Does family management and family-ownership affect firm performance?
A study of 449 European firms during the period 2005-2014.
Amsterdam Business School
Name Josca van Walsum
Number 10210326
BSc in Business Economics Specialization Finance
Supervisor Dr. I.J. Naaborg Completion 15-12-2015
ECTS 12
Abstract
This thesis analyzes the effect of family management and family ownership on firm performance, among firms from the Europe STOXX 600 over the period 2005-2014. To measure firm performance, Tobin’s q and ROA are used. Existing research provides mixed outcomes and is done mainly on firms in the United States. These results do not necessarily need to apply to the European market due to legal and constitutional differences between both markets. The research in this paper shows that firms with one or more family members on the executive management or board of directors, perform significantly better than other firms. The same relation holds for firms with a family-ownership of at least 10%. This is consistent with the hypotheses that the presence of family management and family-ownership creates significant differences in firm performance over non-family firms in Europe. The results show that the benefits of family management and family-ownership outweigh the costs and suggest that firms managed or owned by family members outperform firms managed or owned by outsiders.
2 Table of Contents
1. Introduction……… 3
2. Literature review………...……. 4
2.1. Family management and firm performance………. 4
2.2. Family-ownership and firm performance………. 6
2.3. Conclusion……….... 7
3. Methodology & Data………... 8
3.1. Methodology & hypothesis……….. 8
3.1.1. Hypothesis………. 8
3.1.2. Methodology……….. 8
3.1.3. Measuring firm performance………..…... 9
3.1.4. Definitions of family management and family-ownership….……... 11
3.1.5. Control variables……… 10
3.1.6. Robustness check………....10
3.2. Data & descriptive statistics ………...…………. 11
3.2.1. Sample………..…………. 11
3.2.2. Descriptive statistics……….. 14
4. Empirical results……….. 16
4.1. Model results………..16
4.2. Robustness check ………..18
5. Conclusion & discussion………. 20
3 1. Introduction
In October 2014 McKinsey released a report on founder-and family-owned businesses, stating that family businesses are stronger and more vital around the world than ever before. According to McKinsey’s report 40% of the major listed companies in Europe are founder-or family-controlled. This raises questions on the effect of family members owning or managing a firm, on firm performance. Several studies have been done with mixed outcomes. Anderson & Reeb (2003) and Maury (2006) conclude that a firm performs better when it is family founded. Morck (1988) concludes the opposite, stating that entrenched family firms have objectives other than creating public shareholder value. Holderness & Sheehan (1988) find that family firms have a lower Tobin’s q than non-family firms. They conclude that family managed firms are often affiliated with unqualified family members participating in the management of the firm, instead of non-family members with appropriate qualifications. Morck et. al. (1988) confirm this in their research, stating that family-ownership in the US results into lower firm performance due to managerial entrenchment. Maury (2006) on the other hand, finds that firms actively managed by founding family members experience higher performance due to long-term investment horizons and their association with the family heritage.
This research will measure the effect of family managed and family-owned firms on firm performance specifically in Europe, as opposed to existing research which mainly focuses on the US (Morck et al., 1988, Anderson & Reeb, 2003, Villalonga & Amit, 2006, Miller et al., 2007) or Asia (Fan & Wong, 2002 and Lemmon & Lins, 2003). There are some legal and constitutional differences between these markets and the European market such as: different product regulations, different licensing requirements (Hoekman, 2015) and differences in labor laws and investor protection (La Porta et al., 1998). Because of these differences the results of these existing studies do not necessarily apply to the European market.
Maury (2006) does cover the western European market in his research. However, his research looks at different forms of family control, such as family members in the management or family members holding at least one of the two top officer positions, while this research looks at the effect of family-ownership as well. Anderson & Reeb (2003) mainly focus on family firms in general and also do a sub study on the effect of a family CEO. However they do not research the specific effects of family management and family-ownership separately.
This thesis will research the separate effects of family-ownership and family management. A firm is considered family managed when one or more family members are on the executive management or board of directors. A firm is considered family-owned when a family owns
4 10% or more of the firms equity. Villalonga & Amit (2006) research the effect of family-ownership, management and control on firm value among 508 listed Fortune-500 firms over the years 1994-2000. According to them, the best way to understand if and when family firms perform better than non-family firms, is to make a distinction among the two fundamental elements of management and ownership in the definition of family-firms. It will be tested whether these two different forms of family businesses, have different effects on firm performance. The sample used for this study consists of firms included in the Europe Stoxx 600, which represents large, mid and small capitalization companies across 18 European countries. Two regressions will be done to answer the main question on the relation between family firms and firm performance, where firm performance is measured by the firm’s Tobin’s q and its ROA. The first regression is on the relation between family management and firm performance and the second regression is on the relation between family-ownership and firm performance.
Chapter two summarizes the main theories in existing literature on the relationship between family firms and firm performance. Chapter three describes the empirical methodology and outlines the data of the research. Chapter four shows the regression results and is followed by a final conclusion in Chapter five.
2. Literature review
2.1. Family management and firm performance
In this research a firm is considered family managed when one or more family members holds a position in the executive management or board of directors. Hermalin and Weisbach (1991) investigate the effect of board composition and direct incentives on firm performance among 142 firms from the New York Stock Exchange in the years 1971, 1974, 1977, 1980 and 1983. They state that large management-ownership is often a characteristic of family firms, causing top management positions to be filled by family members instead of qualified managers. Morck et. All. (1998) study the so called ‘Canadian disease’ among 500 Canadian firms in the year 1988. They conclude that firms owned by families experience limited growth due to inherited control. They state that old wealth may entrench management, which results in a lack of entrepreneurship and creativity. Shleifer and Vishny (1997) research the importance of ownership concentrations in corporate governance systems around the world. According to them, managers who are entrenched on their job and no longer qualified to run the firm but resist being replaced, impose great costs. Morck, Schleifer & Vishny (1988) investigate the relation between management ownership and market valuation among firms in a 1980
cross-5 section of 371 Fortune 500 firms. They find a distinction between older and younger firms. Their research demonstrates that when founding family members take top positions at a young firm this has a positive impact on the Tobin’s q, whereas for older firms this has a negative effect.
