Running several dependent against one equation

[Zacross]

Hi everyone,

I am rather familiar with the basic functions of Eviews but really bad with programming in general, and I believe what I want to do requires programming! I hope you can help me!

So that you guys understand in which situation I'm in, here is the data I have at hand: monthly return from 1991 to 1997 of 5 passive management funds and of 658 active management funds.
My task is to check which strategy (passive vs active management) leads to the highest excess return over the sample period (according to CAPM and multivariate models such as Fama-French and others).
For my presentation, I would like to know whether it is possible to run the respective 663 regressions against the same model (again, CAPM or multivariate) and THEN plot the market coefficient on a graph to assess the consistency of the sensitivity to market fluctuation amongs funds.
Actually, I could probably do without such a thing, but I believe it would be a nice plus to present such scatter plot! Besides, it would help me to understand Eviews better :)

Thanks in advance for your help!

Cheers :)

[EViews Gareth]

http://forums.eviews.com/viewtopic.php?f=5&t=1638

[Zacross]

I had already seen that post before :)

I really don't want to sound lazy or anything, but that post is chinese to me... I can somehow figure out that there is a program that is similar to what I would like to do, but I would be completely unable to modify it...

I saw that there is a program made to regress one dependent Y against, for instance, 15 different X. What I want, however, is somehow the opposite! regressing 600+ different Y against the same equation!

Something like the following

Y1 c X1 X2 X3
Y2 c X1 X2 X3
Y3 c X1 X2 X3
.
.
.

And finally store the coefficients of X1 (optionally of X2 and X3)

Problem also is that the post creates its own data within the code whereas mine already exist :(

[EViews Gareth]

Remove the part of the code that generates data.

Inside the loop that does the estimations, rather than changing X, change Y instead.

You're unlikely to find someone willing to just write the program for you. If make an attempt and post what you have, people are more willing to help and point out where you might have made mistakes.