Announcement

Collapse
No announcement yet.
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • Preformation betas with rolling regressions

    Dear Statalisters,

    I am working on my masterthesis Finance at the moment and I try to calculate the factor sensitivity of companies to subsequently create new stock portfolios. I tried to calculate the factor sensitivity using the rolling regression command in Stata. The code looks like this:

    rolling _b, window(60) saving(betas, replace) reject(e(permno) < 24): reg ret factor

    permno is the company_ID.

    Unfortuantely, the code is working very slow and takes to long on large datasets.

    I have searched on the internet and I found the asreg command which does it extremely fast:

    bys permno: asreg ret factor, wind(60) min(24)

    However, when I am testing the results of both commands on small datasets, I won't get the same results.

    Therefore my question, could someone help me with a Stata code which would give me faster results? and where could be the difference between the two codes?


    Thank you very much,

    Karina

  • #2
    I can't believe that e(permno) is a returned result. e(N) seems more likely.

    Tell us more about the results of

    Code:
    tsset

    Comment

    Working...
    X