Hi
I am researching CAPM, Fama-French 3F and Fama-French 5F.
I used the data from their website on the factors (5F 2x3) and the portfolio's formed on Size-BM, Size-INV and Size-OP sorted 2x3 from 07/1963 to 12/2020.
I did a GRS-test and all models failed, then I did individual regressions for returns of each portfolio against each factor model. I am receiving R2 of 95% and up (even for CAPM) and very low alpha values that are considerably lower than what Fama and French found in 2015 (A 5F Asset Pricing Model).
What could be the reason for these results that are better than their?
The best thing to do, in my opinion, is to search for studies which got results that are similar to yours, from which you will get the respective justification(s) that you can compare to yours.
Generally, individual test results are different from portfolios' results due to a number of reasons. I did such tests on a few papers of mine; you may want to check them out.
https://www.investopedia.com/terms/f/famaandfrenchthreefactormodel.asp
- Badges
- Science topic