Hello,
This morning, I found a strange behaviour in the garchFit method. Fitting
with trace=T doesn't give the same fitted value for h.t (or sigma.t) :
> X = garchSim(n=1000)
> head(garchFit(~garch(1,1), data=X, trace=T)@h.t)
....
--- END OF TRACE ---
Time to Estimate Parameters:
Time difference of 1.093 secs
[1] 1.152227e-05 1.167620e-05 1.094984e-05 9.857859e-06 9.280726e-06
8.473238e-06
> head(garchFit(~garch(1,1), data=X, trace=F)@h.t)
[1] 1.152036e-05 1.167270e-05 1.094500e-05 9.852020e-06 9.274241e-06
8.466313e-06
The difference is slight, but I doesn't see why it exists. All other
parameters I checked are equals. I will investigate more deeply in the
code later this afternoon.
Sylvain.
*************************************************************************
This message and any attachments (the "message") are con...{{dropped:13}}
_______________________________________________
R-SIG-Finance@... mailing list
https://stat.ethz.ch/mailman/listinfo/r-sig-finance-- Subscriber-posting only.
-- If you want to post, subscribe first.
