Option-Implied Quantiles and Market Returns (Extended Abstract)

Fear of disastrous tail events embedded in the short-term option contracts is reflected in the long-term equity risk premiums (ERP). A novel formula is proposed to identify the risk-neutral return quantiles from European option prices in a model-free manner. We use this formula to extract risk-neutral return quantiles of the S&P 500 index from January 1996 to June 2019. In the uni-variate predictive regressions, we find the difference between 5% and 95% risk-neutral quantiles (TD) significantly predicts equity risk premiums at horizons of more than one year, based on the standard error estimates of Hansen and Hodrick (1980) corrected for heteroskedasticity, Hodrick (1992) and Newey-West. We argue that TD captures the aversion to disastrous tail events of the market participants and, consistent with this, we find that TD is highly persistent. In the bi-variate predictive regressions that control for the well-known market predictors, TD is complementary to the variance risk premium of Bollerslev, Tauchen and Zhou (2009), which is a significant predictor of the ERP at horizons of less than one year. The correlation of TD with the dividend price ratio is 35% and highly significant, which is consistent with the finding of Campbell and Shiller (1989) that the variation of dividend price ratio is driven mainly by the variation of future discount rates.