Page 1 of 1

Feasible Generalized Least Squares

Posted: Wed Oct 19, 2011 5:12 pm
by fnaticsk
Hi, I can't seem to find out how to re-estimate a model using Feasible Generalized Least Squares, could someone please care to explain how it's performed? It would be greatly appreciated. Thanks!

Re: Feasible Generalized Least Squares

Posted: Wed Oct 19, 2011 5:15 pm
by EViews Gareth
You'll have to be more specific - what do you mean by "re-estimate"?

Re: Feasible Generalized Least Squares

Posted: Wed Oct 19, 2011 8:17 pm
by fnaticsk
Ummm.. I'm not sure, but basically I did a Heteroskedasticity BreuschPagan test and got the regression and now I'm supposed to transform the regression using the FGLS to get a better fit/lower standard errors of the coefficients.

Hope that made sense :?

Edit: Oh and I concluded that Heteroskedasticity was present if that helps..

Re: Feasible Generalized Least Squares

Posted: Wed Oct 19, 2011 8:59 pm
by startz
Most likely the assignment is to use the Breusch-Pagan regression to forecast variances and then use the forecast variances as weights.