AGARCH + rugarch

Just trying to replicate a bit of research which uses it. 

As to the value of it compared with GJR -- I'll defer to your expertise. 

Cheers,
MW

On Nov 19, 2013, at 19:57, alexios ghalanos <alexios at 4dscape.com> wrote:

Sentana (1995) actually worked out more thoroughly the details of this
model which he called quadratic ARCH (QGARCH).

This is not implemented in rugarch. For asymmetric response to shocks
try the GJR (gjrGARCH), TGARCH (model="fGARCH", submodel="TGARCH")
or NAGARCH (model="fGARCH", submodel="NAGARCH")...among others.
If you provide compelling evidence why this model merits inclusion I
will consider it.

Regards,

Alexios


On 20/11/2013 00:34, R. Michael Weylandt <michael.weylandt at gmail.com> wrote:

Is it possible to implement the AGARCH [1] of Engle 90 using rugarch? The aparch models seem close, but don't allow for different powers of epsilon as far as I can see. 

Thanks, 
Michael

[1] I know GARCH terminology is a mess, so I'm using that given at public.econ.duke.edu/~boller/Papers/glossary_arch.pdf (P.2 in my case)