Besides the agency problem family firms might be subject to excessive compensation, related-party transactions or special dividends, all at the cost of the firm’s wealth (Anderson and Reeb, 2003).
Finally, family managed firms may perform worse than other firms during a financial crisis. Lins, Volpin, and Wagner (2013) examine whether and how family control affects firm value and corporate decisions during the 2008-2009 financial crisis among 35 countries. They find that family-controlled firms underperform significantly by cutting relatively more in their investments compared to other firms. They state that survival of the family’s economic interests becomes a priority at the expense of the use of firm resources and outside shareholders.
Family managed firms also experience advantages compared to non-family firms. Anderson and Reeb (2003) state that family businesses may enjoy reputation benefits from the family’s lasting presence in the management of the firm and its effect on external parties, such as suppliers of capital. They state that these suppliers are more likely to deal with family managed firms for extended periods of time since they often remain the same governing bodies for longer, relative to non-family firms where managers and directors shift on a more continuous basis. Therefore, families maintaining a long-term presence in the firm may result in lower costs of debt according to Anderson, Mansi & Reeb (2003), who investigate the relation between family-ownership and the agency cost of debt among S&P 500 firms in the period 1993-1998.
Additionally, firms actively managed by founding family members can experience higher performance due to long-term investment horizons. Maury’s (2006) research on the relationship between family firms in Western Europe in 1998 and their performance, states that one of the main reasons family firms perform better is due to active management. This means that one or more members of the founding family fill a position on the executive management or board of directors. According to Maury (2006), family management reduces the separation of ownership and control, resulting in better firm performance. This creates long-term investment decisions benefiting the firm.
6 2.2. Family-ownership and firm performance
A firm is considered family-owned when the founding-family owns at least 10% of the firms shares. Morck and Young (2003) research the agency problems in large family business groups in the United States. According to them, family-owned firms give rise to an extra set of agency problems compared to widely held firms. They state that the owning family has an incentive to act in their own interest instead of the shareholder’s interest inducing an agency problem. This may lead to suboptimal policies and result into poor firm performance relative to non-family firms.
The owning families are also indicated to be well-informed shareholders, who possess superior information over the outside shareholders (Anderson and Reeb, 2003). Chan, Chen and Hillary (2010) study inside trading and family firms in the S&P 1500 over the years 1997-2006. They find that the stock trades made by founding CEO’s are higher than those made by CEO’s in non-family firms and that the trade forecasts of the firm’s stock returns are better. Anderson, Reeb and Zhao (2012) do research on the relation between organization structure and the information content of short sales among the 2000 largest industrial firms in the US, using data from 2005 through 2007. They conclude that informed trading can undermine investment confidence in the market and might limit capital market development. They also state that family members can be a source of information leakage to outside investors.
Besides the agency problems family-owned firms experience, they also have some benefits over non-family firms. Anderson and Reeb (2003) investigate the relation between founding-family-ownership and firm performance in the S&P500 during the period 1992-1999. They use profitability-based measures on firm performance (ROA) and find that family-owned firms perform significantly better than non-family firms. An additional analysis using Tobin’s q leads to a similar conclusion: they find a 10% higher Tobin’s q in family-owned firms relative to non-family firms. The explanation of Anderson and Reeb (2003) for this difference in performance is that family members understand the business better and see themselves as representatives of the firm. They state that the family wealth is directly linked to the value of the firm and thus creates an incentive to aim for value maximization of the firm. Families will have an incentive to reduce the agency problem by monitoring managers and minimize free-riding costs.
Founding-families also tend to maintain a stake in the firm for a longer period of time, passing the firm on to next generations. As a result, family members have longer horizons than other
7 investors and are therefore more willing to invest in long-term projects. Stein(1989) studies myopic corporate behavior and states that long-term stockholders will discourage myopic behavior and therefore will be less likely to renounce good investments to boost current earnings. This is also mentioned by James (1999) who examines the loss of agency costs when an owner also functions as manager in family firms. He finds that the extended horizon characteristics of family businesses create an incentive to invest according to the market rule (positive NPV projects). He therefore states that family-owned firms invest more efficiently than non-family firms because the business is intended to succeed also when owned by the descendants. Casson (1999) investigates in his paper whether family-ownership requires one special theory, or if its strengths and weaknesses vary across family-owned firms. According to him, founding families view their firms not as much as wealth to spend during their lifetime but more as an asset to pass on to the next generation.
Habbershon and Williams (1999) use a Resource Based View (RBV), to provide a theoretical framework explaining the competitive advantage of family-owned firms from a qualitative point of view. Besides looking at the quantitative advantages of family-owned firms (such as ROA) their research states that firms should be compared at behavioral level as well. They discuss the ‘bundle’ of resources that are a result of family involvement in a business, as the ‘familiness’ of a firm. This includes factors such as: a family-orientated working environment, more efficient communication through ‘family language’, more efficient informal-decision making channels, more commitment to their missions and less managerial politics (Habbershon and Williams, 1999).
While informed trading in family-owned firms is mentioned as a disadvantage, as it undermines investment confidence among the outside shareholders, it can be seen as a benefit as well. According to Anderson, Reeb and Zhao (2012), informed trading can result in better price discovery, facilitating market efficiency and limiting idiosyncratic risks on corporate investment policy.
