How to create a matrix of coefficients, the alpha, t stat and P-values from an automated series of multiple regressions

[Shivam Patel]

Very similar to Benjamin Arndt's question. Only I'm new to STATA and can't work out which parts of his code are relevant to me and how to create the missing code that I need.


My data has the following columns - Date, Fund1,.....Fundx, Factor1, Factor2, Factor3, Rfr

The observations for the Funds are unbalanced, in that some Fund's have observations for all dates, and others have observations for some dates.


The time-series multiple regression I need to perform is -

(Fund1 - Rfr) = the Y variable

Factor1 Factor2 Factor 3 = the X variable

I need Newey West Standard errors.


After the regression for Fund1, I need to store the factor coefficients, the alpha, t-stats and p-values as a new matrix and for the process to automatically repeat for Fund2 all the way to Fundx.


I emailed tech support and they said I'd need to 'programme' STATA (which I have no clue!), forwarding me 'help foreach' 'help local' 'help postfile - which again confuses me.

[Steve Samuels]

Welcome to Statalist, Shivam! Unfortunately, this Forum is for discussing Mata, which is a specialized matrix language for Stata. Post in the General Forum, where more people will see your question.

[Shivam Patel]

I see! I was wondering why no one was replying to me. Thanks. I've posted on the other forum, a moderator can now delete this if they need.