Ownership Structure and Company Performance - Assignment Example

Ownership structure and company performance (A report to the Board of Directors of JP Drew) I. Introduction This report investigated the relationship between ownership structure and company performance based on an available data set. The literature has produced a number of studies on the relationship between company structure and performance. For example, this is the findings of Mudambi and Nicosia (1998, p. 180), Gursoy and Aydogan (1998), and Welch (2003). Welch (2003) is one of the latest and concluded that company structure affects firm performance using regressions with adjusted R squares of only 0.30 or 30%. The adjusted R square indicate the percentage of variation in the dependent variable can be explained by the variation of the independent variables. For the purpose of making a similar investigation, this study used data involving 500 cases. Data are available on market-to-book-value (MBV), identity of the largest owner, “concentration of ownership” assessed in terms of the percentage share of the largest owner, size in terms of the total assets of the company, return on capital in percent (ROCE), and industry type. Therefore, there were six variables. However, because this investigation is limited to the relationship between ownership structure and company performance, there were only four relevant variables: MBV, identity, concentration of ownership, and ROCE. The relevant independent variables for the study are “identity of the largest owner” and the “concentration of ownership” assessed in terms of the percentage share of the largest owners. In contrast, the relevant dependent variables or company performance variables appropriate from the data set are MBV and ROCE. There was no need to use the variable “industry” because it is irrelevant for the focus of our study. Of course, our study can also cover whether the effect of the independent variables differ in each type of industry or whether the specific effect of the independent variables vary depending in manufacturing, services, and primary industries. However, we need not do this. Similarly, we do not have to cover firm size based on assets because firms can be small or large but asset size does not imply performance. Of course, we can also construct empirical investigations that differentiate the effects based on asset size but, again, we do not have to do this. II. Select graphical analysis We focus in this section on variables that may indicate a possible presence or absence between firm ownership structure and firm performance. In Figure 1, we investigate the scattergram pattern if we have the percentage share of the largest owner of a firm and market to book value ratio of the firm. Based on the pattern revealed by Figure 1, it seems obvious the percentage share of the largest owner of the firm and firm performance as indicated by the market to book value ration of the firm are positively related. Figure 1. Scattergram: percentage share of largest owner and market to book value ratio Source: Excel construction of the scattergram on the data set Doing the same between the percentage share of the largest owner of the firm and percentage return to capital or ROCE, we obtain Figure 2. Figure 2. Percentage share of largest owner of firm and returns to capital in % Source: Microsoft Excel’s construction of the scattergram based on the data set Thus, Figures 1 and 2 indicate that company structure is associated with company performance. Figure 1 show that concentration of ownership as indicated by the percentage share of the largest owner of the firm is positively associated with the market to book value of the firm ratio. On the other hand, Figure 2 indicate the firm concentration is positively associated with percentage return to capital of firms. III. Bivariate analysis We can check the relationships suggested by Figures 1 and 2 through bivariate statistics using correlation statistics as per computations of the SPSS or the Statistical Package for the Social Sciences. Table 1. Pearson-r correlation coefficient and significance table at a 1-tail test of select variables Variable(s) Book to market value Percentage return on capital Percentage of ownership of top owner of the company .709 (Significance 1-tail = 0.000) N=500 0.584 (Significance 1-tail = 0.000) N=500 Source: SPSS output Table 1 indicates that the variable “percentage of ownership of the top owner of the company” is positively and significantly correlated with book to market value at a Pearson r correlation coefficient of 0.709. At a 1-tail test, the significance of the Pearson r correlation coefficient is 0.000 that implies that the Pearson r correlation coefficient is significant at critical α = .001. This means that the correlation between the two variables is highly significant. Given the significance level, this means we can reject the null hypothesis of zero correlation to accept the alternative hypothesis of positive correlation at critical α = .001. A similar story can be made between the correlation between the variable“percentage of ownership of the top owner of the company” and percentage returns to capital. The Pearson r correlation coefficient between the two variables is 0.585 which has a significance of 0.000, implying that at critical α = .001, we can reject the null hypothesis of zero correlation to accept the alternative hypothesis of positive correlation between the two variables. IV. Multivariate analysis Multiple regressions constitute one type of multivariate analysis (Hair et al. 1995, p. 13). The basic format of multiple regressions is Y1 = X1 + X2 + … + Xn (Hair et al. 1995, p. 79). This is the format that we will follow for equations 1 and 2 as well as regression equations 1 and 2 that will be based on equations 1 and 2. In turn, regression equations 1 and 2 will be estimated using Microsoft Excel’s statistical function facilities and the data will be obtained will be reflected as regression equation estimates 1 and 2. Thus, the regressions estimates that will be obtained for regression equations 1 and 2 are the multiple regressions that are specific forms of multivariate regression based on Hair et al. (1995, p. 13). The fundamental regression functions for investigation are as follows: Equation 