Exchange Rate Pass through to Stock Prices: A Multi GARCH Approach

This paper analytically examines the impact of exchange rate volatility on stock prices in Nigeria via both symmetric and asymmetric GARCH models. At the onset the descriptive statistics reveals that both series are non-normally distributed as indicated by the Jacque-Bera statistic, also the standard deviation implied that the stock price series is more volatile than the exchange rate. Furthermore both series are reported to be negatively skewed also reference to the kurtosis statistics presented it is observed that both series are leptokurtic distribution. Further the result obtained from the estimated model GARCH models reveals that the PGARCH gives the better fit of the stock prices volatility model given its minimum AIC value. In the symmetric models {GARCH (1, 1) and GARCH-in-Mean} the shocks on stock returns volatility are found to be mean reverting whilst in the asymmetric GARCH models {TGARCH, EGARCH and PGARCH} only EGARCH was found to be non-mean reverting. Further, the asymmetric term in all the 3 models indicates that bad news exerts more shocks on the stock returns volatility than good news of the same magnitude. The post estimation diagnostic test of ARCH effect demonstrate that all the models completely captured the ARCH effect. Immensely the findings of this study shall be of utmost relevance to investors, stock brokers, members of the academia, regulators and monetary authorities.