!
!
!
!
Family ownership and firm performance !
during and around the crisis of 2007-2012!
Bachelor thesis about European family firms during and around the recent !
financial crisis!
!
!
!
!
!
!
!
!
!
!
!
Author: ! ! O. M. Bieze! Student-number: ! 10113312!!
Institution: ! ! University of Amsterdam! ! ! ! ! ! !
Programme:! ! BSc. Economics & Business!
Specialization: ! Finance & Organization!
Field: ! ! ! Organizational Economics!
!
Supervisor: ! ! A. R. S. Woerner!
!
Date:! ! ! June 29, 2016!
Statement of originality
!!
This document is written by student Onno Bieze who declares to take full responsibility for the contents of this document.!
!
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.!
!
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
Abstract
!!
In this paper the relationship between family ownership and firm performance of European pu-blicly listed firms is investigated with a focus on the recent financial crisis of 2007-2012. The fin-dings are ambiguous but suggest, based on two of the three performance measures used, that family firms outperform non-family firms. During the crisis years the findings also point towards the family owned firms performing better than non family owned firms, which is contrary to the expectations. Overall, the results hint at a better performance of family firms, suggesting that, even in times of crisis, family ownership is a good organizational structure.
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
List of contents!
!
Statement of originality! ! ! ! ! ! ! ! ! 2!!
Abstract ! ! ! ! ! ! ! ! ! ! ! 3!!
List of contents ! ! ! ! ! ! ! ! ! ! 4!!
Section 1 ! ! Introduction! ! ! ! ! ! ! ! 5!!
Section 2! ! Literature Review! ! ! ! ! ! ! 8!
! 2.1! Principal-agent theory!! ! ! ! ! ! ! 8!
! 2.2! Long-term vision and investment strategy! ! ! ! ! 9!
! 2.3! Expropriating wealth! ! ! ! ! ! ! ! 10!
! 2.4! Family altruism! ! ! ! ! ! ! ! 11!
! 2.5 ! Predictions of the relationship between family ownership and ! ! !
! ! firm performance! ! ! ! ! ! ! ! 11!
!
Section 3! ! Methodology!! ! ! ! ! ! ! 13!
! 3.1 ! Data sample! ! ! ! ! ! ! ! ! 13! !
! 3.2 ! Family firm variable! ! ! ! ! ! ! ! 15!
! 3.3 ! Crisis years variable! ! ! ! ! ! ! ! 15!
! 3.4 ! Performance variables! ! ! ! ! ! ! 15!
! 3.5! Control variables! ! ! ! ! ! ! ! 16!
! 3.6! Descriptive statistics! ! ! ! ! ! ! ! 16!
!
Section 4! ! Empirical findings! ! ! ! ! ! ! 21!
! 4.1 ! Family firms vs. non-family firms and their performance! ! ! 21! ! 4.2 ! Family firms vs. non-family firms and their performance during ! ! !
! ! crisis years! ! ! ! ! ! ! ! ! 23!
! 4.3 ! Robustness tests! ! ! ! ! ! ! ! 25!
!
!
!
Section 5! ! Conclusion, discussion and limitations! ! ! ! 26!
! 5.1! Conclusion and discussion! ! ! ! ! ! ! 26!
! 5.2! Limitations! ! ! ! ! ! ! ! ! 27!
!
Reference list! ! ! ! ! ! ! ! ! ! 29!!
Appendices! ! ! ! ! ! ! ! ! ! ! 32!!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
1. Introduction!
!
Since 2013 the Center for Family Business at the University of St. Gallen, Switzerland, in coop-eration with EY’s Global Family Business Center of Excellence constructs a list of the 500 lar-gest family firms in the world based on their revenue, named the Global Family Business Index . 1
On their website it is stated that this index provides impressive evidence of the economic power and relevance of family firms in the world. Several studies in the economic literature come to the same conclusion. (Shleifer & Vishny, 1986; La Porta et al., 1999; Claessens et al., 2000; Faccio & Lang, 2002; Anderson & Reeb, 2003). !
! Based on the different results in the economic literature there isn’t reached consensus yet whether family firms are performing better or worse than non-family firms. Anderson and Reeb (2003) find in their study about family owned firms in the S&P 500 index that family firms outperform non-family firms, which is contrary to their hypotheses. However, Holderness and Sheehan (1988) find the opposite to be true. In their research among large U.S. public firms they come to the conclusion that non-family firms perform better using Tobin’s Q as a perfor-mance measure.!
! In the past ten years a lot of corporations have been greatly impacted by the financial crisis, which started in 2007 and ended around 2012. The question remains if both family and non-family firms are impacted in the same way by this latest financial crisis. One of the reasons for the crisis is widely believed to be a lack of long-term vision by decision makers like mana-gers and investors. According to Levie and Lerner (2009), family firms take advantage of a strong long-term vision by building the long-term reputation of the firm and by bonding with their external shareholders through good relationships. This good long-term vision of family firms might suggest that they could be less affected by the financial crisis than non-family firms. ! ! This research will compare the performance of family firms and non-family firms, and there will be a focus on the question if the performance of family firms or non-family firms are more affected by the recent financial crisis. The firms used in this research are publicly listed firms and most of them can be found on the Standard & Poor’s Euro Plus Index, which consists of firms in fifteen developed countries in the mainland of Europe. The family firms are identified using the Global Family Business Index. Some, but not all, of these family firms are present on both the Euro Plus Index and the Global Family Business Index.!
http://www.familybusinessindex.com
! The main research question is formulated as follows: How has family ownership affected firm performance of publicly listed European firms during and around the recent financial crisis of 2007-2012?!
! This research question is going to be answered using a literature review and empirical research. In the next section the existing literature is reviewed and the expectations with respect to the relation between family ownership and firm performance is discussed. In section three the data sample is described, the different key variables are discussed, and some descriptive statis-tics are presented. Section four provides the empirical results between family ownership and firm performance, during and around the times of the financial crisis. Section five concludes the paper and discusses the limitations of this research.!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
2. Literature review!
!
In this section the existing literature is reviewed and the expectations regarding the relationship between family ownership and firm performance is discussed. In the next few paragraphs the most important theories and points are presented in which family firms differ from non-family firms according to the existing literature.!
!
