Working Papers R&D

OWNERSHIP STRUCTURE OF THE 250 LARGEST LISTED COMPANIES IN GERMANY by S. KLEIN* and C. BLONDEL**

2002/123/IIFE

*

Research Fellow, Family Firms, INSEAD, Boulevard de Constance, 77305, Fontainebleau Cedex, France.

**

Senior Research Programme Manager, Family Firms, INSEAD, Boulevard de Constance, 77305, Fontainebleau Cedex, France.

A working paper in the INSEAD Working Paper Series is intended as a means whereby a faculty researcher's thoughts and findings may be communicated to interested readers. The paper should be considered preliminary in nature and may require revision. Printed at INSEAD, Fontainebleau, France. Kindly do not reproduce or circulate without permission.

Ownership Structure of the 250 Largest Listed Companies in Germany Working Paper October 2002

Sabine B. Klein Research Fellow, Family Firms [email protected] Christine Blondel Senior Research Programme Manager, Family Firms [email protected] INSEAD Boulevard de Constance F – 77305 Fontainebleau Cedex France Tel: +33 (0)1 60 72 40 00 Fax:+33 (0)1 60 74 55 00 This research benefited from the financial support of the Tetra Laval Research Fund for the Large Family Firm, which is hereby gratefully acknowledged. Preliminary results were presented in conferences in September 2001 (MGI) and April 2002 (Campden)

EXECUTIVE SUMMARY Although it is often stated that family businesses are the prevalent form of business in most economies, they are mostly treated equivalent with small and medium sized businesses. Apart from the still unsolved family business definition problem this is the reason why there is so little grounded research and science of family businesses in Germany. This study shows that even in the group in which no one would assume a high family influence, namely the 250 largest listed companies in Germany, there is a strong influence of individuals and families. This study follows the design of the French study of Blondel, Rowell & Van der Heyden (2002) to ensure comparability of French and German results. After a short review of French and German family business research, the French study of Blondel et al. is summarised. While definitions and measures of family influence in this study are the same as in the French study, some specificities of the German study will be discussed such as the data sample, the percentage for blocking votes, etc. Main results will be discussed as well in terms of numbers as in terms of capitalization, branches and evolution from 1993 to 1998. The first result is that overall, patrimonial firms (those which are controlled by either individual owners or families) are the prevalent form in Germany and in France in the biggest 250 listed companies. In both countries more than 50% of these companies are patrimonial companies.

Family business studies Family business research dilemmas Families have influenced economy from the very beginning of economic history. Despite that, the family business field as an academic discipline is still in its infancy (Astrachan & Bowen, 2001, p.v). In terms of both publications and Chairs at universities, family business as an academic field still has to emerge in Germany (Klein, 2002). The two biggest issues in the international development of the field are the definition confusion on one hand and the lack of “family business facts” (Shanker & Astrachan, 1996, p.108) on the other. Both issues affect Germany as well. Data published in Germany until 2000 referred to SMEs (e.g. Ballarini & Keese, 1997), had no definition of family business (Albach & Freund, 1989; Wieselhuber & Spannagl, 1988) or were based on extrapolation of income tax sheets (e.g. Schröer & Freund, 1999). Westhead, Cowling & Storey (1997, p.16) showed that the number of family businesses one will get as a result of an empirical study depends very much on the definition chosen. The different definitions therefore are the main reason for which empirical data of different © INSEAD, October 2002.

2

studies cannot be compared. To find a way out of this definition dilemma Astrachan, Klein & Smyrnios (2001) developed a scale to measure the family influence onto any business, taking into account the existing definitions. Instead of dividing all businesses into family businesses and non-family businesses the developed scale enables researchers to assess family influence onto a business on a continuous scale, the “F-PEC” scale. The F-PEC scale comprises three subscales: Power, Experience, and Culture, hence the three letters (PEC – F standing for Family). The F-PEC Power subscale itself includes power ensured via ownership, management and/or governance of the family. In the study reported in this paper only the ownership dimension of family influence will be used because it is well documented and, as will be shown later, is the relevant variable in Germany through which families and/or individuals seek influence. Existing studies on family businesses in France and Germany Cross-country comparisons of family influence onto companies are relatively young. La Porta et al (1999) compared family influence via ownership in the 20 largest publicly traded companies in 27 countries. The results varied from 5 % of these companies with families as main shareholders (U.K.) up to 70% (Hong Kong). Faccio & Lang (2000) found out that the majority of the 3,740 publicly traded companies of their sample were owned by families. Families were main shareholders in 70% of the French companies of the sample, 72% of the German, 65 % of the Italian and 67 % of the Spanish while in the U.K. only 34% of the companies had a family as main shareholder. The first German research on family influence via ownership in large German companies was carried out in 1965 by Pross. His data showed that only 20% of the biggest German companies were majority-owned by a family or an individual (Pross, 1965, p. 113). While Pross only analysed the first owner, Schreyögg & Steinmann (1981, p. 538) analysed the chain of ownership of the 300 biggest companies of all legal forms in Germany in 1972. The aim was to show how many of these companies were respectively controlled i) by owners or ii) by management because of the lack of owners’ influence. Schreyögg & Steinmann (1981, p. 546) found that 50 % of the companies were controlled by owners, and 50 % by management. In terms of turnover, the results differed, with only 35 % of the total turnover of the 300 biggest companies controlled by owners. Witte (1981, p. 742) in contrary looked explicitly for family influence via ownership. He found that 44 % of the Aktiengesellschaften in Germany in 1976 with more than 2000 employees were owned by families or individuals. Although Witte asked for family ownership he did not, as well as Pross and Schreyögg & Steinmann, focus on family business but on ownership influence onto management decisions and return variables. Family business research in Germany is relatively young and up to 2000, there were no “family business facts”, i.e. no actual empirical research based on random sample and precise definition. The first and so far only study that satisfies this criteria was published in 2000 (Klein, 2000a, 2000b). Based on a random sample of more than 1000 German companies of all sizes, Klein (2000a, p. 44) found that nearly 30% of companies with more than one billion DM turnover were family businesses in 1995. Although the definition © INSEAD, October 2002.

