WHICH EPU HAS BETTER PREDICTIVE POWER IN FORECASTING THE RETURN AND VOLATILITY OF STOCK INDICES

Purpose- This study examines the predictive power of domestic and foreign Economic Policy Uncertainty (EPU) indices-specifically from the US, China, Europe, and globally-on the return and realized volatility (RV) of 20 global stock indices using monthly data spanning over 25 years. Methodology- The dynamic connectedness method is applied to analyze the spillover effects of EPUs across stock markets. Findings- The empirical results reveal that EPU impacts RV more significantly than returns, with Global and US EPUs emerging as primary drivers across most regions and periods. Notably, Chinese EPU consistently exhibits minimal influence, while local EPUs have pronounced effects in economies with heightened political, economic, and financial uncertainties. Contrary to expectations, EPUs from major trade partners and regional EPUs do not exhibit superior predictive power compared to Global and US EPUs. These insights are critical for investors, risk managers, and policymakers in optimizing strategies amidst evolving economic uncertainties. Conclusion- The findings indicate that Economic Policy Uncertainty (EPU) has a stronger influence on realized volatility (RV) than on stock returns, with Global and US EPUs serving as the dominant predictors across most regions and periods. In contrast, Chinese EPU consistently has minimal impact, while local EPUs play a significant role in countries with pronounced political and economic instability. Surprisingly, EPUs from major trade partners and regional EPUs do not outperform Global and US EPUs in predictive power. These insights underscore the importance of focusing on global and US-driven uncertainties for strategic decision-making by investors, risk managers, and policymakers. Keywords: Economic policy uncertainty (EPU), stock indices, return, realized variance, dynamic connectedness. JEL Codes: C32, C58, D80, G12, G17