A practical method to estimate cost of capital for South African firms

Abstract

The aim of this work was to test the validity and reliability of the clustering approach as a viable alternative practical tool to estimate cost of capital for JSE listed target firms. Starting from an initial (257 x 8) matrix of raw accounting data (from McGregor BFA) for selected firms, the original eight accounting variables were reduced to three common factors by exploratory factor analysis. The common factors, namely Profitability, Asset efficiency and Growth, were subsequently used as variables to cluster the firms into ten homogeneous groups. From the ten clusters that were extracted, a sample of 31 observations was randomly picked and a paired sample t-test performed to test the null hypothesis of zero difference between the mean of the calculated weighted average cost of capital (WACC) and the mean of cluster analysis-predicted WACC. Based on the results of the t-test, there was insufficient statistical evidence to reject the null hypothesis and accordingly the clustering approach was deemed to be both reliable and valid as an alternative tool to estimate cost of capital for target listed South African firms.