The Cross-Section of Stock Returns before CRSP
94 Pages Posted: 24 Nov 2021 Last revised: 8 Feb 2023
Date Written: February 7, 2023
Abstract
We comprehensively study the cross-section of stock returns out-of-sample using a novel constructed database of U.S. stocks between 1866-1926. Results over this ‘pre-CRSP’ era generally confirm post-1926 CRSP results. The relationship between market beta and returns is flat, the size premium is weak, and momentum, value, and low-risk premia are sizable positive and significant. Importantly, these factor premia do not decay significantly out-of-sample. In addition, we explore recent machine learning models and explanations of factor premia. Overall, we show that findings in the existing literature on traditional equity factors are robust.
Keywords: empirical asset pricing, return anomalies, machine learning, factor premiums, p-hacking, momentum, value, beta, low volatility, size
JEL Classification: G10, G11, N21, N22
Suggested Citation: Suggested Citation