News‎ > ‎

New Working Paper online: Predictability and the Cross-Section of Expected Returns in Models with Long-Run Risks

posted Jun 25, 2016, 10:23 AM by Julian Thimme
My new paper, coauthored by Christian Schlag and Michael Semenischev, identifies a pattern in the cross-section of expected stock returns that is at odds with the majority of macro asset pricing models. In line with model implications, returns on individual stocks in the cross-section are predictable by the price dividend ratio and the variance risk premium of the market portfolio. However, we show that stocks with high betas from predictive regressions have high returns on average. This constitutes a stylized fact that most asset pricing models (e.g. "the" long run risk model of Bansal and Yaron (2004)) cannot explain. We discuss several channels by which a model could explain the discovered pattern and show that the "good and bad uncertainty"-model of Segal et al. (2015) does the job.