International stock market volatility: A data-rich environment based on oil shocks

This paper investigates the predictive ability of oil shocks for international stock market volatility based on a data-rich environment. Our empirical analysis shows that multiple oil shock measures contain valuable information for predicting stock market volatility, in addition to traditional economic variables and uncertainty indices. Moreover, based on the group 7 countries, the least absolute shrinkage and selection operator method and regime-switching model jointly deliver incremental improvement in forecasting accuracy from both statistical and economic perspec-tives. These results are confirmed by robustness checks under different business cycles and market conditions, including the COVID-19 pandemic.