Calculating the square root of the variance of the residual of a regression

[Benn]

Hi Guys,

I am trying to calculate the square root of the variance of the residual of a regression.

Basically if the regression is the following:

R= a+ b1.MKT+ b2.SMB + B3.HML+ B4.UMD + E

Then I am trying to compute X, where

X = √ var(E).

PS: Also, can the result be stored in a SAS dat. file, because I need to merge it to the original data set later on.

Thanks a lot for the help guys! Appreciate it!

[PGStats · Accepted solution]

Proc reg will give you the root mean square error as part of the fit statistics. You can output them in dataset myStats :

proc reg data=myData plots=none;

     model R = MKT SMB HML UMD;

ods output FitStatistics=myStats;

run;

PG