Hi,
I stumbled over the situation of ARCH diagnostic testing in VARs and
found the old thread from 2017 "multivariate ARCH test for VARs". (See
below for a snippet.) It talks about a "univariate" option which isn't
documented in current gretl 2020a. Is this a secret or has it never been
implemented?
thanks
sven
-------
From 2017:
Allin:
Here's one more thought. If the multivariate autocorrelation and ARCH
tests are fully validated, they should probably become the default for
VARs. But as Lutkepohl points out, they both chew up a lot of degrees of
freedom, and he talks about use of univariate tests as a fallback. So we
might go the other way from the current private option and add a
--univariate option to get the univariate versions even when testing a VAR.
Jack:
Excellent solution IMO.