2.1 Principal-agent theory!
First, the principal-agency theory of Jensen and Meckling (1976) is discussed. The basis of this theory is based on the idea that most people have their own self-interest in mind, when acting in a relationship. For example, in a principal-agent relationship both the owner (principal) and the manager (agent) of a firm try to maximize their own utility. The utility maximizing behavior of the manager may not be in the best interest of the owner. There arise agency costs, because the ultimate goals of the principal and the agent aren’t aligned. The agency costs in this case are all costs made by the owner to align the manager’s interests with his own interests, such as incen-tives programs to ensure the manager only engages in activities that are good for the firm, as well as some company policies, which have the goal to guarantee that managers don’t take cer-tain action that would do harm to the owner’s interest.!
! According to Berle and Means (1932) these agency costs are diminished in a firm where the ownership is concentrated and not widely dispersed, and thus should have a positive effect on the value of the firm. However, Demsetz (1983) argues that the ownership structures that emerge in firms are an endogenous outcome of profit maximizing decisions regarding cost ad-vantages and disadad-vantages. Through these decisions a firm will ultimately arrive at the best ownership structure for that particular firm. Thus according to Demsetz (1983), the ownership structure of a firm isn’t based on the preferences of shareholders for a particular ownership structure, but is only based on the preferences of shareholders to maximize the firm’s value. Therefore, the ownership structure of family firms should not have a competitive advantage over the ownership structure of widely held firms, or vice versa.!
! Demsetz and Lehn (1985) think that concentrated ownership creates incentives to de-crease the agency conflicts that arise in the principal-agent relationship between owners and managers. They state that a family’s wealth is so strongly linked with the firm’s performance that the families have a strong incentive to monitor the managers and thus minimize the free-rider problem which is an inherent problem in a firm with widely dispersed ownership. This would
suggest a better performance of family firms, however, their results don’t provide evidence of a positive effect of family ownership on the firm performance. !
! Overall family firms, which have more concentrated ownership than most widely disper-sed non-family firms, probably have less agency costs and less problems with the free-rider problem which should have a positive effect on the firm’s performance.!
!
2.2 Long-term vision and investment strategy!
Another potential benefit of family ownership is the better long-term vision family firms are belie-ved to have with respect to more widely-held firms. Anderson and Reeb (2003) discuss this long-term vision of family firms and they state that there arise reputation concerns from the fami-ly’s sustained presence in the firm. They suggest that the long-term nature of family firms has an effect on external stakeholders, like suppliers and capital providers, through the willingness of these stakeholders to deal with the same governing bodies and practices for longer periods in family firms relative to non-family firms. The building of a good long-term reputation with respect to the firm’s external stakeholders is also mentioned by Levie and Lerner (2009). The reputation of families in family firms can thus create economic value through these good bonds with the firm’s outside stakeholders.!
! According to James (1999), the long-term vision of families provides incentives to invest according to the net present value rule and thus to don’t forgo beneficial investment opportuni-ties because of the current short-term costs. This suggests that family firms have a more effi-cient investment strategy than more widely held non-family firms.!
! This is in line with the findings of Bertrand and Schoar (2006). They state that the family links that bind current generations and future ones provide family firms with a focus on maximi-zing long-term returns and the desire to invest in opportunities that more widely held firms would not pursue.!
! Furthermore, Lins et al. (2013) found that in the recent financial crisis firms that cut in-vestment more have greater stock declines. This link between inin-vestment and stock price decli-ne suggests that profitable investment opportunities are cut. They found that about a third of the underperformance of family firms is explained by this underinvestment. This underinvestment is in line with the results of Faccio et al. (2011) who find in their research that firms that are con-trolled by undiversified shareholders, like family firms, have the tendency to undertake less risky investments than firms that have diversified shareholders. The risky investments which have potentially high benefits, are cut, and the riskless investment opportunities are taken. !
! On the one hand, this is contrary to the view of James (1999) that family firms don’t for-go beneficial investment opportunities. On the other hand, the research of Lins et al. (2013) could be seen in line with the belief of Anderson and Reeb (2003) that the firms survival is an important concern for families, which may be the reason why family firms underinvest in times of crisis.!
! The long-term vision of family firms could be beneficial for the firm performance when regarding the long-term bonds with external stakeholders the families maintain and regarding the investment policies family firms have. However, in times of crisis this long-term vision of fa-mily owned firms could potentially give rise to a worse performance than non-fafa-mily firms. To secure the ultimate survival of the firm for the family, the decision makers of the family firms may undercut high risk, but potentially very valuable investment opportunities, and therefore have a lower performance than non-family firms.!
!
2.3 Expropriating wealth!
A potential disadvantage of family firms is the idea that members in public family owned or con-trolled firms may exploit minority shareholders by engaging in opportunistic activities according to Anderson and Reeb (2003). Harold Demsetz (1983) and Demsetz and Lehn (1985) provide evidence of this statement and observe that the families in family owned firms have the power and the incentive to expropriate wealth from the minority shareholders which suggests severe moral hazard conflicts between the founding family and minority shareholders.!
! According to Johnson et al. (2000), tunneling is the term used to refer to the transfer of resources out of a firm to its controlling shareholder and it comes in two forms. The first form is that a controlling shareholder is simply transferring resources from the firm to the shareholder itself via self-dealing transactions that includes theft and fraud, which are illegal, but also asset sales and contracts such as advantageous asset pricing towards the shareholder.!
! Second, through activities such as dilutive share issues, minority freeze-outs, insider trading and creeping acquisitions, the controlling shareholder can increase his share of the firm and therefore discriminate against minority shareholders. These tunneling activities will be be-neficial to the controlling shareholder, but often aren’t bebe-neficial to the firm as a whole. For the overall firm performance of family firms these factors could have a negative influence.!
!
!
!
2.4 Family altruism!
The term family altruism is the idea that to retain control in firms, families choose members of their family to fill the top positions in the firm because they have trust in each other. However, when filling these top positions in the firm, families, who choose only out of a pool of family members, have on average less competent candidates to choose from. Lubatkin et al. (2007) state that non-family firms thus are more likely to have superior human capital in their firm, be-cause the potential talent pool where their managers are chosen from extends beyond the smal-ler pool of only family members. The research of Smith (2006) supports this view by stating that on average family firms have more major decision makers with less qualifications and who adept less quickly to some management practices than non-family owned firms.!
