in rank. If there was no figure for
the global ultimate owner, the domestic ultimate owner share was used. If this figure was missing, the immediate ultimate owner figure was used. The immediate shareholder of a company is the first level owner in the path towards the ultimate owner. If none of these three ultimate owner classifications was present in a firm, then there subsequently was no owner ith more than 25% of the shares and, subsequently, the company was reported to be widely
ompanies are presented. As indicated in the introduction and above this will be done on three st Europe, the four law families and the separate countries.
significant relation (-,564) (p<α) etween the number of shareholders and the share of the ultimate owner. Table 4.2 presents
e findings of both measures for the European countries. w
held.
4.4 Description of the Ownership Data
Below the descriptive findings regarding shareholder concentration, the share of the ultimate owner, the type of the ultimate owner and the number of pyramidal structures in European c
levels. The levels are West and Ea
4.5 Shareholder Concentration
To analyze the degree of shareholder concentration in Europe, two measures will be used. The first measure is the distribution of ownership concentration of European listed and non listed firms. This will be done using the same procedure as in Claessens and Tzioumis (2006). For each country, the percentage of firms with a shareholder at three thresholds will be calculated. This means that for every firm in a country it will be analyzed whether, the largest owner has a stake above 50,01%; a stake between 25,01% and 50%; or does not have a stake above 25% in the company. Subsequently one can determine whether the firms in a country can be characterized as having relatively dispersed ownership (i.e. highest percentage in the < 25% category), relatively concentrated ownership (i.e. highest percentage in the > 50,01% category) or having a mediocre ownership concentration degree (i.e. highest percentage in the between 25,01% and 50% category or more or less similar figures at each threshold). The second measure of ownership concentration is the average number of shareholders in a company. The more shareholders are present in a company, the smaller their individual shares are and the more dispersed the ownership is expected to be. Performing a simple bivariate correlation analysis confirms this by finding a negative
Table 4.2: Shareholder concentration in Europe
non listed listed
country Obs. > 50,01% 25,01%-50% < 25% average no. shareholders Obs. > 50,01% 25,01%-50% < 25% average no. shareholders Austria 55 87,27% 3,60% 9,09% 5 10 40,00% 0,00% 60,00% 25 Belgium 102 79,41% 2,94% 17,65% 3 21 23,08% 38,20% 38,20% 28 Denmark 111 70,72% 7,20% 22,52% 2 25 16,00% 12,00% 72,00% 24 Finland 43 93,02% 2,33% 4,65% 1 37 10,81% 24,32% 64,86% 45 France 554 92,24% 2,53% 5,23% 2 152 38,82% 23,03% 38,16% 21 Germany 1011 85,56% 7,52% 6,92% 3 126 38,88% 23,02% 38,01% 22 Greece 18 66,67% 22,22% 11,11% 3 20 30,00% 30,00% 40,00% 11 Ireland 28 89,29% 3,57% 7,14% 3 13 7,69% 7,69% 84,62% 44 Italy 277 75,01% 7,58% 17,33% 6 63 52,83% 22,22% 25,40% 20 Netherlands 171 88,30% 7,60% 4,10% 2 64 12,50% 17,19% 70,31% 24 Norway 25 100,00% 0,00% 0,00% 2 5 40,00% 0,00% 60,00% 43 Portugal 20 85,00% 5,00% 10,00% 2 15 33,33% 20,00% 46,67% 22 Spain 353 79,60% 5,38% 15,01% 4 57 31,58% 17,54% 50,88% 63 Sweden 195 94,36% 1,54% 4,10% 1 68 4,42% 30,88% 64,70% 31 Switzerland 151 90,73% 2,65% 6,62% 2 76 30,26% 10,53% 59,21% 23 UK 1810 94,31% 3,26% 2,43% 3 365 3,46% 6,58% 89,86% 41 Belarus 25 88,00% 4,00% 8,00% 2 0 0,00% 0,00% 0,00% 0 Bulgaria 33 78,79% 6,06% 15,15% 2 6 66,67% 16,67% 16,67% 3 Croatia 21 95,24% 4,76% 0,00% 2 14 50,00% 7,14% 42,86% 10 Czech 44 84,01% 6,90% 9,09% 1 7 71,43% 14,29% 14,29% 8 Hungary 15 100,00% 0,00% 0,00% 3 7 42,86% 42,86% 14,29% 28 Poland 179 92,18% 6,70% 1,12% 2 32 34,38% 31,25% 34,38% 13 Romania 64 84,38% 10,94% 4,69% 3 11 100,00% 0,00% 0,00% 5 Russia 62 53,23% 11,29% 35,48% 3 59 38,98% 30,51% 30,51% 7 Serbia 15 80,00% 6,67% 13,33% 1 6 83,33% 16,67% 0,00% 3 Slovak 18 88,89% 5,56% 5,56% 1 3 33,33% 33,33% 33,33% 4 Ukraine 125 54,40% 21,60% 24,00% 3 11 18,18% 45,45% 36,36% 5 Averages 84,10% 6,27% 9,64% 3 35,29% 19,31% 41,69% 21 West Europe 4924 85,72% 5,31% 8,99% 3 1117 25,85% 17,70% 56,43% 30 East Europe 601 81,74% 7,68% 10,58% 2 156 49,01% 21,65% 20,24% 8 English origin 1838 91,80% 3,42% 4,79% 3 378 5,58% 7,14% 87,24% 42,5 French origin 1746 75,82% 9,71% 14,46% 3,1 473 37,93% 24,41% 37,65% 20,6 Scandinavian origin 374 89,53% 2,77% 7,82% 1,5 135 17,81% 16,80% 65,39% 35,8 German origin 1567 88,24% 4,95% 6,81% 2,2 287 44,65% 17,80% 28,46% 12,6
