Return and risk performance: implications for the classification of South African unit trusts.

Abstract

This study focussed on investigating the similarities and difference in the return and risk performance within and across classification categories of South African unit trusts. This study was done with a view of considering the validity and usefulness of industry based classification, in unit trusts selection by investors as well as relative performance comparison by investors, regulators and fund managers alike. The sample was drawn from the equity classification category and NAV data was obtained over a 10-year period from January 2006 to December 2015 from INET BFA and Bloomberg. Using financial and statistical models, a one-way Welch’s analysis of variance (ANOVA) and Tukey’s test were used in addressing the proposition regarding similarities and differences in performance, while cluster analysis using the K-means non-hierarchical technique was utilized in addressing the movements of units trusts away from the initial industry based classification. The study showed that performance of unit trusts within classification categories are not significantly different in all categories tested, while performance across classification categories are also not significantly different in most of the categories. When classification of unit trusts was examined on the basis of return and risk, only the unit trusts in one of the classification categories, retained a composition similar to the industry based classification. This is in contrast to other unit trusts that showed migration across categories. In summary, the study provides evidence that supports the notion that performance of unit trusts in the equity classification category is mostly similar, suggesting possible limited diversification benefits when investing along the same asset type classification category.