3

included influence via ownership, management and governance, ownership proved to be the dominant variable via which families in Germany seek influence. 80 % of all family businesses in that sample were 100 %-owned by families (Klein, 2000b, p. 165). Ownership share of the families declined a) with increasing number of family owners and b) with increasing size of the company. Insofar it is interesting to learn the level of influence reached via ownership by families and/or individuals in the group of the biggest listed companies. As Blondel et al (2002) showed for France, this paper examines a comparable German group of companies - namely the 250 biggest listed companies on the Frankfurt stock exchange. The research design is analogue to the one of the French study to ensure comparability. The French study of Blondel, Rowell & Van der Heyden (2002) To find out whether family and/or individuals also influence large companies in France, Blondel, Rowell & van der Heyden (2002) analysed the ownership structure of the SBF 250, the index of the 250 largest listed companies on Paris stock exchange. The sample is well documented and Blondel et al assumed that the family influence onto listed publicly traded companies of that size would be smaller than in other samples. In other terms, if it is found that this sample includes a significant influence of families and individuals, one could assume that the influence will even be bigger in other samples. Influence via ownership was chosen because ownership structure of publicly traded companies is relatively well documented (Blondel et al, 2002, p. 2). As the French study focused on the ownership of individuals and/or families, Blondel et al tried to avoid the definition problem of family business research by looking for what they termed patrimonial firms. The criteria used in the French study were also used in the German study, with some distinctive features that will be described further. The key features of the French study are: 1. Definition: A patrimonial business is a business where one or several individuals or families are identified as the major ultimate owner or owners. The family does not need to be descendents of the firm’s founder(s). The aim is to understand the role that individuals or families play in the economy, including those individuals or families that purchase businesses. 2. Population: The 250 largest French companies that are listed in the so-called SBF 250, an official list of the 250 largest publicly-traded companies on Paris stock exchange. Studying publicly traded companies allows access to comparable data. 3. Scope: The aim is to look at the relative importance of different types of ownership and at their distribution across economic sectors. The size of ownership stakes and of voting rights were also analysed.

© INSEAD, October 2002.

4

4. Time: The SBF 250 was analysed at two points in time, end 1993 and end 1998, with the aim to understand the evolution of ownership by individuals and families in French business. Ownership types as defined by Blondel, Rowell & Van der Heyden (2002, p. 4) As far as possible the ultimate ownership of the companies listed in the SBF250 was identified. When the main owner of company A was another company B, the owner of B was identified and so on - until the last identifiable owner was found. Blondel et al considered that this search produced identifiable ownership when the stake of the owner(s) in each step of the ownership chain was at least 10% of equity. It should immediately be pointed out that most of the patrimonial firms in the SBF250 had rather concentrated ownership, including absolute majority (above 2/3 of the shares) and simple majority (between1/2 and 2/3 of the shares). At the end of the chain, several cases can occur: An identifiable owner is determined, and it is: . A family, or a group of families, an individual or a group of individuals: this company is called a patrimonial firm; . A corporation with no further identifiable owner: this company is called a single corporate-owned firm; . The state: this company is called state-owned; . A mutual fund, non-profit organization or an employee representative body: this company is called mutual-owned -

No identifiable owner can be determined: this company has spread ownership. It should be noted that this definition of spread ownership includes joint control situations involving two or more companies with equal stakes.

Blondel et al looked at control both in terms of ownership and voting rights. They also examined the degree of control of ultimate owners, taking as degree of control the weakest degree of control in the chain, when ownership was indirect. Summary of the results of the French study The study established that 57 % of the SBF 250 companies were patrimonial companies in 1998, although they only represented 35 % of the overall capitalisation of the 250 biggest companies. Both in number and in capitalisation, the percentage of patrimonial firms increased from 1993 to 1998 in France (from 48 % to 57 % in numbers, from 32 % to 35 % in terms of capitalisation). Except in property, energy/utilities and finance/insurance the patrimonial firms represented more than a third of each sector, in some they even represented more than 2 thirds (automobile, media, etc.) and 100% in luxury.

