Replicating Anomalies

In their paper, “Replicating Anomalies,” researchers Kewei Hou, Chen Xue, and Lu Zhang (“HXZ”) looked at 447 factors from quantitative literature to reproduce the outcomes. They found that 286 factors (64%) had no statistical significance. For the 161 factors that appear robust, their magnitudes are less than reported. Specifically, 115/161 don’t have noteworthy “alphas.” Long story short, there doesn’t seem to be many anomalies that can be explained by a factor. However, there are two exceptions: momentum and value.