«BI Norwegian School of Management Hand-in date: 01.09.2011 Thesis supervisor: Øyvind Bøhren Program: Master of Science in Business and Economics ...»
5.2 Endogeniety Obviously, our study might be prone to the criticism that our model has omitted a variable that is correlated with the explanatory variables – and that it actually is one of those which are related to the origin of the the eﬀects which are reported. This problem can be labeled Endogeniety and the implication might be that coeﬃcients obtained from OLS is biased and of little value. Fortunately, panel data is able to exploit both the time dynamics in addition to the cross sectional information to control for non-observable variables. Another form for endogeniety that concerns the literature on family ﬁrms and growth is that it is hard to determine causality. Causality requires direction, sign and magnitude. In our case, we believe it is hard to determine whether growth causes family ownership, or vice versa. Since we are not aware of any speciﬁc exogenous shocks in our family ownership data, we will restrict ourselves from proposing any causality. The endogenity problem is a problem for both the independent and the dependent variable.
5.3 Fixed Eﬀects Model (FEM) When the number of cross sectional units is large and the number of periods is small, FEM and random eﬀects models (REM) can diﬀer signiﬁcantly. In FEM, 5 Methodology 16 the statistical inference is conditioned on the observed cross-sectional units in the sample. This is appropriate if we strongly believe that the individual or cross sectional units in our sample are not random drawings from a larger sample. We believe that this is the case in our sample, and hence we infer that FEM is an appropriate model to take into consideration (Gujarati, 2003, 650).
5.4 Descriptive statistics Descriptive statistics of the most important variables on overall sample is presented in table 2.. The growth variable variation is high even when we have excluded the smallest companies from our sample. The standard deviation is 26% with an average reel growth of 6%. The family ownership concentration is quite high with an average of 81%. This means that most of the companies that are non-listed are controlled by a family. The size of the companies also diﬀer a lot. Where the largest company has an asset side of 61 billion compared to the smallest of 179 thousand. The average company is 15,76 years, but it varies from 0 to 152 years. The non-family ﬁrms is 1,43 years older than the family ﬁrms. The sample consists of many more family ﬁrms than non-family ﬁrms, which is consistent with Berzins, Bohren, and Rydland (2008). Furthermore, family ﬁrms have a higher debt/asset ratio than non-family ﬁrms, which is in-line with our a priori expectation. This might indicate that family ﬁrms increase their debt in order to ﬁnance their projects.
In table 4 and ﬁgure 1, we show that family ﬁrms as a fraction of sample reduces with size. In the lowest quartile the fraction of family ﬁrms is 95%.
Conversely, the highest quartile has a fraction of just 60 %. This could imply that family ﬁrms reduce their risk as they grow.
In line with our hypothesis, we observe that family ﬁrms grow slower than non-family ﬁrms when we look at real growth after inﬂation (2). There is clearly a discrepancy in the growth rates, and it seems to be time-varying at ﬁrst sight.. Furthermore, it is interesting to note that the gap narrows in the business cycle contraction after the IT bubble (2001 to 2003) but it seems to be expanding during the ﬁnancial crisis (2007-2009). Overall, both growth rates are going in the same direction, but the time varying behavior of the gap is strange though it might be consistent with our ﬁnancial constraint argument, because during the ﬁnancial crisis, the credit spread increased to levels much higher than during the IT-bubble. Because family ﬁrms are more indebted, it is reasonable that growth-gap is wider during the ﬁnancial crisis than. t to but the gap might indicate that family ﬁrms growth is more robust during time of recession. We may suspect that the gap also can be caused by the higher degree of diversiﬁcation and the longer time horizon as mentioned in 5 Methodology
Fig. 2: Family ownership without CEO This ﬁgure shows growth year by year. The sample period is 2001 to 2009.
Growth is yearly growth in operating revenue, In the sample the parent ﬁrms, companies less that 15 million in revenue, ﬁrms with no employees and ﬁnancial companies is excluded. Growth is adjusted for inﬂation.
Multicollinearity is the presence of high degree of correlation among the independent variables (Miller, 2005). Table 7 on page 24 exhibit that the highest correlation is between the industry dummies for manufacturing and trade. and the family dummy and the CEO dummy (with a correlation coeﬃcient of respectively -0.356 and 0.296). It is also not a high correlation between the variable. Based on the correlation matrices. it does not seem that there is an imminent multicollinearity problem.
