Tail comovements of implied volatility indices and global index futures returns predictability

Abstract

This paper investigates whether the comovements across different asset classes of IV indices' tails are informative in predicting global stock index futures returns. The empirical results show that the upper tail of the condition distribution exhibits larger comovement in IV indices compared to the average and lower tail of the conditional distributions. Moreover, from a forecasting perspective, the comovements of the IV indices in the upper tails provide more informative insights for predicting subsequent returns on stock index futures, compared to the lower tails and average of the conditional distributions. Finally, the information extracted from both the lower and upper tails of the conditional distributions yields the best forecasting performance for predicting subsequent returns on stock index futures, surpassing the outcomes of solely focusing on upper or lower tails, as well as the average of the conditional distributions. Overall, our empirical results support the viewpoint suggested by Shefrin (2008) that the heterogeneity of investor beliefs plays a crucial role in asset pricing.