© INSEAD, October 2002.

5

In 1998, 20 % of the French patrimonial firms of the SBF 250 were controlled with absolute majority (more than 2 thirds of the shares); 37 % of the patrimonial firms were controlled with simple majority (more than half of the shares); 26 % were controlled with a “blocking stake” (more than one-third of the shares). In the remaining 17 % of patrimonial companies, families and individuals only had marginal or relative control (between 1/10 and one third of the shares). The difference between the latter was defined as marginal control standing at less than twice the next owner, and relative control at more than twice the next owner (Blondel et al, 2002, p. 9).

The German Study Specificities of the German study Selecting a stock exchange While France is a centralized state, the German structure is completely different, because of a different long term history. Germany is a decentralised country, a conglomerate of 16 small and more or less independent Bundesländern. While Paris is the principal stock exchange in France, in Germany there are 6 different stock markets. However, the most relevant internationally is Frankfurt stock exchange, which therefore was chosen for the German study. Defining the 250 largest companies by capitalisation While France focuses on stock market prices, Germany seems more customer market orientedParis stock exchange measured company size in terms of capitalization (= stock market prize of a company), but German stock exchanges were still using turnover (= what customers paid for the products of a company) up to 1997. Moreover, no official list or index exists in either of the German stock exchanges, at the exception of the DAX (list of 40 most important companies) and the M-DAX (list of 100 companies). For the year 1998, the Frankfurt stock exchange published a list of listed companies with their capitalisations, but for 1993 we could not find any official list of companies with their capitalisations. It was therefore necessary to build this list - which was done with the help of the Institut für Bank, Börsen und Versicherungen, Humboldt Universität Berlin, directed by Professor R. Stehle. The Institut had a list of traded companies with the number of shares and the share price at the end of 1993, which allowed us to select and rank by capitalisation the 250 largest companies. One of the additional problems we faced is the fact that many companies had two classes of shares, those with and those without voting rights: the capitalisation of both classes had to be added to get the overall capitalisation of the company. Finding ownership structure In Germany, companies are bound by law to give information on shareholders’ voting rights, but not on their ownership stakes. Therefore in some cases we could only access the © INSEAD, October 2002.

6

overall voting right structure. Hence this study and its results are based on voting rights rather than on ownership stakes. Another problem was that information concerning ownership structure in 1993 was in some cases hard to obtain. If there was no source that could be trusted and no answer from the company when asked by letter, a missing value for this company was accepted in order not to falsify the data. In total, information was missing for 18 companies in 1993, and for 4 companies in 1998. The percentages shown in our results are calculated on the total of companies with available data, i.e. 246 in 1998 and 232 in 1993. Accounting for country differences in control levels In our analysis of control levels, we took into account the difference in thresholds in France and Germany: when 1/3 of the votes are necessary in France to block some decisions, in Germany ¼ are sufficient. Consistently, absolute control in France starts with 2/3 of votes, while in Germany the minimum is for absolute control is ¾ of votes. We used the same terminology for both the German and the French study, but corresponding percentages are different because of national laws. Economic sectors As one of the long-term aims of this study is to compare German and French data the French economic sector classification was used for the German study. The sectors given by the German companies in some cases did not match optimally with the French sector code system. For instance, none of the German companies called itself agro-alimentary but fast moving consumer goods instead. For reliability reasons, the sector code which was taken was the one that the company itself had chosen on a questionnaire sent to them by Commerzbank in order to publish the “Wer gehört zu wem?”. The sectors given there by the companies were then fitted into the French sector system, which with its 16 sectors is less differenciated then the Commerzbank one. The 60 Commerzbank sectors equal the first level of NACE (Statistical classification of economic activities in the European Community). In some cases when companies had given multiple sectors in which they were active, if no particular sector was dominant the company was judged as “others” in the sector code.

© INSEAD, October 2002.

7

Results of the German study End 1998, patrimonial firms were the prevalent ownership form of the 250 largest listed companies in Germany: 51 % (126 companies) were owned by individuals and/or families. Table 1 shows that the percentage of patrimonial companies in the total decreases with the increasing size of the companies. Of the top 120 companies 39 % were patrimonial, of the top 20 only 15 % were patrimonial. Table 1: Prevalence of patrimonial firms Patrimonial firms (1998)

Number

%

Top 20

3

15 %

Top 120

47

39 %

Top 250

126

52 %

Result 1:

Families and/or individuals own the majority of the 250 largest listed companies in Germany.

Result 2:

Of the biggest 20 listed companies in Germany 1998, 15 % (three companies) were owned by individuals or families.

In terms of capitalisation the patrimonial firms represent 19 % of the overall capitalisation of the 250 companies, 18 % of the top 120 and 13 % of the top 20 (see table 2). There is a simple reason why the percentage of patrimonial firms measured in terms of capitalisation does not significantly increase from top 20 to top 250 whereas it does counted by numbers: the top 20 German companies in 1998 represent 74 % of the overall capitalisation, the top 120 even 98 %. Table 2: Capitalisation of the Top 250 German listed companies Capitalisation (DM)

Total (billion DM)

Patrimonial firms

Top 20

1 183

156

13 %

Top

1 581

292

18 %

Top 250

1 609

310

19 %

Result 3:

%

(Billion DM)

In terms of capitalisation patrimonial firms are smaller on average than nonpatrimonial ones. They only represent 19 % of the overall capitalisation of the largest 250 listed German companies in 1998.