2.3. Conclusion
Existing literature extensively discusses on the different corporate governance structures and especially the ones with family-ownership and management. The literature review shows that there are both potential costs and benefits of family-ownership and family management. The different empirical studies on the overall effect of family-ownership and management have provided various outcomes. Most studies agree that family management can be a major disadvantage when an unqualified family member is appointed a management position instead
8 of an outside manager with the appropriate qualifications. On the other hand a generally recognized benefit of family management is the creation of firm value because of the family’s long-term investment interests. When studying existing theory on family-ownership, most researchers agree on the costs of occurring agency problems. Since the owning family has an incentive to act in their own interest at the expense of other shareholders. A commonly named benefit of family-ownership is the incentive for the owning family to maximize firm value, due to a direct link between family wealth and value of the firm. To analyze whether the benefits outweigh the costs or the costs outweigh the benefits, a more empirical research is needed which will be done in the next chapters.
3. Methodology & Data
3.1. Methodology & Hypothesis 3.1.1. Hypothesis
This research is focused on testing whether family managed and family-owned firms outperform or underperform compared to the rest of the market. The research question is therefore as follows: Does family management and family-ownership affect firm performance? Existing research provides very different results making it difficult to predict an outcome. Anderson & Reeb (2003) and Villalonga & Amit (2006) find significant positive relations while Holderness & Sheehan (1988) and Morck et al. (1988) both find a negative impact. Also most existing research has been done in the US, making it even more difficult to predict the outcome of this thesis for the European market since both markets have legal and constitutional differences (La Porta et al. 1998). Therefore the hypotheses that will be tested in this research are:
(1)
H0 =There is no effect of a firm being family managed on firm performance: β1 = 0
H1 =There is a significant effect of a firm being family managed on firm performance: β1 ≠ 0
(2)
H0 =There is no effect of a firm being family-owned on firm performance: β1 = 0
H1 =There is a significant effect of a firm being family-owned on firm performance: β1 ≠ 0
3.1.2. Methodology
The model used follows the method of Anderson and Reeb (2003). The goal of this research is to find the separate effects of family-ownership and family management on firm performance.
9 To measure this, a multivariate analysis will be done using the following regression equations:
(1a)
Tobin’s q = β0 + β1 (family managed) + β2 (firm size) + β3 (firm risk) + β4 (firm age) + β5
(growth) + β6 (industry) + β7 (crisis) + ε
(1b)
ROA = β0 + β1 (family managed) + β2 (firm size) + β3 (firm risk) + β4 (firm age) + β5 (growth)
+ β6 (industry) + β7 (crisis) + ε
(2a)
Tobin’s q = β0 + β1 (family owns>10%) + β2 (firm size) + β3 (firm risk) + β4 (firm age) + β5
(growth) + β6 (industry) + β7 (crisis) + ε
(2b)
ROA = β0 + β1 (family owns>10%) + β2 (firm size) + β3 (firm risk) + β4 (firm age) + β5
(growth) + β6 (industry) + β7 (crisis) + ε
3.1.3. Measuring firm performance
The dependent variable in this research is firm performance. As in most articles this is measured by Tobin’s q, the ratio of a firm’s market value to the replacement costs of its assets. Return on assets (ROA) will be used as second measure of firm performance. ROA is calculated by dividing the net income by total assets (Anderson & Reeb 2003).
3.1.4. Definitions of family management and family-ownership
Previous articles use different definitions of a family firm. Maury (2006) uses a family dummy that is equal to one when at least 10% of the voting rights is owned by one or more named individuals or families. Villalonga & Amit (2006) define a family firm as a firm whose founder or a family member is an officer, a director, or the owner of at least 5% of the firm’s equity.
This paper uses a dummy for family management and one for family-ownership as the main explanatory variables. They will be the primary indicators of family participation. A firm is considered family managed when at least one founding-family member is active in executive management or the board of directors (Maury 2006). Founding-family is hereby defined as an
10 individual or entire family with a similar last name as the founder of the firm. As a measurement of family-ownership, the fractional equity of the founding family is used, following the definiton of Anderson & Reeb (2003). This means a business is considered family-owned when at least 10% of its shares are in hands of one or more named individual founding-family members or families as a whole. The use of a 10% treshold is confirmed by Porta, Lopez‐de‐silanes & Shleifer (1999), who study ownership structures of large corporations in 27 wealthy economies to identify the ultimate controlling shareholders of these firms. They state that a 10% cutoff creates a significant threshold, where the direct investor is able to influence the management of an enterprise. Some existing papers such as Andres (2008) use different criteria with a cutoff at 20%. The effect of a higher ownership threshold will be tested later in this research as a robustness check. In this robustness-check the percentage treshold for family-ownership will be changed to at least 30%, to see if this increases the measured effect on firm performance. A second robust-check will be done changing the family management definition to see if this increases the measured effect on firm performance. A firm is considered family managed when a founding family member fills the CEO position at the company.
3.1.5. Control variables
Six control variables are used to control for industry and firm specific characteristics. These are mostly the same as used by Anderson & Reeb (2003): firm size, risk, age, growth, industry and a crisis dummy.
Firm size is defined as the natural log of its total assets. Firm risk is measured by the standard deviation of the stock returns (historical volatility). Firm age is given by the natural log of the number of years since incorporation. Firm growth opportunities are measured as is done by Maury (2006), by taking the average growth in net sales. To sort each firm per industry, the industry classification benchmark (ICB) is used. This provides 36 different categories, without the utilities and bank sectors which are excluded from this research. The final variable controls for the economic crisis and is a dummy that equals one for the years 2008-2011 and zero otherwise.
