[Gretl-users] Re: How to apply the HAC option in Gretl?
On Thu, 10 Nov 2022, Cottrell, Allin wrote:
* With missing values, the HAC results differ noticeably, and it
turns
out that R is ignoring the missing values altogether. I can replicate
the R results in gretl if after sprinkling the NAs I do "smpl
--no-missing" then reestablish the time-series property with "setobs 1
1 --special-time-series".
IMHO from a statistical point of view this makes no sense whatsoever.
I wonder if that could be the case with Eviews too.
Probably not. If that was the case, modifying my little example script as
in
<hansl>
nulldata 64
set seed 20221109
x = uniform() < 0.9 ? normal() : NA
y = uniform() < 0.9 ? normal() : NA
smpl --no-missing
setobs 1 1 --special-time-series
ols y const x --robust
</hansl>
would produce the same standard errors that Sven reported earlier today
(and doesn't).
-------------------------------------------------------
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
-------------------------------------------------------