Abstract
This study defines systematic tail risk as a stock’s exposure to market tail events and assesses the impact of it on the cross section of returns from the Johannesburg Stock Exchange (JSE). To determine the extent to which systematic tail risk explains the cross section of returns in the JSE, the study estimates systematic tail risk by combining the statistical concepts of extreme value theory (EVT) and copula. Specifically, the study first characterizes stocks and market tail events under the Block model and subsequently proxies a stock’s systematic tail risk with parameter estimates of an extreme value copula fitted to the bivariate Generalized Extreme Value (GEV) distribution of stock and market tail events. Based on data on JSE All Share Index companies, provided by the JSE for the period of January 2002 through June 2018, results of the traditional asset pricing portfolios formation and crosssectional regressions show that the extreme value copula parameter adequately captures systematic tail risk in the JSE. More importantly, the results support the existence of a systematic tail risk premium in the JSE. Interestingly, the effect of systematic tail risk on the cross section of returns is time-varying and independent from that of risk measures such as beta and downside beta and firm characteristics such as book-to-market (BTM) ratio, size and past returns. In addition, the results provide evidence on the negative impact of the 2008 Global Financial Crisis on crash aversion in the JSE. The practical relevance of these results is of an utmost importance for both academics and finance professionals. The findings implicitly provide support for the downside risk framework as a legitimate perspective on investors’ perception of risk in equity markets and reveal a need to reconsider somehow disfavoured portfolio theories such as the safety-first criterion for asset pricing endeavours.
M.Com. (Financial Economics)