3.1.6. Robustness check
In addition to the main research, two Robustness checks will be done using two different definitions for family management and family-ownership. For family management a dummy variable will be used that is equal to one when a family member is CEO of the firm. For the
11 ownership dummy a threshold percentage of 30% will be used to define a family-owned firm instead of the 10% cutoff used in regression (2). The following regression equations will be used:
(3a)
Tobin’s q = β0 + β1 (family CEO) + β2 (firm size) + β3 (firm risk) + β4 (firm age) + β5
(growth) + β6 (industry) + β7 (crisis) + ε
(3b)
ROA = β0 + β1 (family CEO) + β2 (firm size) + β3 (firm risk) + β4 (firm age) + β5 (growth) +
β6 (industry) + β7 (crisis) + ε
(4a)
Tobin’s q = β0 + β1 (family owns>30%) + β2 (firm size) + β3 (firm risk) + β4 (firm age) + β5
(growth) + β6 (industry) + β7 (crisis) + ε
(4b)
ROA = β0 + β1 (family owns>30%) + β2 (firm size) + β3 (firm risk) + β4 (firm age) + β5
(growth) + β6 (industry) + β7 (crisis) + ε
3.2. Data and descriptive statistics 3.2.1. Sample
For this thesis the sample is formed by firms from the Europe Stoxx 600, which represents large, mid and small capitalization companies across 18 European countries. As is done by Anderson & Reeb, public utilities are excluded because of government regulations that can potentially affect firm performance. Because it is difficult to calculate Tobin’s q for banks, they are also excluded, leaving a sample of 449 firms. The data are observed over a period of ten years: 2005 – 2014, providing 4,490 firm years. Since the research uses data from 2005-2014, all firms incorporated after 2005 are excluded from the sample.
DataStream is used to retrieve the needed financial data of all firms, such as total market value and the return on assets. To get data on shareholder characteristics, shareholder names and the division of management, the Amadeus and Osiris databases from Bureau van Dijk are used.
As is shown below in table I and table II, 22% of all businesses in the sample are family managed firms and 10% are family-owned.
12 Table I
Family and non-family firms per industry (n=449 firms, t=10 years)
Number and percentage of family managed and family-owned firms per country. Family managed firms refer to the firms where the founding family member fills one or more positions on the board of directors or the executive management. Family-owned firms refer to the firms where a family member owns at least 10% of its equity. The percentages of family managed firms and family-owned firms are obtained by dividing the number by the total number of firms.
Country Total firms Family
managed Family-owned %Family managed %Family-owned Austria 4 0 0 0 0 Belgium 11 5 1 45 9 Switzerland 43 17 9 40 21 Germany 52 12 6 23 12 Denmark 14 3 3 21 21 Spain 15 4 0 27 0 Finland 12 0 0 0 0 France 69 24 7 35 10 Great-Britain 133 17 10 13 8 Greece 2 0 0 0 0 Ireland 5 2 1 40 20 Italy 15 5 3 33 20 Luxembourg 3 1 0 33 0 The Netherlands 28 4 3 14 11 Norway 10 0 0 0 0 Portugal 2 1 1 50 50 Sweden 31 4 0 13 0 Total 449 99 44 22 10
13 Table II
Family managed firms and family-owned firms per industry (n=449 firms, t =10 years)
Number and percentage of family managed and family-owned firms per sector according to the Industry Classification Benchmark (ICB). Family managed firms refer to the firms where the founding family member fills one or more positions on the board of directors or the executive management. Family-owned firms refer to the firms where a family member owns at least 10% of its equity. The percentages of family managed firms and family-owned firms are obtained by dividing the number by the total number of firms.
Sector nr. Industry Total firms Family managed firms Family-owned firms % family managed in industry % Family-owned in industry
0530 Oil & gas producers 12 2 0 17 0
0570 Oil equipment services & distr. 8 0 1 0 13
0580 Alternative energy 2 1 0 50 0
1350 Chemicals 24 5 2 21 8
1730 Forestry & paper 2 0 0 0 0
1750 Industrial metals & mining 3 2 0 67 0
1770 Mining 7 1 0 14 0
2350 Construction & Materials 17 3 2 18 12
2710 Aerospace & Defense 11 0 0 0 0
2720 General industrials 8 0 0 0 0
2730 Electronic & electrical equipment 6 0 0 0 0
2750 Industrial engineering 26 3 1 12 4
2770 Industrial transportation 15 3 2 20 13
2790 Support services 25 4 2 16 8
3350 Automobiles & parts 14 5 2 36 14
3530 Beverages 9 5 2 56 22
3570 Food producers 13 3 1 23 8
3720 Household goods & Home constr. 11 4 2 36 18
3740 Leisure goods 1 0 0 0 0
3760 Personal goods 12 7 4 58 33
3780 Tobacco 3 0 0 0 0
4530 Health care equipment & services 14 4 3 29 21
4570 Pharmaceuticals & Biotechnology 21 6 2 29 10
5330 Food & drug retailers 14 5 3 36 21
5370 General retailers 11 4 0 36 0
5550 Media 23 9 2 39 9
5750 Travel & Leisure 20 6 4 30 20
6530 Fixed line telecommunications 11 0 0 0 0
6570 Mobile telecommunications 7 0 0 0 0
8530 Nonlife insurance 20 1 1 5 5
8570 Life insurance 10 1 0 10 0
8630 Real estate investment & services 7 0 0 0 0
8670 Real estate investment trusts 16 1 0 6 0
8770 Financial services 24 10 4 42 17
9530 Software & computer services 10 5 2 50 20
9570 Technology hardware & equipm. 11 2 1 18 9
14 3.2.2. Descriptive statistics
Table III shows the descriptive statistics of the sample of firms divided in table III A,B and C. Table III A presents the mean, median, standard deviation, minimum and maximum values of the variables. Table III B gives the test of means results between family managed and non-family managed firms and the test of means results between owned and non- non-family-owned firms. Table III C presents the correlations between the dependent, independent and control variables.