! Villalonga and Amit (2006) find in their research that family firms, where the founder of the firm is in control as a CEO or if he hired an outside candidate for the position of CEO, per-form better than non-family firms. However, when family firms are run by descendant-CEOs, mi-nority shareholders in those firms are worse off than they would be in non-family firms in which they would be confronted to the classic principal-agent conflict with managers. These findings of Villalonga and Amit (2006) are in line with the belief that on average the founder of a firm could find a more talented candidate for the top decision making positions if he chooses not only from the pool of family members.!
! Thus family altruism could be a potential disadvantage for the performance of family firms if the pool of potential decision makers the family firms are willing to choose from, consists of less competent candidates than the pool where their non-family counterparts are choosing from.!
!
2.5 Predictions of the relationship between family ownership and firm performance!
There is already done a lot of research on the differences between family firms and non-family firms, but whether one type outperforms the other still remains an open question. Berle and Means (1932) suggest that firms where the ownership is concentrated, like family firms, should perform better than widely held firms due to reduced agency costs. This view is challenged by Demsetz (1983). !
! Based on the believed good long-term vision of family firms, according to Anderson and Reeb (2003), and Levie and Lerner (2009), family firms may perform better than non-family firms due to their long-term reputation by creating strong bonds with external stakeholders. Also the investment strategies of family firms are considered to be based on long-term profits, and
profitable investment opportunities are not forgone because of the short-term costs according to James (1999).This statement is backed by Bertrand and Schoar (2006). However, following Lins et al. (2013) family firms may cut investment in times of crisis to secure the ultimate survival of the firm, which is why family firms may underperform in the crisis years. !
! Some potential disadvantages, according to Anderson and Reeb (2003), Demsetz (1983), and Demsetz and Lehn (1985), regarding the firm performance of family owned firms come from the expropriating of wealth and the exploitation of minority shareholders. Furthermo-re, the tunneling activities mentioned by Jonhson et al. (2000), which are used to potentially gain benefits for the family while exploiting the other minority shareholders, could be a negative influence on the firm performance.!
! Then there is the idea that family firms have potential less viable candidates for the deci-sion making top positions in the firm as stated by Lubatkin et al. (2007), because the families only choose from a pool of members of their family. This could mean that family firms have less competent managers and therefore the firm’s performance could be lower than that of compa-rable non-family firms.!
! Based on these different potential advantages and disadvantages of family ownership in a firm, the prediction of the relationship between family ownership and firm performance is most likely positive. The potential benefits of a reduced agency conflict between owners and mana-gers in a family owned firm, and the long-term vision of families in their firms regarding working together with external stakeholders and regarding their investment horizon, outweigh the poten-tial costs of having exploited minority shareholders and having less competent decision makers.! ! This predicted better performance of family firms with respect to non-family firms in the crisis years will most likely be negative, because of the undercutting of profitable investment op-portunities that family firms undertake to secure the survival of the firm.!
! !
!
!
!
!
!
!
!
!
3. Methodology!
!
3.1 Data sample!
The Global Family Business Index, constructed by the University of St. Galen, Switzerland, in cooperation with the Ernst & Young’s Global Family Business Center of Excellence, consists of the 500 largest family firms in the world, based on the yearly revenue. !
! The list consists of both private and publicly listed family firms. This research focuses on the publicly listed European family firms of the Global Family Business Index which are compa-red to the constituents of the Standard & Poor’s Euro Plus Index. This index of the rating agency Standard & Poor’s is a subindex of the S&P Europe 350 index and includes the largest Euro-pean firms which are restricted to the developed countries of the mainland of Europe. The coun-tries included in this index are Austria, Belgium, Denmark, Finland, France, Germany, Ireland, Italy, Luxembourg, Netherlands, Norway, Portugal, Spain, Sweden and Switzerland. !
! For this research there is deliberately chosen not to include firms of the United Kingdom, which are included in the overall S&P Europe 350 index, because of the different management style that most firms in the United Kingdom have. !
! The Global Family Business Index is stripped for this research to include only the firms of these above mentioned European countries, which are all publicly listed firms. The perfor-mance of all the firms in the sample is observed for over a decade, from 2005 until 2015. The total sample includes 75 family firms, with a total of 803 observations, and 197 non-family firms, with a total of 2022 observations. In total the sample thus consists of 272 firms, with a total of 2825 observations. Most firms have data on all the eleven years from 2005-2015, but there are also some firms where not the whole time period is covered in the data.!
! The performance and accounting data of each firm in this sample is measured on an an-nual basis and is found via the database of Compustat, which is a database of financial, statisti-cal and market information on active and inactive global companies throughout the world.! ! Table I provides an overview of the number of family and non-family firms per industry division included in the sample. The different industry divisions are based on the two-digit Standard Industry Code (SIC) by which all the firms in the sample are divided. Firms that belong in the division Finance, Insurance & Real Estate (SIC 60-67) are excluded from the sample, be-cause the firms in this industry division are subjected to different regulations.!
! As shown in table I the firms in the sample are distributed over eight different industry divisions. There are no firms that belong in the Agriculture, Forestry & Fishing division which is !
the only division next to the excluded Finance, Insurance & Real Estate division that isn’t repre-sented in the sample. By far the largest industry division in the sample is Manufacturing, with a total of 167 firms, where roughly 28 percent is indicated as a family firm. Retail Trade is the divi-sion with the highest percentage of family firms. In this dividivi-sion more than half (58.8 percent) of the total firms is considered to be a family firm. Transportation & Public Utilities is with a percen-tage of 13.5 percent the industry division with the lowest percenpercen-tage of family firms in the total sample.!
In Appendix A there is a more detailed table provided, where the firms are divided per industry based on their different two-digit Standard Industry Code (SIC).!
!
!
Table I!Number of family and non-family firms per industry division.!
!
!
Two-digit SIC Industry division description Family firms Frequency Non-family firms Frequency Total firms FrequencyPercent family firms in division 01-09 Agriculture, Forestry & Fishing 0 0 0 -10-14 Mining 1 5 6 16.7% 15-17 Construction 5 6 11 45.5% 20-39 Manufacturing 47 120 167 28.1% 40-49 Transportation & Public Utilities 5 32 37 13.5% 50-51 Wholesale Trade 2 4 6 33.3% 52-59 Retail Trade 10 7 17 58.8% 60-67 Finance, Insurance
& Real Estate excluded excluded excluded
-70-89 Services 3 17 20 15.0%
91-99
Public Administra-tion & Nonclassifi-able Establishments
2 6 8 25.0%
3.2 Family firm variable!
For the analysis of this sample a dummy variable is used to indicate whether a firm is conside-red to be a family firm or a non-family firm. A firm is marked a family firm when it is included in the Global Family Business Index. !
