This paper examines the issue of volatility transmission between stock market and foreign exchange returns for three emerging African countries including, Botswana, Kenya and South Africa using the multivariate VAR-EGARCH model. The application of the multivariate VAR-EGARCH model enables us to determine the degree of volatility transmissions and asymmetry between the stock and foreign exchange markets. The data consist of monthly observations on stock market returns and foreign exchange returns spanning the period February 1997 through January 2006. The data for foreign exchange rates were collected from the International Financial Statistics published by the International Monetary Fund. The stock market data were obtained from the Datastream. The foreign exchange returns are proxied by changes in foreign exchange rates. Specifically, the paper uses the modified augmented Dickey-Fuller unit root tests (DF-GLS) to determine the time series properties of the return series for the sample countries. The standardized residuals from an AR(1)-EGARCH(1,1) model were used as volatility measures. To ascertain the degree of volatility transmission between the stock and foreign exchange markets, the study applies the VAR-EGARCH(p,q) model. The results from the DF-GLS unit root tests indicate that the stock market and foreign exchange return series are level stationary. The preliminary results from the multivariate VAR-EGARCH model show significant evidence of lead-lag relationship between stock market returns and foreign exchange returns for the sample countries. The results also indicate the existence of significant volatility spillover effect between the stock and foreign exchange markets for Botswana, Kenya and South Africa.

JEL Classification:  C32, F31, GI5 Keywords: EGARCH, volatility transmission, stock returns, foreign exchange returns, asymmetry