Time windowFama-MacBeth Regression (fmb)

[Juana Fusteria]

Hello everyone,

I am trying to run a 2 step fama macbeth regression for financial data, for which this thread (https://www.statalist.org/forums/for...eth-regression) has been very useful. I have 170 time periods for the classic excess returns (LHS variable) and on the RHS I have the classic fama french 3 factors and an added crisis variable for 30 equity indices that I treat as stocks. If I understood the process correctly from the aforementioned post, I would have 170 regressions with 30 observations each. Is that correct?

Additionally, I would like the formation period for the betas to be of 12 time periods, but when I use the classic assreg, fmb command, I do not see how to specify the formation period for the betas. Is there a way to do this?

Thanks and kind regards,

Juana

[William Lisowski]

Perhaps you will find the discussions on Attaullah Shah's web pages at the links below helpful.

https://fintechprofessor.com/stata-p...ions-in-stata/

https://www.statalist.org/forums/for...regression-fmb

I will also add that if you can sort your way through the confusion of the early posts in the following topic, the final few posts, where the original poster reaches clarity on how to accomplish something that seems similar to your objective, may prove helpful.

https://www.statalist.org/forums/for...iant-variable/

Let me add one final piece of advice. Please keep clear in your mind that the Fama-French model is different than Fama-MacBeth regression. FMB is a tool often used in the final stage of FF, but it has many other uses as well, as Professor Shah's web pages discuss.

[Juana Fusteria]

Thanks for your reply. I understand that FMB is different from FF but the fame french 3 factors are usually still used on the fama macbeth regressions as I understand. I had already checked professors Shah's website and I asked the question there also because as far as I could see that issue was not addressed there.

[William Lisowski]

In the second of the links I pasted in post #2 my clipboard apparently contained the wrong liink. The following is the correct link to a very recent post to his blog in case you have not seen it.

https://fintechprofessor.com/2019/06...sreg-in-stata/

[Juana Fusteria]

Thanks. I had actually seen that also, but I appreciate your responses.