! The constructors of this index mark a firm as being a family firm when more than 32 per-cent of the voting rights are in the hands of one family. This cut-off at 32 perper-cent is motivated by the idea that around 30 percent of the votes are sufficient to control the general assembly of a publicly listed company in the OECD countries, because on average 60 percent of the votes are present in the general assembly. They decided to be somewhat more conservative with their 32 percent cut-off which is also more conservative than most academic studies.!
! The dummy variable in this research is thus set to one if the firm is included in the Global Family Business Index and is set to zero if a firm in the sample is not included in the index.!
!
3.3 Crisis years variable!
The crisis years and the non-crisis years are also indicated by a dummy variable which is set to one if the year falls in the period of 2007 until 2012, and equals zero if the year is 2005, 2006, or falls in the period of 2013 until 2015. The crisis period is chosen following De Bruyckere et al. (2013), who have named the period of 2007-2012 the European debt crisis.!
!
3.4 Performance variables!
The first performance variable used is return on assets (ROA). Following Anderson and Reeb (2003), the return on assets is calculated in two ways. The first approach is to divide a firm’s net income by the book value of total assets. In the second approach earnings before interest, taxes, depreciation and amortization (EBITDA) divided by the book value of total assets is used. The return on assets ratio shows how well the management of a firm uses its assets to generate earnings.!
! The second performance variable is return on equity (ROE) which is calculated as the firm’s net income divided by the firm’s total shareholder equity. Thus, the return on equity mea-sures the efficiency of a firm generating profits from every unit of shareholder’s equity, and shows how well the management uses the firm’s investment funds to generate earnings growth.! ! In most empirical studies about family firm performance also Tobin’s Q is used as a me-asure to calculate a firm’s performance. However, as pointed out by Dybvig and Warachka (2012), the relationship between Tobin’s Q and firm performance is confounded by endogeneity.
They argue that inefficiency due to underinvestment lowers the firm performance but increases Tobin’s Q. Therefore, Tobin’s Q will not be included as a performance measure in this research.!
!
3.5 Control variables!
To control for the firm-specific characteristics of a firm there are some variables included in this research. The firm size is measured as the natural logarithm of the book value of total assets. The capital expenditure of a firm are the funds used to upgrade or acquire physical assets such as property or equipment and is often used to undertake new projects of investments by the firm. In this research the natural logarithm of the capital expenditure of a firm is used. The long-term debt ratio is calculated as the long-long-term debt divided by the book value of total assets. ! ! Within different industry divisions and during different years there could be some unfore-seen factors that are of a greater influence within certain industry divisions or during certain ye-ars. That is why there’s controlled for the different industries by including dummy variables ba-sed on the different industry divisions and there are dummy variables included to control for the different years separately.!
! Whether a firm is part of the S&P Euro Plus is also indicated by a dummy variable. This variable is set to one if a firm is part of the S&P Euro Plus Index, and zero if a firm isn’t part of the S&P Euro Plus. All the non-family firms are part of this index and some of the family firms are part of this index. This dummy variable is introduced to control for the specific characteris-tics firms in this index may have.!
!
3.6 Descriptive statistics!
Table II presents four panels of descriptive information about the data sample. Panel A provides means, standard deviations, minimum and maximum values of the different key variables in the sample. Panel B shows the correlation matrix for these key variables. Panel C and panel D pro-vide the results of the difference of means tests between family and non-family firms during the crisis years, and the non-crisis years, respectively. In the panels C and D a difference is signifi-cant when the t-statistic at the right column has one star (*), two stars (**) or three stars (***) at a two-tailed significance level of ten, five, or one percent, respectively.!
! Panel A shows that approximately 28 percent of the firms in the sample are family firms. The assets of the firms in the sample are on average worth 33.96 billion dollars, and their aver-age capital expenditure is around 1.88 billion dollars. The firms increase their sales each year with about 7.8 percent. The average firm’s assets are financed with 19.45 percent long-term
debt, and the return on assets, based on net income, is 5.17 percent. The return on assets, ba-sed on the earnings before interest, taxes, depreciation and amortization, are for a standard firm in the sample 12.65 percent. The return on equity is on average 57.70 percent, and is a lot hig-her for non-family firms than for family firms.!
! Panel B presents the correlation matrix of the key variables in the sample. The correlati-on matrix shows a positive relaticorrelati-on between the coefficients of family firms and the return correlati-on as-sets, based on net income. However, the coefficient between return on asas-sets, based on EBIT-DA, and family firms gives a negative relation. This is also the case for the coefficient between the return on equity and family firms.!
! In panel C the difference of means tests between family firms and non-family firms in the crisis years are presented. In these years the family firms are on average 15.73 billion dollars less worth than their non-family counterparts. Their capital expenditure is lower by about 0.58 billion dollars. The assets of the family firms are less financed with long-term debt, and their sa-les have a higher growth rate on average.!
! When it comes to the performance measures family firms have a higher return on as-sets, based on net income than non-family firms, with a difference of 0.32 percent. However, the return on assets, based on EBITDA, is lower for family firms, with a difference of 0.29 percent. Also the return on equity is much lower for family firms than for non-family firms.!
! Panel D provides the same difference of means tests but for the non-crisis years. It shows that in these years the family firms are on average 19.02 billion dollars less worth than non-family firms. The family firms have a lower capital expenditure than non-family firms by about one billion dollars. The assets of the family firms are less financed with long-term debt, but in the non-crisis years the family firms have a lower growth of their yearly sales.!
! When it comes to the performance measures family firms have a lower return on assets, based on both net income and EBITDA, than non-family firms, with a difference of 0.22 percent and 0.36 percent respectively. In these years the return on equity is also a lot lower for family firms with respect to the non-family firms.!
! The statistics shown in panel C and panel D point towards the possibilitythat family firms perform relatively better in the crisis years than non-family firms, because the difference in re-turn on assets, based on net income, is positive in these years and the negative differences, based on the other return on assets ratio and the return on equity ratio, are smaller in these ye-ars than they are in the non-crisis yeye-ars.!
!
!
!
Table II: Panel A!
