Relationship Between Ownership Structure and Dividend Policy in the Sugar Industry in Kenya

Abstract

The study sought to establish the scrutinize the link between ownership structures and disbursement policy in the sugar business in Kenya; to measure the link between government ownership and dividend policy amongst sugar establishments in Kenya; and explore the link between private dividend policy and ownership amongst sugar establishments in Kenya. To realize the objectives the study used descriptive study design. The target population was eleven operational sugar factories in Kenya. The study utilized secondary information that entailed time series data gathered over a period of approximately 5 years; from 2013-2017. The information collected comprised of dividend pay-out ratio, proportion of common shares of the state divided by cumulative common shares in issue; proportion of common shares of the institution divided by cumulative common shares in issue; percentage of common shares held by foreign investors divided by cumulative common shares in issue; ROE, Liquidity ratio. Normality test was measured by the Shapiro Wilk test, The Shapiro Wilk test pointed out that the data points contained in the sample were not normally distributed. Heteroscedasticity was measured by the Breusch-Pagan (BP) test which showed that there was no heteroscedasticity in the data set imply that there was equal variability in the data set. Multicollinearity was measured by variance inflation factor. VIF values of government ownership, institutional ownership, foreign ownership, liquidity and pay-out ratio are less than 10 showing have fair multicollinearity. Profitability has a VIF greater than 10 indicating strong multicollinearity. The variables under study were analysed for their descriptive statistics which were presented in the form of mean, standard deviation, maximum, minimum and skewness. The findings as per the descriptive statistics revealed that other than liquidity all other variables had standard deviation greater than the mean implying a high volatility in the sense that a change in the data values implies unpredictability of the variable. Correlation analysis was used to determine whether a change in another variable accompanies a change in another variable. From the correlation matrix above, the findings reveal that there is a weak positive correlation(r=0.035) between dividend pay-out and government ownership. There is a weak positive correlation (r=0.080) between dividend pay-out and institutional ownership. Foreign ownership and dividend pay-out were weakly negative correlated (r=-0.053) with dividend pay-out. There was a weak positive correlation (r=0.191) between ROE and divided pay-out. Liquidity was also weakly positive correlated (r=0.191) with dividend pay-out. From the model summary the study deduced that the coefficient of determination r-squared was 0.22861. This suggests 22.861% of the variation in dividend pay-out is accounted for by model or that the model is 22.861% efficient in estimating the relationship. The study concluded that all the five independent variables had a positive effect on dividend pay-out. The extent of the positive impact on dividend pay-out varied form one variable to another as indicated by different coefficients. Nonetheless, ROE was concluded to have had a stronger influence on dividend pay-out. The study was also conducted with the aim of establishing associations between dividend pay-out and government ownership, institutional ownership, foreign ownership, ROE and liquidity. It was concluded that government ownership, institutional ownership, ROE and liquidity had weak positive correlation with dividend pay-out while foreign ownership was weakly negatively correlated with dividend pay-out.