Saving period effects from pool regression

For questions regarding programming in the EViews programming language.

Moderators: EViews Gareth, EViews Moderator, EViews Jason, EViews Matt

trent_saunders
Posts: 2
Joined: Sat Jun 19, 2021 4:49 pm

Saving period effects from pool regression

Hi,

Is there a simple way to save a series of the period fixed effects estimated in a pooled regression?

I'm currently looping over the period fixed effects coefficients and saving them as separate series. However, this gets a bit messy when estimating multiple regressions, each with a different number of explanatory variables (E.g. depending on the number of explanatory variables in the regression, the first time period dummy might be c(9), c(12), etc).

I also tried using the @effects command, though this seems to save the cross-section and period fixed effects in the same vector.

Cheers,
Trent

EViews Gareth
Fe ddaethom, fe welon, fe amcangyfrifon
Posts: 13364
Joined: Tue Sep 16, 2008 5:38 pm

Re: Saving period effects from pool regression

Unfortunately @effects is the only way to do it. If you have both cross and period effects, you'll have to lop off the first X elements of the vector, where X is the number of cross-sections you have.

trent_saunders
Posts: 2
Joined: Sat Jun 19, 2021 4:49 pm

Re: Saving period effects from pool regression

Thanks. I found an ad hoc solution that works for what I am doing. If I estimate the cross-section fixed effects as cross-section specific intercepts (rather than using cx=f as an estimation option), then @effects only retrieves the period fixed effects. E.g.

pool pool1 nsw vic qld wa sa
pool1.ls(per=f) y_? x1_? x2_? @cxreg c
vector v1 = pool1.@effects