© INSEAD, October 2002.

8

Other Ownership Types The analysis of ownership types reveals that the second largest group of companies were owned by other companies, which ultimately had no single identifiable owner. This group, called single corporate companies accounted for 21% of the companies end 1998. The next group, composed of companies with spread ownership, represented 18% of the companies. State-controlled companies accounted for 7%, with the remaining 3% under “mutual” control. Thus the common assumption that stock exchanges are made of companies with diffused ownership is wrong in the case of Frankfort stock exchange. Coming back to the question of control first raised by Pross (1965), our research shows that there are less management-controlled companies than owner-controlled companies in the top 250 listed German companies in 1998. Patrimonial companies are owned by an individual or/and families that have more influence (by definition) than the other possible owners, at the exception of the State. Whether families use their power to influence the strategy, the human resource management, the investments of the company can not be proved neither rejected based on the data of this study. Companies with spread ownership and companies owned by other companies with ultimate spread ownership (the single corporate companies), are controlled by their management more than by their owners. There is, by definition, no owner who is powerful enough to influence the company´s policies via ownership share. In summary, 52 % of the top 250 companies in 1998 were owner-controlled (patrimonial) and 39 % manager-controlled (spread and single corporate). However, the balance was reversed amongst the larger companies, with 39 % of the 120 largest companies controlled by owners in 1998, and 48 % controlled by management. Evolution from 1993 to 1998 A comparison of our results with the Schreyögg and Steinmann’s (1981, p. 546) would evidence a stable number of owner-controlled companies over the past 20 years in Germany. However, because of differences in the studies, this cannot be stated as a result but only formulated as a hypothesis. In fact, the results of our study do not confirm this hypothesis. The number of patrimonial firms in the 250 largest listed companies in Germany increased from 39% to 52% of the companies. In parallel, companies with spread ownership increased in number during these 5 years, from 12 to 18%. At the same time, the single-corporate firms, firms owned by other firms with no ultimate identifiable owner, decreased from 38% in 1993 to 21% in 1998. This development can be interpreted as the breaking-up of the so-called “Deutschland-AG”, a synonym for the network of German companies that over the years developed reciprocal ownership and inter-dependence. Indeed, it was a priority of German government in the 1990s to proceed with the “Entflechtung” (detanglement) of German industry. With the globalisation of the international financial markets this inter-dependence had negative influence onto the market value of the companies.

© INSEAD, October 2002.

9

Table 3: Ownership types of German listed companies in 1993 and 19981 Ownership type

1993 (Top 250)

1998 (top 250)

Patrimonial

39 %

52 %

Single corporate

38 %

21 %

State

8%

7%

Mutual

4%

3%

Spread

12 %

18 %

Total

100 %

100 %

Result 4:

From 1993 to 1998 there are clear differences in terms of ownership type. The number of patrimonial firms increased by 41%, the number of spread companies increased by 67%, and the number of single corporate firms decreased by 41%.

There was almost no difference between 1993 and 1998 for the top 20. There was a slight decrease in terms of patrimonial firm capitalisation, but not significant (from 15 % to 13 %). In the top 120, a slight increase of capitalisation of patrimonial firms took place, from 17 % in 1993 to 18 % in 1998, not significant either. The evolution of capitalisation was influenced by the high percentage represented by the top 20 companies within the overall capitalisation. Table 4: Patrimonial firms in 1993 and 1998 Patrimonial firms (%)

In top 20

120

250

In 1993

15 %

25 %

39 %

In 1998

15 %

39 %

52 %

In 1993

15 %

17 %

19 %

In 1998

13 %

18 %

19 %

Number (%)

Capitalisation (%)

Result 5 : The number of patrimonial firms in the top 20 was stable from 1993 to 1998, while it increased significantly in the top 120 and the top 250.

1

Percentages calculated on the number of companies with ownership data, ie 246 companies in 1998, 232 in 1993.

© INSEAD, October 2002.

10

Result 6:

Although the numbers increased there are only slight differences in terms of capitalisation, but not significant ones.