Table III A: Summary statistics
Provides the summary statistics of the used sample, which consists of 449 firms and is measured over a period of 10 years, from 2005-2014. This gives a total of 4,490 firm years. The variable Ln(total assets) is measured as the natural log of the total book value of total assets and controls for firm size. Historical volatility is the standard deviation of the stock returns and controls for firm risk. Age is the number of years since the firm has been incorporated and Ln(age) is its natural log. Growth controls for the growth opportunities of the firm and is measured by taking the average growth percentage in net sales. Crisis is a dummy variable that equals one for the years 2008-2011 and zero otherwise. Tobin’s q is calculated by the ratio of the firm’s market value to the replacement costs of its assets and measures firm performance. ROA is found by dividing net income by total assets and is a second way to measure firm performance. The variable %family-ownership is measured in the percentage of equity the founding family holds. Family owns>10% is a dummy variable that equals one when the founding family holds at least 10% of the firm’s equity. Family management is a dummy variable equal to one when one or more members of the family are present in the board of directors or excecutive management.
Variable mean Median S.D. Min Max
Ln(total assets) 15.96716 15.82754 1.702533 10.78792 21.59574 Historical volatility 26.08641 24.95 7.776761 5.65 63.1 Firm age 54.2448 40 42.69513 10 325 Growth% 9.662977 6.33 25.55673 -91.26 549.29 Crisis .4 0 .4899503 0 1 Tobin’s q 1.162030 .7971559 1.232362 .0060760 9.890254 ROA 5.981799 5.005533 8.242905 -80.92842 67.86717 % Family-ownership 4.161382 0 13.3045 0 89.2 Family Owns > 10% .0967033 0 .2955858 0 1 Family Managed .2217295 0 .4154561 0 1
Table III A shows that the average Tobin’s q of the sample is 1.16 and the average return on assets is 5.98. Of the sample, 22.17% is family managed which means that at least one of the founding family members is active in the board of directors or excecutive management. The table also shows that 9,67% of all firms has family-ownership of at least 10%. The highest
15 percentage of shares that a family owns of its firm is 89.2%; an abundant majority ownership. The average firm age is 54 which indicates the sample firms are well established but this is partly due to removal of businesses that were incorporated after 2005. The average annual growth of sales in this sample is 9.66% and 40% of the observations of variables were measured during the crisis.
Table III B: Test of means
Table III B shows a difference of means test between family managed firms and non-family managed firms and between family-owned firms and non-family-owned firms. Hereby the t-statistic of the difference of means are shown at 10%, 5% and 1% significance level: *, **, *** respectively.
Family managed Non-family
managed
t-statistic Family-owned
Non-family-owned t-statistic Ln(total assets) 15.6641 16.0131 5.8106*** 15.5361 15.9887 5.4053*** Historical volatility 26.0355 25.9588 -0.2671 24.7578 26.1200 3.4040*** Firm age 3.7468 3.6773 -2.3947** 3.7171 3.6919 -0.6364 Growth % 11.6407 9.0889 -2.7983*** 10.2589 9.5823 -0.5153 Tobin’s q 1.6005 1.0615 -12.1114*** 1.9089 1.0972 -13.1675*** ROA 7.7618 5.6255 -7.2789*** 9.7745 5.6700 -10.1131***
Table III C: Correlation matrix
Provides the correlation matrix for the above mentioned dependent, independent and control variables. Beneath the correlation value, the significance is shown in parantheses.
(1) (2) (3) (4) (5) (6) (7) (8) (9) (1)Tobin’s q 1 (2)ROA 0.6396 (0.0000) 1 (3)Ln(total assets) -0.4023 (0.0000) -0.2444 (0.0000) 1 (4)Historical volatility -0.0609 (0.0001) -0.1680 (0.0000) -0.1408 (0.0000) 1 (5)Ln(age) -0.0828 (0.0000) -0.0128 (0.3997) 0.2459 (0.0000) -0.1439 (0.0000) 1 (6)Growth 0.0647 (0.0000) 0.0602 (0.0001) -0.0882 (0.0000) 0.0731 (0.0000) -0.0674 (0.0000) 1 (7)Crisis -0.1190 (0.0000) -0.0840 (0.0000) 0.0104 (0.4823) 0.1726 (0.0000) -0.0147 (1.0000) -0.0808 (0.0000) 1 (8)Family Owns>10 0.1958 (0.0000) 0.1498 (0.0000) -0.0807 (0.0000) 0.0528 (0.0007) 0.0097 (0.5245) 0.0081 (0.5926) -0.0000 (1.0000) 1 (9)Family Managed 0.1811 (0.0000) 0.1089 (0.0000) -0.0871 (0.0000) 0.0042 (0.7894) 0.0367 (0.0167) 0.0425 (0.0052) 0.0000 (1.0000) 0.5081 (0.0000) 1
16 Table III B presents the comparison of means test between family managed and family-owned firms and non-family firms. It shows that family managed and family-owned firms on average are smaller than non-family firms. Family managed firms appear to have an average Tobin’s q of 1.6 and family-owned firms have an average Tobin’s q of 1.9. For non-family firms this is 1.06, which is significantly lower. The same relation holds for the firm’s return on assets: family managed and family-owned firms have a significantly higher average ROA than non-family firms.
Finally table III C provides a correlation matrix of the key explanatory variables of this research. It shows that both the family-ownership dummy variable and the family management dummy variable have a positive relation with the measures of performance: Tobin’s q and ROA. The correlation between family management and family-ownership is 0.5081. This indicates that both variables move differently (otherwise their correlation would have been close to 1), which makes the seperation between family management and family-ownership a meaningful distinction for this research . In addition, both family dummies show a negative association with the natural log of total assets which measures firm size, indicating that family managed and family-owned firms might be relatively smaller firms.
