Abstract
This study develops a framework for the diversification of a domestic portfolio exposed to systemic risk within a common financial market. The developed framework intends to examine the optimal allocation problem and the investors’ risk tolerance in two financial market uncertainty regimes obtained from the systemic risk measure. To this end, the study makes use of the Conditional Value at Risk (CVaR) based on the Extreme Value Theory (EVT) and the Generalized Autoregressive Heteroscedasticity (GARCH) model. The CVaR is thereafter used to create two sub-portfolios; i.e., the Adverse Returns Portfolio (ARP) and the Favorable Returns Portfolio (FRP). The ARP1 and the FRP2 represent the set of portfolio returns observed during a financial crisis due to systemic risk, and during normal financial market period respectively. A quadratic Mean-Variance portfolio optimization problem is then applied to these two types of portfolio returns in order to identify investment allocations and performances during financial crisis resulting from a systemic risk and during normal financial market period. Using a sample of daily log return series of nine Johannesburg Stock Exchange (JSE) sector indices; the findings of this study show that JSE sectors that are positively correlated with the benchmark index (All-Share Index: ALSI) tend to contribute more in maximizing the ARP while the sectors that are negatively correlated with the ALSI tend to maximize the FRP. Investors who are aware of the behavior of these two portfolios can protect their investment capital during financial crisis resulting from a systemic risk. Furthermore, the study finds that the efficient ARP has better performance measures than the benchmark. However, the inverse is true for the FRP. These findings are consistent with different levels of risk aversion considered in this study.
M.Com. (Financial Economics)