Summary statistics for the different key variables in the sample. It provides the means and standard devi-ations for the total of sampled firms, and for family firms and non-family firms separately. The last two
co-lumns of the table provide the minimum and maximum values of all the firms in the sample.!
!
! ! ! !!
!
!
!
!
Mean StandardDeviation Min. Max.
Family/non-family (Family/non-family=1) 0.28 0.45 0.00 1.00
All Family
Non-family All Family
Non-family Total Assets ($000,000,000) 33.96 21.64 38.85 65.85 51.26 70.23 0.21 986.40 Ln(total assets) ($000,000) 9.48 8.96 9.69 1.36 1.26 1.35 5.34 13.80 Capital Expendi-ture ($000,000,000) 1.88 1.32 2.11 6.59 4.94 7.13 0.00 12.47 Ln(cap. ex.) ($000,000) 6.16 5.66 6.36 1.62 1.49 1.63 -2.92 11.73 Long-term debt ratio (%) 19.45 16.84 20.49 14.38 13.07 14.74 0.00 114.32 Sales growth (%) 7.80 8.58 7.49 83.57 65.35 89.82 -100.00 3437.30
ROA - net
in-come (%) 5.17 5.23 5.15 8.90 5.68 9.91 -89.79 253.86 ROA - EBITDA
(%) 12.65 12.42 12.74 7.69 7.27 7.85 -35.86 76.93
!
!
!
!
Table II: Panel B!
Correlation matrix for the key variables in the sample.!
!
!
!
!
!
!
!
!
!
!
!
!
!
Family/non-family ROA - net income ROA - EBITDA ROE Long-term debt ratio Sales Growth Ln (total assets) Ln (total cap. ex.) Family/ non-family 1.000 ROA - net income 0.0062 1.000 ROA - EBITDA -0.0172 0.5717 1.000 ROE -0.0522 0.0641 0.1099 1.000 Long-term debt ratio -0.1107 -0.2523 -0.1961 -0.0205 1.000 Sales Growth 0.0083 0.0017 -0.0541 -0.0045 0.0082 1.000 Ln (total assets) -0.2715 -0.0589 -0.0796 0.1496 0.0700 -0.0368 1.000 Ln (total cap. ex.) -0.2172 -0.0342 0.0376 0.1497 0.0522 -0.047 0.8868 1.000Table II: Panel C!
Difference of means tests for variables between family and non-family firms during the crisis years. The stars behind the t-statistics indicate whether the difference of means is significant on a one (***), five (**),
or ten (*) percent two-tailed significance level.!
!
!
!
!
!
Table II: Panel D!
Difference of means tests for variables between family and non-family firms in the non-crisis years. The stars behind the t-statistics indicate whether the difference of means is significant on a one (***), five (**),
or ten (*) percent two-tailed significance level.!
!
!
!
Family firms (crisis years) Non-family firms (crisis years) Difference of means t-statistic Total Assets (in billions) 22.55 38.28 -15.73 4.4174*** Total Cap. Ex. (in billions) 1.48 2.06 -0.58 1.6226Long-term debt ratio 17.21 21.36 -4.15 4.9840***
Sales growth 9.56 3.06 6.49 -2.3963**
ROA - net income 5.25 4.93 0.32 -0.7036
ROA - EBITDA 12.48 12.77 -0.29 0.6724 ROE 12.15 70.16 -58.01 1.9923** Family firms (non-crisis years) Non-family firms (non-crisis years) Difference of means t-statistic Total Assets (in billions) 20.52 39.54 -19.02 4.4980*** Total Cap. Ex. (in billions) 1.12 2.17 -1.05 2.4362**
Long-term debt ratio 16.38 19.41 -3.03 3.5914***
Sales growth 7.11 14.26 -7.15 0.8638
ROA - net income 5.20 5.42 -0.22 0.3570
ROA - EBITDA 12.35 12.71 -0.36 0.7610
4. Empirical findings!
!
This section presents the empirical findings of this research. The interest lies in the relationship between family firms vs. non-family firms and their performance, and special interest is placed how this relationship differs in the crisis years. To examine this relationship between these va-riables two models of linear regression are used. In the models there will be controlled for total assets, total capital expenditure, long-term debt ratio, sales growth, industry, year, and for the presence of a firm in the S&P Euro Plus index. ! !
! In table III the results of these models are presented. In model A the firm performance, measured by three different performance measures, is regressed on the dummy variable family firm while controlling for different other variables. Model B places the focus on the differences between family firms and non-family firms with the crisis years as a dummy variable. An interac-tion term of family firms in combinainterac-tion with the crisis years (family*crisis) is included to show the differences of family firms in the crisis years, with respect to the non-family firms. In these models a coefficient is significant when the underlying t-statistic within the parentheses has one star (*), two stars (**), or three stars (***) at a two-tailed significance level of ten, five or one per-cent, respectively.!
!
4.1 Family firms vs. non-family firms and their performance!
Model A of table III hints that, with respect to the performance measures ROA - net income and ROA - EBITDA, family firms perform better than non-family firms. The coefficients of the family firm variable is positive in these cases, with 0.0118 and 0.0200 respectively. This shows that the factor, being a family firm, is positively related with the performance of a firm. These coefficients are significant on a two-tailed one percent significance level. These findings comply with the re-sults that Anderson and Reeb (2003) found in their research about listed companies of the S&P 500 index. Although the coefficients for the family firm variable on both the return on assets me-asures found in this research are somewhat higher than the coefficients Anderson and Reeb (2003) found. !
! The coefficient of being a family firm, when looking at the return on equity performance measure has a negative value of -0.6789, and thus suggests the contrary with respect to the other performance measures. This two-tailed one percent significant negative value hints that
family firms perform worse than their non-family counterparts based on this performance mea-sure.!
!
!
Table III: Model A!
Linear regression results of family firms vs. non-family firms and firm performance. The t-values are inclu-ded within the parentheses underneath the coefficients. The stars behind the t-values indicate whether
the above standing coefficient is significant on a one (***), five (**), or ten (*) percent two-tailed ! significance level.!
!