Presence by economic sector Patrimonial firms were prevalent in the Luxury sector, Services General and Property. In these sectors they represent 80 % and more of the companies in terms of numbers. In the Transport sector none of the companies was found to be patrimonial and only 2 of the 10 companies in the Energy/Utility sector were patrimonial firms (20 %). In the Finance/Insurance sector 17 % are patrimonial firms. In sectors where they were predominant in numbers, patrimonial firms accounted for an even higher stake of capitalisation, while in sectors where patrimonial firms were underrepresented in terms of number, their share of capitalisation was almost non-existent (Energy/Utility 1%, Finance 2%). The other sectors could be divided in three groups: a) percentage of capitalisation significantly lower than the percentage of number: Pharmacy, Automobile, Fast Moving Consumer Goods, General Industry, Services Leisure & Tourism and Construction, b) percentage of capitalisation and of number almost identical: Holding and High Technology, and c) percentage of capitalisation significantly higher than percentage of number (Retail, Media). Table 5: Repartition of patrimonial firms by sector in 1998 (based on top 250) Number of companies Luxury Holding Retail Automobile Media Services – General F.M. Consumer Goods Hi Technology Pharmacy Services –Leisure & Tourism Transport Gen Industry Construction Property Energy/Utilities Finance/Insurance Other TOTAL © INSEAD, October 2002.

Total sector 4 15 22 11 5 7 28 9 14 3 2 48 14 5 10 29 22 243

Patrimonial 4 7 12 6 3 6 16 3 6 2 0 27 8 4 2 5 15 122

% number 100 % 47 % 55 % 55 % 60 % 86 % 57 % 33 % 43 % 66 % 0 56 % 57 % 80 % 20 % 17 % 68 % 50 %

% cap. 100 % 54 % 92 % 14 % 93 % 97 % 32 % 38 % 12 % 32 % 0 27 % 25 % 90 % 1% 2% 48 % 19 %

11

Result 7:

Although patrimonial firms were found in nearly every sector there were sectors in which they were dominant and sectors in which they were nearly non-existent in terms of capitalisation.

Result 8:

Sectors in which patrimonial firms were dominant are luxury, retail, services general, sectors in which they were weak are transport, finance and energy.

Ownership type per sector changed a lot between 1993 and 1998. Sectors like Holding and Fast Moving Consumer Goods grew, Energy/Utility and Finance/Insurance shrank. Patrimonial firms seem to be successful in terms of survival of this ownership type in growing industries and fast moving industries while they are less successful in concentrating branches. In the Finance sector for example the total number of companies in the sector decreased significantly. 17% of the companies in this sector are patrimonial companies but they only represent 1 % of the overall capitalization of this sector. Goehler (1993, p. 316) stated that family businesses do have an advantage in the first two stages of the life cycle (founding and growth) while they have relevant disadvantages in the last two (maturity and decline). Looking at the sector analysis this seems to point to the same direction. Energy, Utility and Finance as fast concentrating markets are old sectors and the patrimonial companies are declining in this sector. Because the sector analysis is too broad to verify this it should be formulated as a grounded hypothesis to be verified/falsified in further research. Hypothesis A:

Patrimonial firms are less successful in terms of survival in old, concentrating sectors than in growing or fast moving sectors.

Hypothesis B:

In concentrating sectors patrimonial firms do not reach the size to compete with non-patrimonial firms.

Table 6: Changes within sectors in the top 250 from 1993 to 1998 Total sector

Patrimonial

Luxury

-1

+1

Holding

+4

+5

Retail

+7

+3

Automobile

0

+3

Media

+4

+2

Services – General

+5

+5

F.M. Consumer Goods

+8

+ 17

Difference 1998 vs. 1993 (number of companies in the Top 250)

© INSEAD, October 2002.

12

Hi Technology

0

-2

Pharmacy

-1

0

Agro alimentary

-

-

Services –Leisure & Tourism

0

+1

Transport

-3

-2

General Industry

0

+16

Construction

-2

-2

Property

-1

+1

Energy/Utilities

-16

-3

Finance/Insurance

-21

-2

TOTAL Looking at the overall change between 1993 and 1998 one will find that 44 patrimonial companies that were listed in 1993 were still listed and patrimonial in 1998. 41 patrimonial companies dropped out, either they disappeared from the stock exchange, or they were not among the biggest 250 any more, or they were not any longer patrimonial firms in 1998. Only 4 of the 41 former patrimonial companies of the 1993 list were still on the 1998 list but not patrimonial any longer. Some of the former patrimonial companies were consolidated with their former mother companies, which were and remained patrimonial. Conversely, in two cases patrimonial firms present in the 1993 list had spin-offs that also became part of the 1998 list (e.g. Fresenius AG in 1993, Fresenius AG and Fresenius Medical Care in 1998). In 1998 there were 84 patrimonial firms that had not been patrimonial or on the list in 1993. Result 9: There was a relevant entry of patrimonial firms between 1993 and 1998 in the list of the biggest 250 listed companies (+84 companies), and exits of only less than half that number (-41). About half of the 90 patrimonial companies in 1993 (44) remained on the list as patrimonial companies in 1998. Level of control Looking at the level of control in 1998 one must be aware that the level of control relates to the voting shares if voting stock and overall stock are different. A large majority of the patrimonial firms were controlled with more than half of the voting shares by the individuals or families owning them. Specifically, 61 % of the patrimonial firms showed majority or absolute control levels in 1998. Only 11 % of the patrimonial firms showed less than a blocking vote.

© INSEAD, October 2002.

