[Gretl-users] Re: About the PMG package
On Wed, 24 Mar 2021, Sven Schreiber wrote:
Again, it's hard to discuss this without having seen the
screenshot. But
if we're just talking about the plain direct Eviews output, then these
are the MG estimates. The SEs are the so-called non-parametric estimates
from the cross-sectional dispersion of the heterogeneous coefficients.
(Applying the plain Eviews output to an individual unit would be wrong,
given the short-run heterogeneity. Eviews also provides the
heterogeneous coefficients in some post-estimation view.)
OK, mystery solved, thanks to this video:
https://youtu.be/JDVr5FIL78Q
The guy says, at 2:50, that for convenience they are diplaying the average
of the individual coefficients. In practice, this (apart from the
different ordering):
<hansl>
include PMG.gfn
set verbose off
open oecddata.gdt --frompkg=PMG
list X = DP LNDI
bundle b = PMG(LPC, X, 1)
bundle myopt = defbundle("estshortrun", 1)
b = PMG(LPC, X, commonlag, , myopt)
printf "\naverage coefficients:\n%12.5f\n", meanr(b.fullcoeffs)
</hansl>
This is close enough to the output you get from Eviews in the video. It
remains too be seen how the standard errors are computed.
-------------------------------------------------------
Riccardo (Jack) Lucchetti
Dipartimento di Scienze Economiche e Sociali (DiSES)
Università Politecnica delle Marche
(formerly known as Università di Ancona)
r.lucchetti(a)univpm.it
http://www2.econ.univpm.it/servizi/hpp/lucchetti
-------------------------------------------------------