I'm currently trying to finalize the new forecasting capabilities in
gretl for version 1.4.0.
I have recently added out-of-sample forecasting with standard errors
for VAR systems, and also for GARCH models. I'm not quite sure
about the latter, and if anyone who knows more about garch than I do
is willing to take a look at the garch-related functions in
forecast.c in the gretl lib directory, I'd be grateful.
In case anyone is able to test gretl's new forecasting routines
against other programs on Windows, I will make a new Windows build
shortly (i.e. by Monday morning).
--
Allin Cottrell
Department of Economics
Wake Forest University, NC