4. Empirical results 4.1. Model results
This chapter contains the model results of the regressions performed to answer the main question of this research. The results of the multivariate analysis described in chapter 3.1, are shown below in table IV. First the results of regression 1 present the effect of the dummy
family managed on Tobin’s q (1a) and ROA (1b). Both show a positive effect of the dummy family managed on firm performance. The results suggest that Tobin’s q in family managed
firms is 26.43% higher compared to non-family firms (family managed coefficient(1a)/ mean Tobin’s q of non-family firms). The same way, the coefficients suggest that ROA is 16.74% higher in family managed firms than in non-family firms (family managed coefficient(1b)/ mean ROA of non-family firms). Regression (2a) and (2b) test for the effect on the dummy variable family owns>10%, and show a similar result. A firm being owned for at least 10% by family members has a positive effect on its Tobin’s q and ROA. The results show that a family-owned firm has 41.70% higher Tobin’s q than non-family firms (family-owned coefficient(2a)/ mean Tobin’s q of non-family firms). For ROA this effect is 33.21% higher (family-owned coefficient(2b)/ mean Tobin’s q of non-family firms). All measured effects are significant at the 1% level.
17 Concerning control variables, the size of a firm is found to have a negative effect on firm performance. Also firm risk and the dummy used for crises years have a negative relation to both Tobin’s q and ROA. For firm age and firm growth, a positive association is found between both variables and firm performance. These results are generally consistent with findings in existing literature.
Table IV
Regression results
This table shows the results of regressing firm performance on family-ownership. Two different models are used to measure the effect on both Tobin’s q and ROA, giving a total of 4 regressions. Tobin’s q is defined by the firm’s market value divided by its replacement costs of assets. ROA is calculated by dividing net income by total assets. The first column (1) reports the regression on the dummy variable family managed, which is equal to one when at least one of the founding family members is present in the board of directors or executive management. The second column (2) reports the regression for the family owns>10% dummy variable, this is equal to one when the family owns 10% or more of its firm’s shares. Ln(total assets) is measured by the natural log of a firm’s total assets to control for firm size. Historical volatility is the standard deviation of stock returns and controls for firm risk. To control for firm age the natural log of age is taken in the Ln(age) variable. Growth is measured by annual net sales growth percentage. A dummy variable crisis is used that is equal to one when the data are covering the years 2008-2011. Finally all regressions include dummies for the four-digit industry codes to account for industry-specific effects. The t-values are shown in parentheses and the significance is stated by *,**,*** at the 10%, 5%, 1% level, respectively.
Tobin’s q Tobin’s q ROA ROA
(1a) (2a) (1b) (2b) intercept 6.1663 (25.62)*** 6.205512 (25.93)*** 32.6313 (19.35)*** 29.3371 (12.63)*** Family managed 0.2791 (6.55)*** 0.9390 (3.15)*** Family owns>10% 0.4403 (7.74)*** 1.863 (4.67)*** Ln(total assets) -0.2664 (-22.34)*** -0.2688 (-22.66)*** -1.2122 (-14.51)*** -1.2187 (-14.66)*** Historical volatility -0.0196 (-7.97)*** -0.0185 (-7.56)*** -0.2262 (-13.14)*** -0.2210 (-12.86)*** Ln(age) 0.0062 (0.26) 0.0106 (0.45) 0.2980 (1.81)* 0.3075 (1.87)* Growth 0.0020 (2.52)** 0.0021 (2.69)*** 0.0281 (4.99)*** 0.0282 (5.03)*** crisis -0.2421 (-6.90)*** -0.2438 (-6.98)*** -0.5544 (-2.25)** -0.5623 (-2.29)** Observations 3935 3945 3939 3949 Adj. R-squared 0.2718 0.2754 0.1358 0.1282
18 4.2. Robustness checks
In this section more specific research is done. Instead of using a dummy for family management, a dummy that indicates a family CEO will be used. Also instead of using a dummy variable for family-ownership of 10% or higher, a dummy is used for ownership of at least 30%. This to test whether the measured effect mentioned above, increases when family-ownership increases. Apart from these different dummy definitions, the same models are used as in the previous chapter.
The results are shown below in table V. Regression (3a) and (3b) give the effect of the dummy family owns>30% on firm performance and suggest for both Tobin’s q and ROA, a slightly weaker positive relation than was measured for ownership of at least 10%. Tobin’s q is in this case for family-owned firms 28.9% higher than for non-family firms, as opposed to 41.7% in the previous section. ROA is in this case 20.9% higher for family-owned firms compared to non-family firms, whereas in the previous section this was 33.21%. The coefficient on Tobin’s q is hereby significant at the 1% level and the coefficient on ROA is significant at the 5% level. These results are in line with the research of Anderson & Reeb (2003), who find a nonlinearity between firm performance and family-ownership. Their results show that performance gains of family-ownership begin to taper off at 30.8% ownership. In other words, beyond this level performance begins to decline but is on average still better compared to non-family-owned firms. Research by Morck et al. (1988) also suggests that the relation between equity ownership and firm performance is nonlinear, due to incentive changes of the equity claimant as the percentage of ownership increases.
The second regression shows a negative effect of a family member being CEO of the firm, on both Tobin’s q (4a) and ROA(4b). Villalonga & Amit (2006) research the effect of family-ownership, management and control on firm value among 508 listed Fortune-500 firms over the years 1994-2000. They find two separate effects of a family CEO on firm performance. Their results show a positive effect when the founder himself is CEO of his own firm. However when the family members of a later generation (descendants of the founder) fill a CEO position the effect on firm performance is negative. They state this difference arises due to decreasing entrepreneurial and leadership skills among later generations. The average firm age of the sample used in this thesis is 54, which might indicate mostly descendant-CEO’s instead of founder-CEO’s, explaining the negative coefficient. The found coefficients for the family CEO dummy are not significant, however this can be because only 19 firms in this sample have a family member as CEO.