13

Table 7: Levels of control (Top 250 in 1998) Patrimonial firms’ control

Number of companies

% of total patrimonial

Absolute (> 75% of votes)

24

19 %

Majority (50% to 75%)

53

42 %

Minority (25% to 50%)

35

28 %

Marginal (10% to 25%)

14

11 %

(1 missing) TOTAL PATRIMONIAL

127

100 %

Result 10: 61 % of the patrimonial firms in 1998 were controlled by majority or absolute control. Looking at the development over time the numbers shown in table 8 are interesting. The level of control decreased on average within the group of patrimonial firms from 1993 to 1998. In numbers, the number of patrimonial firms controlled through at least a majority of voting rights increased from 69 to 77 companies. However, at the same time, the number of patrimonial firms controlled with minority or less increased from 21 companies to 49. As a result, while 77% of the patrimonial firms were at least majority controlled in 1993, this number dropped to 61% in 1998 Table 8: Development of level of control from 1993 to 1998 1993

1993

1998

1998

Number of patrimonial firms

% of total

Number of patrimonial firms

% of total

Absolute (> 75 % of votes)

33

37 %

24

19 %

Majority (50% to 75 %)

36

40 %

53

42 %

Minority (25 % to 50%)

17

19 %

35

27 %

Marginal (10% to 25 %)

4

4%

14

11 %

TOTAL

90

100 %

126

100 %

Patrimonial firms’ control in the top 250

Result 11: Over the 5 years from 1993 to 1998 the level of control in the patrimonial firms decreased on average. The number of companies controlled through at least

© INSEAD, October 2002.

14

majority grew from 69 to 77 companies, while the number of companies controlled with minority or less increased from 21 to 49. Of the 44 patrimonial firms that were present within the 250 largest listed companies both in 1993 and 1998, we find 18 companies (42 %) that remained at the same level of control, 19 companies (42 %) in which the families and /or individuals decreased their level of control, and only 7 companies (16 %) in which the families and/or individuals increased their level of control. Also, if absolute control is equivalent to 1, majority equals 2, minority 3, and marginal 4, the new patrimonial entrants on the list had an average control level of 2.4 while those exiting the list had a control level of 1.9. Result 12: The alteration in the level of control of patrimonial companies came from two sources: first; a decrease in control of the companies that remained in the list of top 250; second, a lower control level on average of the new entrants compared to the exiting ones. Conclusion Following the design of the French study of Blondel et al, this study analysed the 250 largest publicly traded companies on the Frankfurt Stock Exchange. Frankfurt stock exchange was chosen because it is the biggest and the internationally most important of the six German stock markets. Our aim was to identify ultimate ownership at two points of time, end of 1993 and end of 1998. The high percentage of patrimonial firms, firms in which the ultimate owner is either an individual or a family, can be viewed as the most surprising result. 52 % of the companies in 1998 were patrimonial firms following the definition of the French study. The percentage of patrimonial firms in Germany over the given period increased by 41% from 90 companies in 1993 to 127 in 1993, which is an even higher increase than in France (18 %). In terms of absolute numbers 83 firms entered the list of patrimonial firms between 1993 and 1998, and only 42 left it, while half of the 90 patrimonial companies in 1993 (44) remained on the list as patrimonial companies in 1998. Concerning the sector analysis the study led to two main results: Although patrimonial firms are found in nearly every sector there are sectors in which they are dominant and sectors in which they are nearly non-existent in terms of capitalisation. Sectors in which patrimonial firms are dominant are Luxury, Retail, Services General. Sectors in which they are weak are Transport, Finance and Energy. The sector analysis turned out to be not differentiated enough to verify or falsify the hypothesis that patrimonial firms are less successful in old sectors. A more refined study is needed to examine this interesting hypothesis. The last area of this study concerned the level of control within the patrimonial companies. Nearly two thirds (61 %) of the patrimonial firms in 1998 were controlled by majority or © INSEAD, October 2002.

15

absolute control. Over the 5 years from 1993 to 1998 the average level of control in the patrimonial firms decreased significantly. While the number of patrimonial companies with at least majority control increased from 69 to 77 companies, the number of companies with minority or marginal control increased from 21 to 49. The alteration in the level of control comes from two sources; the family and/or individuals that owned companies in 1993 and whose companies stayed in the list decreased their level of control and at the same time the companies that entered the list had a lower control level on average than the ones that exited. The study illustrated surprising commonalities between Paris and Frankfort stock exchanges. In both countries, more than half of the 250 largest traded companies were patrimonial, and their number increased substantially from 1993 to 1998. In both countries, the patrimonial companies were concentrated in the lower range of capitalisation. The analysis of the presence of patrimonial firms by sector evidenced some differences, but the main difference was found in the levels of control, overall lower and most importantly decreasing in Germany, contrarily to France. One aim in the future would be to compare the French and German data more in depth. But then this only should be the first step of a broader international project where the key economic countries can be compared on the basis of the design of the French study. It also would be interesting to analyse the largest non-listed German companies and compare these to the listed ones. But there will be even more severe problems in getting the relevant data of this group of companies.

© INSEAD, October 2002.

