Page 1 of 1

Generate long-run forecasts

Posted: Tue Oct 10, 2017 3:51 am
by mchan
Hi,

I have a very basic question. I estimated a DOLS and FMOLS equation and would like to generate a long-run forecast using the coefficients. This is before building an error-correction model.

I proceed by using Proc > Forecast however I noticed that the series generated is very different (much larger) from my manual calculation using the eviews command genr y=c(1)*x1+c(2)*x2+...

Why is this so? What is the Proc > Forecast function calculating then?

Thanks

Re: Generate long-run forecasts

Posted: Tue Oct 10, 2017 6:36 am
by EViews Gareth
Click on View->Representations to see the forecast string