Economic Policy Uncertainty and Bank-Level Stock Returns Volatility of the United States: A Mixed-Frequency Perspective

Abstract

This paper examines the predictive power of the monthly Economic Policy Uncertainty (EPU) index for the daily return volatility of 149 banking stocks in the United States, over the period from August 2000 to August 2023. Using a Glosten, Jagannathan, and Runkle-generalized autoregressive conditional heteroscedasticity-mixed data sampling (GJR-GARCH-MIDAS) model, the results show that EPU plays a significant role in predicting the volatility of banking stocks. Across the groups of large, medium, and small banks, banking stock volatility tends to increase in response to EPU, highlighting the significance of heightened uncertainties surrounding future government economic policies for the volatility of banking stocks. EPU has a stronger impact on large banks, possibly due to their complexity, funding structure, perceived ‘too big to fail’ status, and systemic interconnectedness. The outperformance of the GJR-GARCH-MIDAS-EPU model holds in an out-of-sample analysis, regardless of market capitalization and forecast horizons.