Author(s): Rangan Gupta
This paper hypothesizes that corporate managers’ sentiment can predict aggregate stock market volatility. Using a k-th order nonparametric causality-in-quantiles test, we show that manager sentiment is a stronger predictor for volatility than stock return, especially when one accommodates for misspecification in the linear predictive model via a nonparametric datadriven approach. But, predictability is completely absent at extreme ends of the conditional distribution of return, and at the upper end of the same for volatility.