5.5 Regressions We have intable 5 shown the results from our panel data test. Regression 1 to 4 shows the relationship between family gatherings and growth. In 5 and 6, we introduce ﬁnancial performance. In 7 and 8 we add the variable for the debt ratio. In the last two regressions we include the variable for distinguishing between management and ownership. We have chosen to look at real growth and therefore inﬂation-adjusted growth variable comes into perspective.
Family company variable is negative and signiﬁcant at 1% level in all regression. It varies from -0,0183 to -0,0211.Size and age have a negative coeﬃcient.
Size varies from -0,1232 to -0,1369. Age varies from -0,0003 to -0,0005. Size is signiﬁcant on 1% level in all regressions. Age is not signiﬁcant in any of the regressions. Financial performance is positive and signiﬁcant on 1% level.
The variable ranges from 0.4771 to 0.5341. The leverage ratio is positive and signiﬁcant on 1% level. The variation of the coeﬃcient ranges between 0.1143 and 0.1244. The CEO variable is negative and signiﬁcant on 5% level. The coeﬃcient is -0,0119.
The family dummy variable is negative supporting our hypothesis. Family 5 Methodology 21 companies grow slower than non-family ﬁrms. Size and age have a negative eﬀect on growth. This is consistent with what ﬁrm dynamics studies have shown. The age variable is non signiﬁcant. This is possibly because the size variable absorbs a lot of the same things as the age variable. The ROA is positive showing that higher ﬁnancial performance generate higher growth.
The puzzling ﬁnding is that higher leverage gives higher growth. This is not in line with our thought that ﬁnancial constrained ﬁrms grow slower. Huynh and Petrunia (2010) explains this with fast growing ﬁrms lever up to get most out of their investment opportunities. As we thought, the alignment of management and control create a negative growth.
The regression result indicate that family ﬁrms grow slower than non-family ﬁrms. This is, after adjusting for common ﬁrm dynamic variables. Financial strength is important, but the ﬁrms with high leverage grow faster. Therefore the A3 problem seems to be less important for ﬁrm growth. The A1 problem is an important issue, but the regression shows that it not the sole reason for the lower growth.
5.6 Robustness5.6.1 Alternative methodology
There are several alternatives to the ﬁxed eﬀects model. We have conducted the random eﬀect model on the sample in table 6. We have tested for the appropriateness of FEM using the Hausman Test for Correlated Random Eﬀects.
This test compares a Random eﬀect model with the FEM Random Hausman.
The results are presented in table 6. The Hausman test yields a p-value of 0 which allows us to reject the null hypothesis, i.e. that a random eﬀect model is appropriate, and we prefer using our FEM model.
5.6.2 Alternative model speciﬁcation
Tab. 7: Correlation and multicollinearity This table shows correlation matrix for the dependent variable Growth, the independent variables Debt/asset, ROA, and various control variables for Norwegian limited liability ﬁrms in the period 2000-2008. Growth is yearly growth in operating revenue, Firm age is the number of years from establishment until the observation year, Firm size is total asset Leverage is debt over total assets. In the sample the parent ﬁrms, companies less that 15 million in revenue, ﬁrms with no employees and ﬁnancial companies is excluded. Growth is adjusted for inﬂation.
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) Family 1.000 Age -0.034 1.000 Total asset -0.119 0.039 1.000 Debt/Asset 0.067 -0.127 -0.028 1.000 ROA 0.050 0.005 -0.011 -0.270 1.000 CEO 0.296 -0.009 -0.080 0.035 0.063 1.000 Agriculture -0.028 -0.017 0.027 -0.028 -0.016 -0.007 1.000 Construction 0.069 0.014 -0.022 -0.005 0.052 0.031 -0.062 1.000 Energy -0.129 0.006 0.053 -0.049 -0.008 -0.069 0.130 -0.024 1.000 Manufacturing -0.084 0.078 0.046 -0.098 -0.033 -0.074 -0.089 -0.160 -0.035 1.000 Multi -0.031 0.029 0.053 -0.021 -0.012 -0.006 -0.041 -0.073 -0.016 -0.105 1.000 Service -0.002 -0.050 -0.020 0.069 -0.041 -0.041 -0.103 -0.185 -0.040 -0.266 -0.122 1.000 Trade 0.074 -0.020 -0.043 0.038 0.052 0.097 -0.119 -0.214 -0.046 -0.307 -0.141 -0.356 1.000 Transport -0.066 0.005 0.018 0.046 -0.012 -0.059 -0.031 -0.056 -0.012 -0.080 -0.037 -0.093 -0.108 1.000
6 Conclusion 25
5.6.3 Conclusion on robustness
We have tested for an alternative model and proxies in this section. The results from the alternative model is similar to our base case. The FEM seems to be the better model-choice when compared to REM. The alternative proxy ∆asset generates negative growth for family ﬁrm. This is consistent with the result with operating revenue as dependent variable. This section shows that our result is robust and consistent even when we change dependent variable and model.
