Risk estimation in the thinly traded JSE environment

Abstract

Received evidence has come to the fore which suggests that the major source of bias in the estimation of beta coefficients on the JSE can be attributable to the thinly traded phenomenon. In this paper the suitability of a beta estimation procedure which corrects for the effects of thin trading is investigated for JSE stocks. The investigation reveals that implementation of the correction procedure results in substantial improvements in beta estimation. Furthermore, on the basis of the empirical investigation suggestions are made regarding the parameters that should be included in the beta estimator.Onlangse getuienis dui daarop dat die vernaamste bron van sydigheid van beta-koeffisiente op die JEB aan 'dun' handel toegeskryf word. In hierdie artikel word die toepasbaarheid van 'n skattingsprosedure wat vir die effek van 'dun' handel korrigeer, ondersoek. Die ondersoek dui aan dat die korrigeringsprosedure wesenlike verbeterings in die skatting van beta-koeffisiente lewer. Verder word op grond van 'n empiriese ondersoek aanbevelings gemaak betreffende die parameters wat in die skatter vir die betas ingesluit moet word.