Return Predictability: Learning from the Cross-Section

52 Pages Posted: 10 Mar 2015 Last revised: 28 Mar 2017

Date Written: February 2, 2016

Abstract

Long histories of returns are needed but often lacking when estimating the equity premium. This paper studies stock return predictability from the perspective of a Bayesian investor who has access to international data. Learning across countries arises whenever this investor believes that international return processes share a common distribution. The model allows for samples of different lengths and introduces economic constraints on equity premium forecasts. Empirically, estimates are more reliable, an effect that manifests itself both in- and out-of-sample. International predictability also appears much less heterogeneous than previously reported.

Keywords: Stock return predictability, Bayesian panel VAR, Asset allocation

JEL Classification: C22, C23, G11, G12, G15

Suggested Citation

Pénasse, Julien, Return Predictability: Learning from the Cross-Section (February 2, 2016). Available at SSRN: https://ssrn.com/abstract=2575725 or http://dx.doi.org/10.2139/ssrn.2575725

Julien Pénasse (Contact Author)

University of Luxembourg ( email )

4 Rue Albert Borschette
Luxembourg, L-1246
Luxembourg

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Downloads
485
Abstract Views
2,689
rank
73,514
PlumX Metrics