We have found indications that family ﬁrms grow diﬀerently compared to nonfamily ﬁrms. Furthermore, we have also seen that these family ﬁrms vary much more in their growth. The regression shows that a part of the slower growth could be explained by the separation of ownership and control. We ﬁnd no support for our suspicion that ﬁnancial constraints were the determining factor in the diﬀerence between growth of family ﬁrms and non-family ﬁrms.
When we were adjusting for the known controls, such as age, industry and ﬁrm size, there was still some residual eﬀects of family ownership on ﬁrm growth that is unexplained. We found the diﬀerence between non-family ﬁrms and family ﬁrms growth rates to be time-varying and consistently positive.
Our paper shows that exploitation of the minority owners could be an issue in non-listed companies. Family ﬁrm ownership could be beneﬁcial for the family, but not for the society.
6 Conclusion 26References Berzins, J., O. Bohren, and P. Rydland (2008): “Corporate ﬁnance and governance in ﬁrms with limited liability: Basic characteristics,” CCGR Research Report, (01/2008).
Bohren, O. (2011): Eierne, styret og ledelsen. Fagbokforlaget.
Coad, A. (2007): “Firm growth : a survey,” Documents de travail du centre d’economie de la sorbonne, Universite Pantheon-Sorbonne (Paris 1), Centre d’Economie de la Sorbonne.
Coad, A. (2009): The growth of ﬁrms: a survey of theories and empirical evidence, New perspectives on the modern corporation. Edward Elgar.
Coad, A., and W. Holzll (2010): “Firm growth: empirical analysis,” Papers on Economics and Evolution 2010-02, Max Planck Institute of Economics, Evolutionary Economics Group.
Coad, A., and J. P. Tamvada (2008): “The Growth and Decline of Small ﬁrms In Developing Countries,” Papers on Economics and Evolution 2008Max Planck Institute of Economics, Evolutionary Economics Group.
Commission, E. (2003): “COMMISSION RECOMMENDATION of 6 May 2003 concerning the deﬁnition of micro, small and medium-sized enterprises,” Oﬃcial Journal of the European Union.
Evans, D. S. (1987): “Tests of Alternative Theories of Firm Growth,” The Journal of Political Economy, 95(4), pp. 657–674.
Grossman, S. J., and O. D. Hart (1980): “Takeover Bids, the Free-Rider Problem, and the Theory of the Corporation,” Bell Journal of Economics, 11(1), 42–64.
Gujarati, D. N. (2003): Basic Econometrics fourth edition.
Hamelin, A. (2009): “Do small family businesses have a peculiar attitude toward growth? Evidence from French SMEs,” Working Papers CEB.
Huynh, K. P., and R. J. Petrunia (2010): “Age eﬀects, leverage and ﬁrm growth,” Journal of Economic Dynamics and Control, 34(5), 1003–1013.
Koller, T., M. Goedhart, D. Wessels, T. Copeland, McKinsey, and Company (2005): Valuation: measuring and managing the value of companies, Wiley ﬁnance series. John Wiley & Sons, Inc.
Meulbroek, L. (2001): “The Eﬃciency of Equity-Linked Compensation: Understanding the Full Cost of Awarding Executive Stock Options,” Financial Management, 30(2).
Miller, J. (2005): The Chicago guide to writing about multivariate analysis, Chicago guides to writing, editing, and publishing. University of Chicago Press.
Morck, R., D. Wolfenzon, and B. Yeung (2005): “Corporate governance, economic entrenchment, and growth.,” Journal of Economic Literature, 43(3), 655–720.
Myers, S. C. (1977): “Determinants of corporate borrowing,” Journal of Financial Economics, 5(2), 147 – 175.
Porta, R. L., F. Lopez-De-Silanes, and A. Shleifer (1999): “Corporate Ownership Around the World,” Journal of Finance, 54(2), 471–517.
Schulze, W. S., M. H. Lubatkin, R. N. Dino, and A. K. Buchholtz (2001): “Agency Relationships in Family Firms: Theory and Evidence,” Organization Science, 12(2), 99–116.
Sraer, D., and D. Thesmar (2007): “Performance and Behavior of Family Firms: Evidence from the French Stock Market,” Journal of the European Economic Association, 5(4), 709–751.
Villalonga, B., and R. Amit (2010): “Family Control of Firms and Industries,” Financial Management, 39(3), 863–904.