Abstract

Identifying the factors that drive the cross-section of expected returns is challenging for at least three reasons. First, the choice of testing approach (time series versus cross-sectional) will deliver different sets of factors. Second, varying test portfolio sorts changes the importance of candidate factors. Finally, given the hundreds of factors that have been proposed, test multiplicity must be dealt with. We propose a new method that makes measured progress in addressing these key challenges. We apply our method in a panel regression setting and shed some light on the puzzling empirical result that the market factor drives the bulk of the variance of stock returns but is often knocked out in cross-sectional tests. In our setup, the market factor is not eliminated. Further, we bypass arbitrary portfolio sorts and instead execute our tests on individual stocks with no loss in power. Finally, our bootstrap implementation, which allows us to impose the null hypothesis of no cross-sectional explanatory power, naturally controls for the multiple testing problem.

Scientific Portfolio AI- Generated Summary

This paper explores the challenges of identifying the factors that drive the cross-section of expected returns. The authors argue that many studies in finance suffer from data mining and overfitting, leading to the identification of “lucky” factors that do not hold up to out-of-sample testing. They propose a framework for testing candidate factors that accounts for the multiple testing problem and adjusts for estimation uncertainty.

The authors begin by discussing the difference between time series and cross-sectional testing approaches for identifying factors that drive expected returns. They argue that cross-sectional tests are more appropriate for identifying factors that are robust across time and that capture differences in expected returns across assets. However, cross-sectional tests are subject to data mining and overfitting, which can lead to the identification of spurious factors that do not hold up to out-of-sample testing.

To address this problem, the authors propose a framework for testing candidate factors that accounts for the multiple testing problem and adjusts for estimation uncertainty. They argue that this framework can help to identify factors that are robust to data mining and overfitting and that hold up to out-of-sample testing.

The authors then apply their framework to a set of candidate factors and find that many of the factors that have been identified in the literature are not robust to their testing approach. They argue that this highlights the importance of testing candidate factors using a framework that accounts for the multiple testing problem and adjusts for estimation uncertainty.

Finally, the authors discuss the implications of their findings for asset pricing research. They argue that researchers should be cautious when interpreting the results of cross-sectional tests and should use a framework that accounts for the multiple testing problem and adjusts for estimation uncertainty. They also argue that researchers should focus on identifying factors that are robust to their testing approach and that hold up to out-of-sample testing.

Overall, the paper provides a valuable contribution to the literature on asset pricing and highlights the importance of testing candidate factors using a framework that accounts for the multiple testing problem and adjusts for estimation uncertainty.