A re-examination of the predictability of stock returns and cash flows via the decomposition of VIX

This paper investigates return and cash flow predictability via the decomposition of VIX. The squared VIX index is decomposed into expected return variations (ERV) and variance risk premium (VRP). Without imposing a strong assumption on the dynamics of the return variations, I examine the predictability via the generalized method of moments (GMM) approach with appropriately chosen instruments. Empirical analysis shows the short-term return predictability of VRP and the short- and long-term cash flow predictability of ERV.