ownership structure based on these findings with 75,82% of the non listed firms having an owner of more than 50,01% of the shares. These findings seem to obstruct the hypothesis that good legal infrastructures facilitate dispersed ownership. In general, it can be argued that European non listed companies are characterized by high ownership concentration. The only main difference with the findings of Claessens and Tzioumis is, that this study finds on average slightly more non listed firms to have an owner between 25% and 50%.
Looking at the listed companies, one can notice that larger differences exist compared to the non listed firms. Overall, the listed companies face relatively more dispersed ownership as was the case with the non listed firms. Focussing firstly on the regional level, it can be noticed that West European companies are characterized by relatively more dispersed ownership compared to the East European companies. Only 25,85% of the West European listed companies have a shareholder of more than 50,01% compared to 49,01% of the East European firms. 56,43% of the West European firms are reported to have no shareholder with more than 25% of the shares compared to only 20,24% in East Europe. Even without the UK, the West European companies still can be characterized by relatively more dispersed ownership compared to their East European counterparts. Without the UK, 27,35% of the West European firms were found to have an owner with more than 50,01% of the shares compared to 54,20% of the firms having no majority owner.
countries are also characterized by the lowest percentage of firms having no owner with more than 25% of the shares (28,46%). Closely behind the German legal origin countries one finds the French legal origin countries with 37,93% and 36,65% of the respective figures. Scandinavian and English origin countries can subsequently be characterized as relatively dispersed with 17,81% and 65,39% for Scandinavian legal origin countries and 5,58% and 87,24% for English origin countries of the respective figures. These findings seem to agree with the hypotheses that the legal infrastructure of a country is related to the degree of shareholder concentration. This will be empirically studied in section six.
To test if there is a significant difference between West and East Europe in terms of shareholder concentration, a simple independent samples T-test was performed. We use the average share of the ultimate owner in a firm, descriptively analyzed for each country in the next sub section, to determine if a significant difference exists in ownership concentration between East and West Europe. To perform this, the share of the ultimate owner of each company was divided by the average share of the European ultimate owner. This average is 87,22% for non listed firms and 54,44% for listed firms. The new variable that was created by this procedure is thus a ‘dispersion ratio’ which for each company indicates if it is more concentrated or dispersed compared to the average ultimate owner in Europe. If the ultimate owner share of an owner in a listed company is for example 60%, the dispersion ratio would be (60/ 54,44) 1,10 which indicates that the company has more concentrated ownership compared to the average European firm. Using this procedure, we can compare the means of both groups of companies and determine which group (i.e. East or West European firms) have more concentrated ownership figures. The results of this independent samples T-test are presented in table 4.3
Table 4.3: Independent samples T-test; difference East and West Europe regarding shareholder concentration
non listed firms
Region N Mean Std. Deviation Std. Error Mean Sig. West 4924 1,012 0,280 0,004 ,000
East 601 0,905 0,333 0,014
listed firms
Region N Mean Std. Deviation Std. Error Mean Sig. West 1117 0,645 0,332 0,010 ,000
East 156 0,939 0,454 0,036
