A1217
Title: Firm characteristics and the cross-section of stock returns: A tale of two tails
Authors: Daniele Bianchi - Queen Mary University of London (United Kingdom) [presenting]
Pedro Moravis Venturi - Queen Mary University of London (United Kingdom)
Abstract: The role of firm characteristics is explored to predict the cross-section of stock returns through the lens of a flexible Bayesian variable selection prior to being embedded in an otherwise conventional parametric portfolio choice. The main results show that model uncertainty is pervasive, and there is little evidence in favor of sparse models. Yet, there is a trade-off between sparsity and shrinkage when maximizing the portfolio's expected utility: while a heavy-tailed sparsity-inducing prior reduces uncertainty on which firm characteristics matter, it also produces strikingly less diversified portfolios with more extreme weights. As a result, when transaction costs are factored in, a dense model that allows for selecting many characteristics while shrinking their impact on the optimal portfolio choice is more adequate to capture the out-of-sample variation of stock returns.
