Re: [Gretl-devel] coint2 with exogenous variables
On Sat, 5 Nov 2011, Andreas Noack Jensen wrote:
Den 04/11/2011 kl. 18.16 skrev Allin Cottrell:
> So I'll try to track down
> the source of the difference with Tr76 added.
I have looked at the problem again and found the error. It is an
easy fix, I guess. The restricted variables are lagged one period.
For that reason the sample is also changing if the restricted
variable is excluded. The attached modified script gives exactly
the same results in Gretl and Ox.
Yes, I came to the same conclusion, and I've just committed a change
to gretl CVS that renders the results the same (from gretl and Ox)
for your original script.
As a side note I have a small suggestion. It is unfortunate to
test cointegration rank in the models with unrestricted
deterministic terms so maybe they should not be there. If you
really want to work with a model with unrestricted terms after a
rank test procedure you can just make the exclusion restriction on
he cointegration vectors and I think that is the "right" way to
end up in that model. At least I shall suggest that the default is
one the models with restricted terms.
You're suggesting that we shouldn't allow including unrestricted
deterministic terms in cointegration testing (though we should allow
them in VECM specifications, having determined cointegrating rank)?
Allin Cottrell