Page 1 of 1

Forecasting with GMM

Posted: Thu Feb 05, 2015 12:08 pm
by henriqueld
Good afternoon!

I have used GMM to estimate a Phillips Curve in Eviews 8. When I try to forecast though, I get only one observation (the last one from my sample)! I have data up until 2014Q1, so that´s up until when I get the "forecasts". What can I do to get real forecasts beyond 2014Q1??

Thanks!

Re: Forecasting with GMM

Posted: Thu Feb 05, 2015 12:33 pm
by EViews Gareth
Post your workfile.

Re: Forecasting with GMM

Posted: Thu Feb 05, 2015 12:41 pm
by henriqueld
I can't. The file is too heavy. What I'm estimating through a new system is:

dlog(ipca_livres) = cp(10)*log(1+expected_inflation) + cp(21)*dlog(ipca(-1)) + cp(30)*d(lcrb) + (1-cp(10)-cp(21)-cp(30))*d(lcrb(-1)) + cp(40)*lhiato + cp(50)*d(loil_oil_hp) + cp(61)*(lipa_ipca(-1)-lipa_ipca_hp(-1)) + cp(71)*q1 + cp(72)*q2 + cp(73)*q3 + cp(74)*q4 + cp(80)*crise

@inst lhiato(-4 to -8) lcrb(-1 to -4) loil_oil_hp(0 to -4) ipca(-1 to -4) expected_inflation(-1 to -4) ltbill(0 to -8) lvix(0 to -8)

Is it possible to get forecasts beyond the range of the data?

Re: Forecasting with GMM

Posted: Thu Feb 05, 2015 1:10 pm
by EViews Gareth
Impossible to say without seeing the workfile.