Programming the statistical procedures from SAS

many mistakes in my F-test part. help me please. thank you very much!

Reply
Occasional Contributor
Posts: 15

many mistakes in my F-test part. help me please. thank you very much!

here is some explain about my question. I need some help. I am kind of following Fama French (1993). I have 50 portfolios. I have to run a time series regression of return of each portfolio on Rm-Rf, SMB, HML. To test 3 factor model of FF, I have to test whether the intercepts in each regression above are jointly significantly different from 0. To do that, the best way is to use Gibbons, Ross, Shanken (1989) test. First I calculate the covariance matrix of the 4 risk factors using proc corr and output the matrix using ods function. Then I calculate the covariance matrix of error terms using proc syslin - SUR. (I am not sure how big a system can proc syslin support) Finally, I use proc iml to calculate the GRS F-Stat. Thank you very much. Regards, Duong *

Ask a Question
Discussion stats
  • 0 replies
  • 211 views
  • 0 likes
  • 1 in conversation