Cross-sectional uncertainty and stock market volatility: New evidence

This study investigates the potential of cross-sectional uncertainty (CSU) to predict stock market volatility. Empirical findings reveal that the newly developed variance-based index conditioned on economic policy uncertainty exhibits greater predictive power than the widely used economic policy uncertainty index. Sparse methods consider multiple predictors and perform well. Further research has demonstrated that the CSU index contains more valuable information and delivers better predictive performance and economic value, especially under financial crises. Our study extends the application of the CSU index and provides novel evidence for volatility prediction.

Automated summary

This study explores the potential of cross-sectional uncertainty (CSU) in predicting stock market volatility. Empirical findings demonstrate that a newly developed variance-based index, conditioned on economic policy uncertainty, has stronger predictive power compared to the widely used economic policy uncertainty index. Sparse methods that consider multiple predictors perform well in this context. Further research shows that the CSU index contains valuable information and delivers better predictive performance and economic value, especially during financial crises. This study extends the application of the CSU index and provides novel evidence for volatility prediction.