The past few weeks we’ve highlighted a set of research papers that go back and forth on the validity of the Fama and French “5-factor model.”
A sampling of the research:
- The Fama French 5-Factor Paper
- The Kewei, Xue, and Zhang (KXZ) 4-Factor Paper (critically assesses the FF 5-factor model)
- Cakici investigates the FF 5-Factor International Evidence
- FF investigate the international evidence for their 5-Factor model
We summarize the debate/controversy in a couple of recent posts:
Now we have a new paper from Back, Kapadia, and Ostdiek that investigates the merits of both the FF 5-Factor and the KXZ 4-Factor.
We test the recent Fama-French five-factor model and Hou-Xue-Zhang four-factor model using test assets from Fama-MacBeth regressions, which are bets on particular characteristics or covariances that are neutral with respect to others. Monte Carlo evidence shows that the tests are unbiased, despite errors in variables. Test assets that are bets on characteristics (covariances) have positive (negative) alphas. In particular, neither model can explain returns related to investment rates, the Fama-French model cannot explain momentum, and the Hou-Xue-Zhang model cannot explain value. The rejections are economically significant.