Forecasting Pakistan stock market volatility: Evidence from economic variables and the uncertainty index
Maria Ghani,Qiang Guo,Feng Ma,Tao Li
DOI: https://doi.org/10.1016/j.iref.2022.04.003
2022-04-01
Abstract:This study examines the impact of the economic policy uncertainty index (EPU) and macroeconomic variables on the volatility of the Pakistan stock market using the GARCH-MIDAS (mixed data sampling) model. The model allows us to observe whether those variables contain valuable information to forecast stock market volatility. Our empirical findings show several outcomes. First, our out-of-sample results show economic policy uncertainty index has predictive power to forecast Pakistan stock market volatility. Second, among all variables, oil prices are the most powerful predictor of volatility with a higher out of sample R square value. Third, all macroeconomic variables including exchange rate, short-term interest rate, money supply M2, foreign direct investment, gold prices, inward remittances, industrial production, and consumer price index (proxy for inflation) contain useful information for stock market volatility forecasting. However, the long-run interest rate is an ineffective indicator of volatility during the sample period study. Finally, we find that the combination forecast information is also useful for volatility forecasting.
economics,business, finance