Abstract
M.Com.
The objective of this study is to investigate the sensitivity of the South African commercial banks to the market, interest rate and exchange rate risk exposures. The study estimates a GARCH (1,1) model using the above variables including their conditional variances. To investigate the impact of the risk premium factor, a GARCH-in-Mean model with implied volatility of the exogenous variables as explanatory variables is used. The research relies on data of the JSE Top 40 companies and the major commercial banks. The data series ranges from 2003 to 2016. Using the TED spread, the data is split into three sub-samples the period prior to the crisis, during the crisis and the post-crisis period. It was found that the bank stock returns are sensitive to the market, interest rate and exchange rate risk. The banks are found to be influenced mostly by the 1-year and 10-year rates during the low volatility periods, while during the crisis period the impact extends to even shorter periods of 1-month, 3-month and 6-month yields. Furthermore, the banks are found to be more sensitive to the exchange rate during the low volatile periods, while the small banks are the most affected during the high volatility periods.
Regarding the conditional variance, the study found that the bank stock returns follow a GARCH generating process. Furthermore, the study found that the conditional volatility from the GARCH-in-Mean model was irrelevant in pricing the bank stocks during the high volatility periods. The conditional variance of the GARCH-M was estimated with an inclusion of the implied volatility of the exogenous variables: market, interest rate and foreign exchange rate returns. The study found that the parameters have a very low significance overall and the impact of the volatility from the market and foreign exchange rate tended to decline during the high volatility period; while the effect of the interest rate volatility rises during the same period.