16

References Albach, H. & W. Freund (1989): Generationswechsel und Unternehmenskontinuität: Chancen, Risiken, Maßnahmen; Gütersloh Astrachan, J. & S.Bowen (2001): Editor´s Notes In: FBR XIV, 2, p.v – vii Astrachan, J., Klein, S. & K. Smyrnios (2002): The F-PEC Scale of Family Influence: A Proposal for Solving the Family Business Definition Problem In: FBR XV (1), p. 45-58 Ballarini, K. & Keese, D. (1997): Strukturmerkmale kleiner Familienunternehmen anhand regionalisierter Daten In: Internationales Gewerbearchiv, 1997, 4, p. 251 - 267 Blondel, C., Rowell, N. & L.Van der Heyden (2002): Prevalence of Patrimonial firms on Paris Stock Exchange: Analysis of the SBF 250 in 1993 and 1998, INSEAD working paper Commerzbank AG (Hrsg.): wer gehört zu wem – Beteiligungsverhältnisse in Deutschland 18. Aufl. 1994, 19. Aufl. 1997, 20. Aufl. 2000 Faccio, M. & L. Lang (2000): The Separtion of Ownership and Control: An Analysis of Ultimate Ownership in Western European Corporations, Working Paper Goehler, A.W.(1993): Der Erfolg grosser Familienunternehmen im fortgeschrittenen Marktlebenszyklus – dargestellt am Beispiel der deutschen Brauwirtschaft, Dissertation HSG St.Gallen 1993 Klein, S. (2000a): Familienunternehmen: Theoretische und empirische Grundlagen; Wiesbaden Klein, S. (2000b): Family Businesses in Germany: Significance and Structure In: FBR XIII, 3, p. 157 - 181 La Porta, R., Lopez-de-Silanes, F. & A. Shleifer (1999): Corporate Ownership around the World In: The Journal of Finance, LIV, 2 Liedtke, R. (1994): Wem gehört die Republik? Die Konzerne und ihre Verflechtungen, Frankfurt Pross, H. (1965): Manager und Aktionäre in Deutschland, Frankfurt Schreyögg, G. & H. Steinmann (1981): Zur Trennung von Eigentum und Verfügungsgewalt – Eine empirische Analyse der Beteiligungsverhältnisse in deutschen Großunternehmen In: ZfB 51. Jg, Heft 6, p. 533 - 557 Schröer, E. & W. Freund (1999): Neue Entwicklungen auf dem Markt für die Übertragung mittelständischer Unternehmen, IfM- Materialien Nr. 136 Shanker, M. & J. Astrachan (1996): Myths and Realities: Family Businesses´ Contribution to the US Economy- A Framework for Assessing Family Business Statistics In: FBR IX, 2, p. 107 - 119 © INSEAD, October 2002.

17

Westhead, P., Cowling, D. & D. Storey (1997): The Management and Performance of Unquoted Family Companies in the United Kingdom, Coventry Wieselhuber, N. & J. Spannagl (1988): Situation und Zukunftsperspektiven von InhaberUnternehmungen in der Bundesrepublik Deutschland, München Witte, E. (1981): Der Einfluß der Anteilseigner auf die Unternehmenspolitik In: ZfB 51.Jg, Heft 8, p. 733 - 779

© INSEAD, October 2002.

18

Appendix Appendix 1: Top 250 1993 and 1998 Ownership Type year

Ownership type

Number

%

Valid %

1993 Valid

patrimonial single corporate state mutual spread Total System

90 87 19 9 27 232 18 250 127 51 16 7 45 246 4 250

36,0 34,8 7,6 3,6 10,8 92,8 7,2 100,0 50,8 20,4 6,4 2,8 18,0 98,4 1,6 100,0

38,8 37,5 8,2 3,9 11,6 100,0

Missing Total 1998 Valid

Missing Total

patrimonial single corporate state mutual spread Total System

Cumulative % 38,8 76,3 84,5 88,4 100,0

51,6 20,7 6,5 2,8 18,3 100,0

51,6 72,4 78,9 81,7 100,0

Valid Percentage 24,8 44,4 11,1 5,1 14,5 100,0

Cumulative Percentage 24,8 69,2 80,3 85,5 100,0

39,2 26,7 9,2 3,3 21,7 100,0

39,2 65,8 75,0 78,3 100,0

Appendix 2: Top 120 1993 and 1998 Ownership Type Ownership type year

Ownership type

Number

Percent

1993 Valid

patrimonial single corporate state mutual spread Total System

29 52 13 6 17 117 3 120 47 32 11 4 26 120

24,2 43,3 10,8 5,0 14,2 97,5 2,5 100,0 39,2 26,7 9,2 3,3 21,7 100,0

Missing Total 1998 Valid

patrimonial single corporate state mutual spread Total

© INSEAD, October 2002.

19

Appendix 3: Top 20 1993 and 1998 Ownership Types year

Ownership type

1993

patrimonial single corporate state mutual spread Total patrimonial single corporate state spread Total

1998

© INSEAD, October 2002.

