News and Asset Pricing: A High-Frequency Anatomy of the SDF

Abstract

We rely on a unique set of high-frequency factors to robustly estimate an intraday Stochastic Discount Factor (SDF). Exploiting the precisely timed jumps in the estimated SDF together with real-time newswire data, we identify and precise the news that is priced. We find that news related to monetary policy and finance on average account for the largest portion of the variation in the SDF and the tangency portfolio risk premium, followed by news about international affairs and macroeconomic data. Reflecting investors changing economic concerns, we also uncover significant temporal variation in the relative importance of the news that matters. Relying on a standard mimicking portfolio approach, we further document marked differences in the way in which the news, and the compensation therefor, manifest in the “factor zoo.”

Keywords: SDF, high-frequency factors, jumps, news, risk premiums

JEL Classification: C58, G12, G14