Dear Ox-users
I'm tryng to model some stock indices with the g@rch package 2.3 (console
vesion) and Ox 3.1. There is no convergence for EGARCH and FIEGARCH for all
series and for all distributions for the innovations. However, using EVIEWS
and the BHHH for EGARCH(1,1)~N, there is convergence with few iterations.
Could someone explain this? Any help will be welcome.
One of the series, a brazilian stock indice, is attached.
Regards
Washington Santos da Silva
Federal University of Lavras - Brazil
|