ROA - net income ROA - EBITDA ROE
Intercept 0.0710 0.1809 -3.6767 (3.18)*** (7.41)*** (-3.95)*** Family firm 0.0118 0.0200 -0.6790 (2.98)*** (4.83)*** (-4.01)*** Ln (total assets) -0.0085 -0.0312 0.4011 (-2.52)** (-10.03)*** (-3.73)*** Ln (cap. ex.) 0.0031 0.0247 0.3234 (-0.64) (7.55)*** (-1.70)*
Long term debt
ratio -0.1537 -0.0852 -1.1557
(-6.19)*** (-5.88)*** (-1.48)
Sales growth -0.0001 -0.0049 0.0408
(-0.07) (-5.61)*** (1.95)*
S&P Euro Plus 0.0352 0.0489 -1.0885
(6.50)*** (10.30)*** (-3.74)*
SIC - Division included included included Year dummy included included included
R-squared (%) 9.71 18.75 3.12
Number of
4.2 Family firms vs. non-family firms and their performance during crisis years!
Model B presents the relationship of the crisis years on the firm performance of both family and non-family firms. In this model the crisis coefficient shows the relationship between non-family firms and firm performance during the crisis years, and the crisis coefficient plus the!
interaction term family*crisis shows the relationship between family firms and firm performance during the recent financial crisis. ! !
! The crisis coefficients are negative for both the return on assets, based on net income, as well as the return on equity, with -0.0080 and -0.3907 respectively . This suggest a negative relationship between the non-family firms and the firm performance during the crisis years. At both these performance measures, when adding the positive coefficient family*crisis to the crisis coefficient, the result remains negative which points towards family firms also performing worse during the crisis years. This complies with the idea of the crisis having a negative influence on a firm’s performance. The positive coefficients for the family*crisis interaction term at these per-formance measures point towards family firms performing better than non-family firms during the crisis years, yet these coefficients are not significant.!
! Looking at the return on assets, based on EBITDA, the coefficients suggest the contrary. The positive crisis coefficient of 0.0635 hints at a positive relationship for non-family firms and the firm performance during the crisis years. The coefficients of the crisis variable and the fami-ly*crisis interaction terms together also suggests a positive relationship for family-firms and the firm performance in the crisis years. Here the small negative family*crisis coefficient suggests that family firms perform slightly worse than non-family firms during the crisis, although this coefficient is not considered to be significant.!
! The results found in this paper are not in line with the results of Lins et al. (2013), who wrote earlier about family firm performance in the recent crisis. They found that family controlled firms underperform when compared to non-family controlled firms during the crisis years. In this paper the results are ambiguous. Based on the ROA - EBITDA there could be suggested that family firms indeed underperform in comparison with non-family firms. However, the coefficient of -0.0003 is very small and not statistically significant. The coefficients of the family*crisis va-riable on the other performance measures are contrary to the findings of Lins et al. (2013). However, they used crisis period return as a performance measure and also another time period is used to indicate the crisis years. Therefore, the results cannot be compared directly but the overall conclusion of the results hints at a different outcome.!
!
Table III: Model B!
Linear regression results of family firms vs. non-family firms and their performance during crisis years. The t-values are included within the parentheses underneath the coefficients. The stars behind the
t-valu-es indicate whether the above standing coefficient is significant on a one (***), five (**), or ten (*) percent two-tailed significance level.!
!
!
!
ROA - net income ROA - EBITDA ROE
Intercept 0.1183 0.1809 -3.3969 (7.25)*** (7.40)*** (-3.80)*** Family firm 0.0083 0.0202 -0.7542 (1.35) (3.76)*** (-2.66)*** Crisis -0.0080 0.0635 -0.3907 (-1.13) (3.22)*** (-2.13)** Family*Crisis 0.0058 -0.0003 0.1248 (0.95) (-0.05) (0.35) Ln (total assets) -0.0085 -0.0312 0.4011 (-2.52)** (-10.03)*** (3.73)*** Ln (cap. ex.) 0.0030 0.0247 0.3232 (-0.64) (7.55)*** (1.70)*
Long term debt
ratio -0.1535 -0.0853 -1.1522
(-6.19)*** (-5.88)*** (-1.48)
Sales growth -0.0002 -0.0049 0.0395
(-0.10) (-5.60)*** (1.84)*
S&P Euro Plus 0.0352 0.0489 -1.0882
(6.50)*** (10.30)*** (-3.73)***
SIC - Division included included included Year dummy included included included
R-squared (%) 9.73 18.75 3.12
Number of
4.3 Robustness tests!
The coefficients of the different variables on the return on equity suggests that the results may be driven by outliers. The high minimum and maximum values for some variables shown earlier in panel A of table II shows that there are indeed values that are very different from the mean of the return on equity. Therefore a robustness test of model A is included in appendix B, and for model B a robustness test is included in appendix C. In the robustness tests the outliers are ex-cluded. !
! For model A the robustness test provided in appendix B shows similar results regarding the relationship between family ownership and the firm performance, based on the return on assets performance measures. However, for the return on equity, the results differ drastically. This shows that the results are mainly driven by outliers.!
! For model B the robustness test in appendix C hints also at the results being driven by outliers. This robustness test points towards family firms performing worse during the crisis ye-ars than non-family firms, which is contrary to the findings in model B. This is however more in line with the results of Lins. et al. (2013) and therefore more in line with the predictions before-hand of family firms undercutting investment in times of crisis.!
!
! ! ! !!
!
!
!
!
!
!
!
!
!
!
!
!
!
5. Conclusion, discussion and limitations!
!
5.1 Conclusion and discussion!
The objective of this research paper is to provide some insights on the performance of family firms in developed countries in the mainland of Europe, more precisely the following counties: Austria, Belgium, Denmark, Finland, France, Germany, Ireland, Italy, Luxembourg, Netherlands, Norway, Portugal, Spain, Sweden and Switzerland. The goal is to present empirical evidence whether publicly listed family firms perform better or worse than publicly listed non-family firms, and there is made an attempt to compare the impact of the recent financial crisis of 2007-2012 on family firms and non-family firms.!
! The empirical part of this research presents some insight with respect to answering the question what kind of relationship there is between family ownership and firm performance, du-ring and around the recent financial crisis. The linear regression models hint that, on the basis of both return on assets performance measures, family firms outperform their non-family coun-terparts. These findings are in line with the findings in the influential paper of Anderson and Reeb (2003). On the basis of the return on equity family ownership is negatively related with the firm performance, which is contrary to the other results.!
! Furthermore, the results suggest that the financial crisis of 2007-2012 negatively influen-ces the performance of both family as non-family firms based on the ROA - net income, and the ROE, but the results suggest the contrary for the ROA - EBITDA.!