19 Table V
Regression results robustness checks
This table shows the results of regressing firm performance on family-ownership and management. Two different models are used to measure the effect on both Tobin’s q and ROA, giving a total of 4 regressions. Tobin’s q is defined by the firm’s market value divided by its replacement costs of assets. ROA is calculated by dividing net income by total assets. The first column (3) reports the regression results using the dummy family owns>30% which is defined as the founding family owning at least 30% of the firm’s equity. The second column (4) reports the regression for the family CEO dummy variable, this is equal to one when the CEO of the firm is a family member. Ln(total assets) is measured by the natural log of a firm’s total assets to control for firm size. Historical volatility is the standard deviation of stock returns and controls for firm risk. To control for firm age the natural log of age is taken in the Ln(age) variable. Growth is measured by annual net sales growth percentage. A dummy variable crisis is used that is equal to one when the data are covering the years 2008-2011. Finally all regressions include dummies for the four-digit industry codes to account for industry-specific effects. The t-values are shown in parentheses and the significance is stated by *,**,*** at the 10%, 5%, 1% level, respectively.
Tobin’s q Tobin’s q ROA ROA
(3a) (4a) (3b) (4b) intercept 6.2583 (18.85)*** 6.2975 (26.22)*** 32.9136 (19.18)*** 29.7707 (12.81)*** Family CEO -0.1496 (-1.74)* -0.1284 (-0.21) Family owns>30% 0.3073 (4.39)*** 1.1703 (2.39)** Ln(total assets) -0.2770 (-23.17)*** -0.2728 (-22.91)*** -1.2332 (-14.51)*** -1.2358 (-14.85)*** Historical volatility -0.0193 (-7.92)*** -0.0194 (-7.87)*** -0.2215 (-12.80)*** -0.2248 (-13.07)*** Ln(age) 0.0141 (0.60) 0.0202 (0.44) 0.3380 (2.03)** 0.3074 (1.87)* Growth 0.0020 (2.65)*** 0.0020 (2.60)*** 0.0257 (4.69)*** 0.0279 (4.97)*** crisis -0.2415 (-6.92)*** -0.2411 (-6.87)*** -0.5933 (-2.39)** -0.5504 (-2.24)** Observations 3981 3945 3985 3949 Adj. R-squared 0.2662 0.2679 0.1274 0.1346
20 5. Conclusion & discussion
This research examines the separate effects of family management and family-ownership on firm performance among firms from the Europe STOXX 600 index over the period 2005-2014. Similar research has been done before, however this is mainly focused on the US or only parts of Europe. Also existing papers presented mixed outcomes. This is due to the use of different definitions where each study examines a slightly different form of family involvement.
The first model which tests for the effect of family management, shows a significant positive relation with both Tobin’s q and ROA. The results present that for family managed firms, Tobin’s q is 26.43% higher and ROA is 16.74% higher than for non-family firms. Therefore it can be concluded that the benefits of having one or more family members in the executive management or board of directors, exceed the costs. This is in line with Maury (2006), who explains that family managers have long-term investment horizons and that family management reduces the separation between ownership and control, benefiting the firm. The second model tests for family-ownership and also shows a significant positive effect on firm performance. Anderson & Reeb (2003) have similar results in their research and explain that family wealth is directly linked to firm value, thus creating an incentive for family members to aim for value maximization of the firm and to reduce the agency problem. The results show that for family-owned firms Tobin’s q is 41.70% higher and that ROA is 33.21% higher than for non-family firms.
More specific research on firms that solely have a family CEO instead of family management gives a negative effect on firm performance. This is in line with Villalonga & Amit (2006), who conclude that descendant-CEOs destroy firm value.
Further testing shows that an increased family-ownership of at least 30% still has a positive effect on firm performance, although weaker than the effect found with ownership of 10% or more. Anderson & Reeb (2003) find the same result when investigating nonlinearities among increasing percentages of ownership. They conclude that the performance gains of family-ownership begin to taper off around the 30% cutoff, due to incentive changes of the equity claimant as the percentage of ownership increases.
The results of this thesis know some limitations. The decision whether an individual is defined as a family member of the founder or not is based on similar last names. However in older firms where several generations have passed, last names may have changed over time due to for example marriage. This means in this thesis, that when a family member of the founder meets the conditions for family-ownership or family management but doesn’t carry
21 the same last name as the founder, the firm will wrongfully be defined as non-family. Secondly this research treats Europe as a homogenous market, whilst there are some legal and constitutional differences among the various European countries. Also, the presence of family managed firms and family-owned firms might vary among different European countries. Therefore a suggestion for future research might be to do the analysis of family management and family-ownership per country. Additionally, this research finds a negative relation between a family CEO and firm performance. Existing literature explains that there is only a positive effect on firm performance when the founder himself is CEO, due to decreasing entrepreneurial and leadership skills among later generations (Villalonga & Amit, 2006). The average firm age of the sample used in this thesis is 54 which might indicate mostly descendant-CEO’s, explaining the negative result. However to fully understand and explain this outcome, more extensive research is needed on the different effects of founder-CEO’s and descendant-CEO’s on firm performance in Europe. Finally this paper doesn’t examine the distinction between different firm sizes. Therefore, a suggestion for future research might be to test for the effects on firm performance for small firms and large firms separately, to see whether the results found in this thesis still hold.
22 6. References
Anderson, R. C., Mansi, S. A., & Reeb, D. M. (2003). Founding family ownership and the agency cost of debt. Journal of Financial Economics, 68(2), 263-285.
doi:10.1016/S0304-405X(03)00067-9
Anderson, R. C., & Reeb, D. M. (2003). Founding‐ family ownership and firm performance: Evidence from the S&P 500. Journal of Finance, 58(3), 1301-1327. doi:10.1111/1540-6261.00567
Anderson, R. C., Reeb, D. M., & Zhao, W. (2012). Family‐Controlled firms and informed trading: Evidence from short sales. Journal of Finance, 67(1), 351-385.
doi:10.1111/j.1540-6261.2011.01714.x
Andres, C. (2008). Large shareholders and firm performance - an empirical examination of founding- family ownership. Journal of Corporate Finance, 14(4), 431-445.
doi:10.1016/j.jcorpfin.2008.05.003
Baek, J., Kang, J., & Suh Park, K. (2004). Corporate governance and firm value: Evidence from the korean financial crisis. Journal of Financial Economics, 71(2), 265-313. doi:10.1016/S0304-405X(03)00167-3
Bennedsen, M., Nielsen, K. M., Perez-Gonzalez, F., & Wolfenzon, D. (2007). Inside the family firm: The role of families in succession decisions and performance. The Quarterly
Journal of Economics, 122(2), 647-691.
