The role of the political cycle in the relationship between economic policy uncertainty and the long-run volatility of industry-level stock returns in the United States

In this study, we investigate how US economic policy uncertainty (EPU) drives the long-run components of volatilities in industry-level stock markets. We use a modified specification of GARCH-MIDAS and find that EPU increases the long-run volatility of the industrials and materials industries and decreases it in 4 of the 10 industries considered here: consumer staples, healthcare, information technology and materials. In addition, we add a dummy variable for the political cycle (PLC) to study whether the relationship between EPU and the volatility of industry returns is significantly different under different political regimes. The results imply that a Republican presidency dampens the effects of EPU on the long-run volatility of the consumer staples, healthcare and information technology industries. We also decompose the aggregated EPU into 11 category-specific EPUs to explore the detailed relationship between category-specific EPU and long-run volatility driven by aggregate EPU. The results for the category-specific EPU are consistent with the findings for the aggregate EPU. In particular, the weakened effect of PLC on the relationship between EPU and the long-run volatility of industry-level returns is also confirmed by MIDAS regression with beta weight scheme.