! Regarding the differences of family firms and non-family firms during the crisis years the results point towards a slight advantage in favor of the family firms. Based on two of the perfor-mance measures, the family firms seem to outperform the non-family firms during the crisis, and with the third performance measure the result is only a very small negative. However, these coefficients are not statistically significant, so no definitive conclusion about the performance differences of family and non-family firms during the crisis can be made. !
! The findings may hint that due to a better long-term vision of family owned firms, these firms are less prone to forgo profitable investments even when times are tough, because they invest according to the net present value rule, which is the view stated in the research of James (1999).!
! Overall this research also could be an indicator that the agency costs are lower for family firms than non-family firms, and therefore the performance of family firms is better than the per-formance of non-family owned firms.!!
! Although some suggestions are made in this conclusion, the main research question could unfortunately not be answered with a clear answer, because the results come to different conclusions regarding the different performance measures, and some of the results are not sta-tistically significant.!
!
5.2 Limitations!
There are some limitations regarding this research. First of all, the definition of when a firm is considered to be a family firm is quite conservative. The Global Family Business Index used to identify the family firms, indicates a firm as being family owned if 32 percent of the voting right is in the hands of the family. In most academic studies there’s a 20 percent ownership share used, to state that a firm is a family firm. Also this research doesn’t consider other family involvement, like having family members on the board of directors, which could be a factor in the performan-ce of firms.!
! Furthermore, the data sample has some limitations. There are some missing values within the different variables, because they are missing in the database of Compustat. Especial-ly within the famiEspecial-ly firms these missing values are maybe so many that the results presented are biased in one direction. However, there is no reason to believe that these values are either ex-ceptionally high or low, because the firms in the sample with the missing values weren’t the lar-gest or smallest firms based on the other variables in the data.!
! The data used in this research does not seem to be totally normally distributed. In the empirical part of this research however only linear regression models are used which don’t pre-dict the data points very well if the data isn’t normally distributed.!
! Also there are some extreme values that might bias the results presented. Therefore ro-bustness tests are provided in appendix B and appendix C. These roro-bustness tests show that based on the return on equity measure the results in this paper are mainly driven by outliers. In Appendix C also can be seen that the results differ on the basis of the other two performance measures. Thus this robustness test shows that the results in model B are driven by outliers.! ! Finally there could be an endogeneity problem, because there is reason to believe that family ownership is affected by firm performance. Family firms may only stay in the hands of families, because they perform well, and this could partly explain the positive coefficient of the family firm dummy variable on the two return on assets performance measures. This endogenei-ty problem could seriously affect the conclusions made in this research paper.!
! Concluding there are some limitations when regarding the empirical research, however, the findings of this research may give other researchers the incentive to do a more thorough research about the differences in family owned and non-family owned listed European firms in the crisis years.!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
Reference List
!!
Anderson, R. C., & Reeb, D. M. (2003). Founding- family ownership and firm performance: ! ! evidence from the S&P 500. The Journal of Finance, 58(3), 1301-1328.!
!
Berle, A., & Means, G., (1932). The modern corporation and private property. New york, NY: ! ! Harcourt, Brace, & World.!
!
Bertrand, M., & Schoar, A. (2006). The role of family in family firms. The Journal of Economic ! Perspectives, 20(2), 73-96.!
!
Claessens, S., Djankov, S., & Lang, L. H. P. (2000). Separation of ownership from control of ! ! East Asian firms. Journal of Financial Economics, 58(1-2), 81-112.!
!
De Bruyckere, V., Gerhardt, M., Schepens, G., & Vander Vennet, R. (2013). Bank/sovereign risk ! spillovers in the European debt crisis. Journal of Banking and Finance, 37(12), ! !
! 4793-4809.!
!
Demsetz, H. (1983). The structure of ownership and the theory of the firm. Journal of Law and ! ! Economics, 25(2), 375-390.!
!
Demsetz, H., & Lehn, K. (1985). The structure of corporate ownership: Causes and ! ! ! consequences. Journal of Political Economy, 93(6), 1155-1177.!
!
Dybvig, P. H., & Warachka, M. (2012). Tobin’s Q does not measure firm performance:! ! Theory, empirics, and alternative measures. (Working Paper). Retrieved from SSRN ! ! eLibrary website: http://phildybvig.com/papers/CGOE_revise16.pdf!
!
Faccio, M., & Lang, L. H. P. (2002). The ultimate ownership of Western European corporations. ! ! Journal of Financial Economics 65(3), 365–395.!
!
Faccio, M., Marchica, M., & Mura, R. (2011). Large Shareholder Diversification and Corporate ! ! Risk-Taking. Review of Financial Studies, 24(11), 3601-3641.!
!
Holderness, C. G., & Sheehan, D. P. (1988). The role of majority shareholders in publicly held ! ! corporations. Journal of Financial Economics, 20, 317–346.!
!
James, H., 1999. Owners as managers, extended Horizons and the family firm. International ! ! Journal of Economic Business, 6(1), 41–55.!
!
Jensen, M. C., & Meckling, W. H. (1976). Theory of the firm: Managerial behavior, agency costs ! and ownership structure. Journal of Financial Economics, 3(4), 305-360.!
!
Johnson, S., La Porta, R., Lopez-de-Silanes, F., & Shleifer, A. (2000). Tunneling. American ! ! Economic Review, 90(2), 22-27.
!
!
La Porta, R., Lopez-de-Silanes, F., & Shleifer, A. (1999). Corporate ownership around the world. ! The Journal of Finance, 54(2), 471-517.!
!
Levie, J., & Lerner, M. (2009). Resource mobilization and performance in family and non- family ! businesses in the United Kingdom. Family Business Review, 22(1), 25-38. !
!
Lins, K. V., Volpin, P., & Wagner, H. F. (2013). Does family control matter? International ! ! ! evidence from the 2008–2009 financial crisis. Review of Financial Studies, 26(10), !
! 2583-2619.!
!
Lubatkin, M. H., Ling, Y., & Schulze, W. S. (2007). An organizational justice-based view of self-! ! control and agency costs in family firms. Journal of Management Studies, vol. 44(6), !
! 955-971.!
!
Shleifer, A., & Vishny, R. (1986). Large shareholders and corporate control. The Journal of ! ! Political Economy, 94(3), 461–488.!
!