Burkart, M., Panunzi, F., & Shleifer, A. (2003). Family firms. Journal of Finance, 58(5), 2167-2202. doi:10.1111/1540-6261.00601
Casson, M. (1999). The economics of the family firm. Scandinavian Economic History
Review, 47(1), 10-23. doi:10.1080/03585522.1999.10419802
Chan L.H.L., Chen T.Y. & Hilary G. (2010). Insider trading and family firms. Working paper, University of Hong Kong.
23 DeAngelo, H., & DeAngelo, L. (2000). Controlling stockholders and the disciplinary role of
corporate payout policy: A study of the times mirror company. Journal of Financial
Economics, 56(2), 153-207.
Edmans, A. (2009). Blockholder trading, market efficiency, and managerial myopia. Journal
of Finance, 64(6), 2481-2513. doi:10.1111/j.1540-6261.2009.01508.x
Habbershon, T. G., & Williams, M. L. (1999). A resource‐ based framework for assessing the strategic advantages of family firms. Family Business Review, 12(1), 1-25.
doi:10.1111/j.1741-6248.1999.00001.x
Hermalin, B., & Weisbach, M. S. (1991). The effects of board composition and direct incentives on firm performance. Financial Management; Financ.Manage., 20(4), 101-112.
Hoekman, B. (2015). Fostering transatlantic regulatory cooperation and gradual multilateralization.18(3), 609-624. doi:10.1093/jiel/jgv028
Holderness, C. G. (2009). The myth of diffuse ownership in the united states. Review of
Financial Studies, 22(4), 1377-1408. doi:10.1093/rfs/hhm069
Holderness, C. G., & Sheehan, D. P. (1988). The role of majority shareholders in publicly held corporations: An exploratory analysis. Journal of Financial Economics, 20, 317-346. doi:10.1016/0304-405X(88)90049-9
James, H. (1999). Owner as manager, extended horizons and the family firm. International
Journal of the Economics of Business, 6(1), 41-55. doi:10.1080/13571519984304
Javetski, B., Murphy, C., & Staples, M. (2014). Perspectives on Founder- and Family-Owned Businesses. McKinsey & Company.
Jiraporn, P., & Dadalt, P. (2009). Does founding family control affect earnings management? Applied Economics Letters, 16(2), 113-119.
24 Lemmon, M. L., & Lins, K. V. (2003). Ownership structure, corporate governance, and firm
value: Evidence from the east asian financial crisis. Journal of Finance, 58(4), 1445-1468. doi:10.1111/1540-6261.00573
Lins, K. V., Volpin, P., & Wagner, H. F. (2013). Does family control matter? international evidence from the 2008– 2009 financial crisis. The Review of Financial Studies, 26(10), 2583-2619. doi:10.1093/rfs/hht044
Maury, B. (2006). Family ownership and firm performance: Empirical evidence from western european corporations. Journal of Corporate Finance, 12(2), 321-341.
doi:10.1016/j.jcorpfin.2005.02.002
Miller, Danny., Isabelle Le Breton-Miller, Richard H. Lester, Albert A. Cannella Jr., 2007. Are family firms really superior performers?, Journal of Corporate Finance. Vol. 13: 829-858.
Morck, R., Shleifer, A., & Vishny, R. W. (1988). Management ownership and market valuation: An empirical analysis. Journal of Financial Economics, 20, 293-315. doi:10.1016/0304-405X(88)90048-7
Morck, R. K. (1998). In Strangeland D. A., Yeung B. (Eds.), Inherited wealth, corporate
control, and economic growth: The canadian disease?. Cambridge, Mass: Cambridge,
Mass. : National bureau of economic research ; Cambridge, Mass. : NBER Host].
Morck, R., & Yeung, B. (2003). Agency problems in large family business groups. Entrepreneurship: Theory and Practice, 27(4), 367.
Porta, R. L., LopezdeSilanes, F., Shleifer, A., & vishny, r. w. (1998). Law and finance. Journal of Political Economy, 106(6), 1113-1155. doi:10.1086/250042
Porta, R., Lopez‐de‐silanes, F., & Shleifer, A. (1999). Corporate ownership around the world. Journal of Finance, 54(2), 471-517. doi:10.1111/0022-1082.00115
Prez-Gonzlez, F. (2006). Inherited control and firm performance. American Economic
25 Shleifer, A., & Vishny, R. W. (1997). A survey of corporate governance. Oxford, UK:
doi:10.1111/j.1540-6261.1997.tb04820.x
Stein, J. (1989). Efficient capital- markets, inefficient firms - a model of myopic corporate- behavior. Quarterly Journal of Economics; Q.J.Econ., 104(4), 655-669.
Stein, J. C. (1988). Takeover threats and managerial myopia. The Journal of Political
Economy, 96(1), 61. doi:10.1086/261524
Villalonga, B., & Amit, R. (2006). How do family ownership, control and management affect firm value? Journal of Financial Economics, 80(2), 385-417.
doi:10.1016/j.jfineco.2004.12.005
Wang Xiao-Wei, & Hou Nai-Kun. (2006). Empirical research on the correlation between the
agency costs and corporate value based on ownership's concentration degree doi:10.1109/ICMSE.2006.314032
Wang, Yangbo., Haoyong Zhou, 2012. Are Family Firms Better Performers during Financial Crisis?, Working paper, Keele Management School.
Yeung, B. (1998). Inherited wealth, corporate control and economic growth: The canadian
