A Review of Systematic Risk Estimation on the JSE

Abstract

This paper reviews some recent developments in the area of beta estimation on the JSE. it turns out that the dual phenomena of extreme thin trading and mild segmentation between the industrial and mining boards on the JSE impact on the techniques used for efficient beta estimation. On the basis of a simulation study as well as an empirical investigation, the trade-to-trade approach to beta estimation with a Bayesian adjustment has shown itself to be superior to other thin trading correction procedures. Further evidence of the perceived segmentation between the industrial and mining sectors of the JSE is revealed in the pricing of securities. As a consequence it is argued that the relevant Financial and Industrial or Mining Index should be used as a market proxy for beta estimation. The non-normality of security returns and possible time variance of security betas is also discussed in the context of the JSE.