I am trying to replicate the Fama-French three-factor model. I am having issues in the dataset.
1) Using the CRSP/COMPUSTAT merged dataset for firm fundamentals from WRDS, and then dropping data on financial and non-US firms, I have much fewer firms than what FF have in the dataset in French's website. Any suggestions on why this is happening?
2) I am unable to obtain data as far back as 1926 from WRDS.
Any suggestions would be very helpful.