Predicting the tail behavior of financial times stock exchange/Johannesburg Stock Exchange (FTSE/JSE) closing banking indices : extreme value theory approach

The incidence of rare but extreme events appears to be significant in worldwide financial markets. In this chapter we apply extreme value theory (EVT) distributions to predict extreme losses of five South African (SA) financial times stock exchange/Johannesburg Stock Exchange (FTSE/JSE) closing banking indices. The effectiveness of risk measures for measuring risk of investment is also explored. A 5-day time series for the period of 02 January 2008 to 20 April 2018 is used. The MS(2)-EGARCH(1,1) showed that there is a regime persistence in all the banks, implying that the new obtained series is independently and identically distributed (i.i.d). It is therefore concluded that the generalized Pareto distribution (GPD) is a better distribution than the generalized extreme value (GEV) in estimating extreme loses and that the computation of economic capital using Glue-value-at-risk (VaR) is more conservative than using other risk measures under the GEV distribution.