Smith, M. (2006). An empirical comparison of the managerial development of family and non- ! family SMEs from Australia’s manufacturing sector. Journal of Enterprising Culture,!
!
Villalonga, B., & Amit, R. (2006). How do family ownership, control and management affect firm ! ! value? Journal of Financial Economics, 80(2), 385–417.!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
Appendices!
!
Appendix A!
Number of family and non-family firms per industry based on the two-digit standard Industry Code (SIC).!
Division Two-digit SIC
Industry description Family firms
Frequency Non-family firms Frequency Total firms Frequency Percent family firms in industry
Mining 10 Metal, Mining 1 0 1 100.0%
13 Oil & Gas Extraction 0 5 5 0.0%
Construction 15 General Building Contractors
0 2 2 0.0%
16 Heavy Construction, Except Building
5 4 9 55.6%
Manufacturing 20 Food & Kindred Products
3 9 12 25.0%
21 Tobacco Products 0 1 1 0.0%
23 Apparel & Other Textile Products
3 1 4 75.0%
25 Furniture & Fixtures 1 0 1 100.0%
26 Paper & Allied Products 0 5 5 0.0%
27 Printing & Publishing 1 4 5 20.0%
28 Chemical & Allied Products
8 27 35 22.9%
29 Petroleum & Coal Products 4 5 9 44.4% 30 Rubber & Miscellaneous Plastics Products 1 5 6 16.7%
31 Leather & Leather Products
2 0 2 100.0%
32 Stone, Clay, & Glass Products 5 5 10 50.0% 33 Primary Metal Industries 4 9 13 30.8% 34 Fabricated Metal Products 0 2 2 0.0%
35 Industrial Machinery & Equipment
5 15 20 25.0%
36 Electronic & Other Electric Equipment
1 13 14 7.1%
37 Transportation Equipment
6 11 17 35.3%
38 Instruments & Related Products
3 7 10 30.0%
Division Two-digit SIC
!
!
!
39 Miscellaneous Manufacturing Industries 0 1 1 0.0% Transportation & Public Utilities 42 Trucking & Warehousing 0 3 3 0.0% 44 Water Transportation 1 0 1 100.0% 45 Transportation by Air 0 5 5 0.0% 47 Transportation Services 2 4 6 33.3% 48 Communications 2 20 22 9.1%Wholesale Trade 50 Wholesale Trade - Durable Goods
1 1 2 50.0%
51 Wholesale Trade - Nondurable Goods
1 3 4 25.0%
Retail Trade 52 Building Materials & Gardening Supplies 1 1 2 50.0% 53 General Merchandise Stores 0 3 3 0.0% 54 Food Stores 4 2 6 66.7%
55 Automative Dealers & Service Stations
1 0 1 100.0%
56 Apparel & Accessory Stores 2 0 2 100.0% 57 Furniture & Homefurnishings Stores 0 1 1 0.0%
58 Eating & Drinking Places
1 0 1 100.0%
59 Miscellaneous Retail 1 0 1 100.0%
Services 70 Hotels & Other Lodging Places 0 1 1 0.0% 73 Business Services 3 10 13 23.1% 78 Motion Pictures 0 1 1 0.0% 79 Amusement & Recreation Services 0 1 1 0.0% 80 Health Services 0 2 2 0.0% 87 Engineering & Management Services 0 2 2 0.0% Public Administration & Nonclassifiable 99 Non-Classifiable Establishments 2 6 8 25.0% Total 75 197 272
Industry description Family firms
Frequency Non-family firms Frequency Total firms Frequency Percent family firms in industry Division Two-digit SIC
!
!
Appendix B!
Robustness test. Linear robust regression results of family firms vs. non-family firms and firm performan-ce, where the outliers are excluded. The t-values are included within the parentheses underneath the coefficients. The stars behind the t-values indicate whether the above standing coefficient is significant on
a one (***), five (**), or ten (*) percent two-tailed ! significance level.!
! ! ! ! !
!
!
!
!
ROA - net income ROA - EBITDA ROEIntercept 0.0923 0.2151 -0.1579 (3.61)*** (6.89)*** (-1.90)* Family firm 0.0119 0.0186 0.0364 (4.56)*** (5.88)*** (4.29)*** Ln (total assets) -0.0139 -0.0348 0.0167 (-9.59)*** (-19.83)*** (3.55)*** Ln (cap. ex.) 0.0094 0.0279 0.0094 (7.86)*** (19.26)*** (2.43)**
Long term debt
ratio -0.0857 -0.0276 -0.0959
(-13.61)*** (-3.77)*** (-4.68)***
Sales growth -0.0010 -0.0045 -0.0016
(-1.05) (-3.73)*** (-0.51)
S&P Euro Plus 0.0280 0.0385 -0.0123
(8.44)*** (9.59)*** (-1.14)
SIC - Division included included included
Year dummy included included included
R-squared (%) 9.71 18.75 3.12
Number of
Appendix C!
Robustness test. Linear robust regression results of family firms vs. non-family firms and their performan-ce during crisis years, where the outliers are excluded. The t-values are included within the parentheses underneath the coefficients. The stars behind the t-values indicate whether the above standing coefficient
is significant on a one (***), five (**), or ten (*) percent two-tailed significance level.!
!
ROA - net income ROA - EBITDA ROE
Intercept 0.1249 0.2148 -0.0566 (12.29)*** (6.88)*** (-1.71)* Family firm 0.0134 0.0199 0.0366 (3.91)*** (4.81)*** (3.28)*** Crisis -0.0041 0.0379 -0.0170 (-1.06) -1.3100 (-1.35) Family*Crisis -0.0025 -0.0022 -0.0003 (-0.67) (-0.50) (-0.02) Ln (total assets) -0.0139 -0.0348 0.0168 (-9.59)*** (-19.83)*** (3.55)*** Ln (cap. ex.) 0.0094 0.0279 0.0094 (7.87)*** (19.27)*** (2.43)**
Long term debt ratio -0.0857 -0.0275 -0.0959
(-13.61)*** (-3.75)*** (-4.68)***
Sales growth -0.0010 -0.0044 -0.0016
(-1.03) (-3.71)*** (-0.51)
S&P Euro Plus 0.0280 0.0385 -0.0123
(8.44)*** (9.58)*** (-1.14)
SIC - Division included included included
Year dummy included included included
R-squared (%) 9.73 18.75 3.12
Number of