1: MBV = f (identity, concentration of ownership) Equation 2: ROCE = f (identity, concentration of ownership) MPV and ROCE are measures of company performance. In contrast, identity and concentration of ownership are measures or variables associated with “ownership structure.” The f in Equations 1 and 2 refers to “function”. MBV and ROCE are both continuous or ratio variables. “Concentration of ownership” is also a continuous or ratio variable. However, the variable “identity” or “identity of the largest owner is” is a nominal or discrete non-orderable variable. For the purpose of regression estimation, we can convert variable “identity of the largest owner” or identity into a four dummy variables: Dummy Variable 1 (“dum1”): Equal to “1” if bank but zero otherwise Dummy Variable 2 (“dum2”): Equal to “1” if institutional investor but zero otherwise Dummy Variable 3 (“dum3”): Equal to “1” if non-financial company but zero otherwise Dummy Variable 4 (“dum4”): Equal to “1” if family or single-person owned It is easy to see that our dummy variables are mutually exclusive. Using the four dummy variables, we converted equations 1 and 2 to regression equations 1 and 2: Regression Equation 1: MBV = f (identity, dum1, dum2, dum3, dum4) Regression Equation 2: ROCE = f (identity, dum1, dum2, dum3, dum4) We can use either the SPSS or Microsoft Excel’s statistical functions to compute for the regression statistics. For a few statistics, Microsoft Excel is just as good as the SPSS. Using Microsoft Excel to compute the regression, we obtained regression estimates1 and 2: Regression Estimate 1. Dependent variable “Market to Book Value Ratio”  Regressors or Independent Variables Estimate of Coefficients Standard Error t Stat P-value Intercept 0.4143 0.0643 6.4471 0.0000 Con 0.0373 0.0016 23.1864 0.0000 dum1 0.1624 0.0295 5.5027 0.0000 dum2 0.0937 0.0300 3.1232 0.0019 dum3 0 0 65,535 n.a. dum4 -0.0412 0.0321 -1.2853 0.1993 Regression Statistics Adjusted R Square 0.5454 Standard Error 0.2409 Observations 500 Regression Estimate 2. Regression data for dependent variable ROCE or “return on capital in percent”  Regressors or Independent Variables Coefficients Standard Error t Stat P-value Intercept 17.5112 1.4961 11.7044 0.0000 Con 0.5962 0.0374 15.9221 0.0000 dum1 -0.1677 0.6872 -0.2440 0.8073 dum2 0.1921 0.6984 0.2751 0.7834 dum3 0 0 65535 n.a. dum4 -0.6266 0.7467 -0.8391 0.4018 Regression Statistics Adjusted R Square 0.3352 Standard Error 5.6080 Observations 500 Regression Estimate 1 indicates that except for dummy variables 3 and 4, all independent variables significantly explain the market value to book value ratio of a firm for around 500 companies. The significance of an independent variable in explaining the dependent variable is indicated by the P-value of the independent variable. P-values lower than 0.10 can be considered significant in explaining the value of the dependent variable by the independent variable. Thus, based on this the variable concentration or Con or the percentage share of the largest owner can significantly explain the market to book value of the company. Regression Estimate 1 suggests that the greater the concentration of ownership of the largest owner, the higher the performance of the firm in terms of market to book value. Similarly, Regression Estimate 1 also suggests that when the largest owner of the firm is a bank or an institutional company, the higher the market to book value of a firm. However, dummy variable 3 yielded a zero coefficient, implying that when the largest owner is non-financial company, there is no contribution to the market to book value of a firm. Regression Estimate 1 indicates that based on the values for dummy variable 4 or when the top investor is an individual or family, the market to book value of the firm is negatively affected but the effect, however, is not significant based on the p-value 0.1993. The adjusted R square value for regression equation 1 is 0.5511, suggesting that the independent variables can jointly explain about 55.1% of the variations in the dependent variable On the other hand, Regression Estimate 2 shows that except for the variable “concentration of ownership” or the ownership share of the largest owner of the firm, all dummy variables either do not explain the significantly or has a zero coefficient. Based on the adjusted R square, the regression equation can only explain only about 33.52% of the variations in the dependent variable. Thus, Regression Estimate 2 indicate that ownership structure based on whether banks, institutional investors, non-financial organizations, and whether families or individuals owned the firm cannot explain the dependent variable “ROCE” or “return on capital in percent”. This indicates that with regard to ownership structure, it is only the concentration of the largest owner that positively affects the “return on capital in percent”. Thus, in summary, based on the two regressions, we have good evidence to claim or conclude that ownership structure affects corporate performance. V. Conclusion Thus, based on the foregoing discussion, there is adequate empirical foundation to claim that the company structure affects firm performance. At the same time, however, we cannot adequately dismiss an alternative view on the data: profitability or firm performance may be the ones promoting corporate concentration. In any case, statistical data are consistent with and even supports the view that company structure influences firm performance. References Gursoy, G. and Aydogan, K., 1998. Equity ownership structure, risk-taking and performance: An empirical investigation in Turkish companies. Paper presented in the International Global Finance Conference, Ankara, Turkey. Hair, J., Anderson, R., Tatham, R., and Black, W., 1995. Multivariate data analysis with readings. New Jersey: Prentice-Hall, Inc. Moore, D. and McCabe, G., 1999. Introduction to the practice of statistics. 3rd Ed. New York: W. H. Freeman and Company. Mudambi, R. and Nicosia, C., 1998. Ownership structure and firm performance: evidence from the UK financial services. Applied Financial Economics, 8, 175-180. Welsch, E., 2003. The relationship between ownership structure and performance in listed Australian Companies. Australian Journal of Management, 28 (3), 287-306. Read More