Re: [Gretl-users] test-down option reports different results in adf and coint
Am 27.05.2015 um 13:07 schrieb Koray Simsek:
I noticed a few threads in which the logic of --test-down was discussed
(whether the maximum possible data is used for each lag vs the current
set-up) but could not really find anything with respect to this
particular difference in adf and coint. I am not sure whether this is a
bug or not.
Hi,
I think the bottom line was that it's a matter of taste and convention.
I guess you could argue that the adf and coint commands should share the
same convention, yes. But it's also not so clear why it's a problem that
they don't. If the qualitative results change because of the influence
of a few initial values, you're in trouble anyway...
cheers,
sven