How can I implement the Fama MacBeth Method in Matlab?

Hello All,

I´m currently trying to compute a Fama MacBeth Regression, which is executed in two steps.

1. You calculate time-series regressions for each asset/portfolio to obtain your betas.

2. You run a cross-sectional regression for every period n to determine your risk premium.

Although it doesn`t sound to complicated I can`t advance because one thing is unclear to me: To be more specific here some information to my data: In my dataset I have 25 Portfolios and 4 factors. I have all my time-series betas ready and stored for the cross-sectional regression but now I get to the problem. Do you make n ( for every period) regressions for EACH portfolio and calculate a risk premium separately for each portfolio. In my case that would be 25 times n regressions. Or do you take the mean of all portfolios and thus only compute n regressions ( one for every period).

I am very thankful for any suggestions/answers ! With kind Regards, A. G.

[Nayyab Mir]

did you manage to figure that out? I am also stuck on the same

[Dennis Schock]

same question few weeks later. Did you manage to figure that out? Same problem...