Abstract
In an efficient stock market the stock prices embody the expected course of local and global economy which might impact the future prices in some way. This paper examines whether the volatility and dynamic linkages of Pakistani stock market with the US stock market are improved if local and foreign macroeconomic variables are augmented in a Multivariate GARCH model. Using the BEKK specification of Engle and Kroner (1995) with local and foreign macroeconomic variables as exogenous variables we estimate a multivariate GARCH model and use the Wald test to investigate whether the stock market volatility is significantly changed due to the local and foreign economic conditions. The monthly stock returns and some key macroeconomic variables are employed from July, 1997 to December 2015. Forecasts are evaluated using three measures namely, R2 (coefficient of determination); Mean Absolute Percentage Error (MAPE) and Median Absolute Percentage Error (MdAPE). We also tested the sensitivity of forecast by using the global financial crisis (GFC) dummy to investigate whether the financial crisis has altered the volatility forecast. Although the both local and global variables significantly impact on the stock market volatility of Pakistan but the local macroeconomic variables contribute more than global to improve the forecast of Pakistan’s stock market volatility. The results are to some extent sensitive to inclusion of the GFC dummy.