Number Percentag Valid e Percentage 3 15,0 15,0 9 45,0 45,0 1 5,0 5,0 2 10,0 10,0 5 25,0 25,0 20 100,0 100,0 3 15,0 15,0 4 20,0 20,0 3 15,0 15,0 10 50,0 50,0 20 100,0 100,0

Cumulative Percentage 15,0 60,0 65,0 75,0 100,0 15,0 35,0 50,0 100,0

20

Appendix 4: Sector analysis 1993 and 1998 of patrimonial firms

year

Sector

1993 luxury media pharmacy property retail services-general servicesleisure/tourism transport others automobile construction energy/utilities finance/insurance f.m.consumer goods general industry hi technology holding Total 1998

luxury media pharmacy property retail services-general servicesleisure/tourism others automobile construction energy/utilities finance/insurance f.m.consumer goods general industry hi technology holding Total

© INSEAD, October 2002.

Number Percentage 1 3 1 6 3 9 1 1

1,1 3,3 1,1 6,7 3,3 10,0 1,1 1,1

Valid Percentage 1,1 3,3 1,1 6,7 3,3 10,0 1,1 1,1

Cumulative Percentage 1,1 4,4 5,6 12,2 15,6 25,6 26,7 27,8

2 2 3 10 5 7 9

2,2 2,2 3,3 11,1 5,6 7,8 10,0

2,2 2,2 3,3 11,1 5,6 7,8 10,0

30,0 32,2 35,6 46,7 52,2 60,0 70,0

19 5 3 90

21,1 5,6 3,3 100,0

21,1 5,6 3,3 100,0

91,1 96,7 100,0

4 3 6 4 12 6 2

3,1 2,4 4,7 3,1 9,4 4,7 1,6

3,1 2,4 4,7 3,1 9,4 4,7 1,6

3,1 5,5 10,2 13,4 22,8 27,6 29,1

15 6 8 2 5 16

11,8 4,7 6,3 1,6 3,9 12,6

11,8 4,7 6,3 1,6 3,9 12,6

40,9 45,7 52,0 53,5 57,5 70,1

27 3 8 127

21,3 2,4 6,3 100,0

21,3 2,4 6,3 100,0

91,3 93,7 100,0

21

Appendix 5: Sector analysis 1993 and 1998 of the top 250 year

Sector

1993 luxury media pharmacy property retail services-general servicesleisure/tourism transport others automobile construction energy/utilities finance/insurance f.m.consumer goods general industry hi technology holding Total 1998 luxury media pharmacy property retail services-general servicesleisure/tourism transport others automobile construction energy/utilities finance/insurance f.m.consumer goods general industry hi technology holding 18 oder Total © INSEAD, October 2002.

Number

Percentage 1,2 2,0 ,4 6,0 2,4 6,0 ,8 1,2

Valid Percentage 1,2 2,0 ,4 6,0 2,4 6,0 ,8 1,2

Cumulative Percentage 1,2 3,2 3,6 9,6 12,0 18,0 18,8 20,0

3 5 1 15 6 15 2 3 5 4 11 16 26 50 20 48 9 11 250

2,0 1,6 4,4 6,4 10,4 20,0 8,0 19,2 3,6 4,4 100,0

2,0 1,6 4,4 6,4 10,4 20,0 8,0 19,2 3,6 4,4 100,0

22,0 23,6 28,0 34,4 44,8 64,8 72,8 92,0 95,6 100,0

1 4 5 14 5 22 7 3

,4 1,6 2,0 5,6 2,0 8,8 2,8 1,2

,4 1,6 2,0 5,6 2,0 8,8 2,8 1,2

,4 2,0 4,0 9,6 11,6 20,4 23,2 24,4

2 22 11 14 10 29 28 48 9 15 1 250

,8 8,8 4,4 5,6 4,0 11,6 11,2 19,2 3,6 6,0 ,4 100,0

,8 8,8 4,4 5,6 4,0 11,6 11,2 19,2 3,6 6,0 ,4 100,0

25,2 34,0 38,8 44,4 48,4 60,0 71,2 90,4 94,0 100,0 34,4

22

Appendix 6: Level of ownership of patrimonial firms 1993 and 1998 level of ownership

year

Number Percentage

1993

Valid

1998

Valid

absolute majority minority marginal Total absolute majority minority marginal Total System

Missing Total

33 36 17 4 90 24 53 35 14 126 1 127

Valid Cumulativ Percentage e Percentage 36,7 36,7 36,7 40,0 40,0 76,7 18,9 18,9 95,6 4,4 4,4 100,0 100,0 100,0 18,9 19,0 19,0 41,7 42,1 61,1 27,6 27,8 88,9 11,0 11,1 100,0 99,2 100,0 ,8 100,0

Appendix 7: Development of control level of 44 patrimonial firms that stayed in the list from 1993 to 1998 Diff Contr Lev

Number Percentage -2 -1 0 1 2 Total

7 11 18 6 1 43

© INSEAD, October 2002.

16,3 25,6 41,9 14,0 2,3 100,0

Valid Percentage 16,3 25,6 41,9 14,0 2,3 100,0

Cumulative Percentage 16,3 41,9 83,7 97,7 100,0

23