0
$\begingroup$

Hey I have to regress abnormal returns for 202 companies over a period of 10 years (between 2004-2013) against their respective PE ratios and a Profitability measure for the same time period. Thus I have to run a pooled regression. I have read online on how to do it...but still am not sure of the process and also did not find any appropriate examples, or one similar to my case. Here my dependent variable are the returns of the 202 stocks over a 10 yr time period. And the two independent/ explanatory variables are their PE ratios, current and next year and profitability measures respectively (both of which are also considered fr the same time period of 10 years). So please could anyone tell me whether it is possible to run this regression in eviews?

I was confused regrading what should i select in the estimation box options. Firstly, what should I consider as my common coefficient/cross section coefficient..in the estimation box options. Also should I go for fixed effects/common/ random effects intercept option?

The equation I constructed to test whether there are any PE, Expected PE or time lag PE effects on Abnormal Returns is as follows:

Abnormal Return (t) = α+β1 (PE t-1)+β2(PE t)+β3 (PE t+1)+ β4(Profitability t)+ε​

Please help me out anyone by telling me how can I run a pooled regression for my dataset and equation..or is it even possible..to take these data from three different sheets and accordingly run it because otherwise I have to take mean returns of all companies and mean PE of all companies and run the regression which would be less satisfactory to my dissertation research. Thanks alot in advance.

Regards, Samiha

$\endgroup$
-1
$\begingroup$

Pooled regression is only possible with panel data. Make sure your dataset is a paneldataset. So you need a timedimension and a crosssectional dimension. If so you can easily select your data and run a Pooled OLS regression. It is the standard regression method for panel data so you don't have to change settings.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.