Title: Term structure of forward moments and equity premium predictability
Authors: Abderrahim Taamouti - Durham University Business School (United Kingdom) [presenting]
Abstract: A recently established aggregation property of the second and third moments of returns is considered to construct forward moments extracted from option prices. We show that according to standard affine no-arbitrage models, the forward moments should exhibit a factor structure, while the equity risk premium should also be an affine function of the same state variables. In light of this, we use dimensionality reduction techniques to extract, from the forward moments, the common factors that maximize the covariance between these factors and the equity premium. We show empirically that a small number of factors can explain the equity premium, both in-sample and out-of-sample, better than most traditional predictors. Moreover, we document that the inclusion of forward skewness into the analysis improves the asset return predictability and thus show that forward moments encapsulate important information about future market returns.