[R] A mistake in garchFit()? {fGarch}

[Yohan Chalabi]

>>>> "TY" == Ted Young <unixunix99 at gmail.com>
>>>> on Tue, 30 Dec 2008 11:53:23 -0800 (PST)

   TY>
   TY> Hello,
   TY>
   TY> I was using garchFit {fGarch} to fit some GARCH processes.
   TY> I noticed that the result contains Log Likelihood value
   TY> (right above
   TY> Description), but when I use .. at fit to retrieve Log Likelihood
   TY> value,
   TY> the sign switched.
   TY> I am confused about which value I should choose to report...
   TY> Any help here?
   TY>
   TY> Thanks a lot!
   TY>
   TY> Ted

the result contains the maximized Log likelihood. 

But internally the negative Log likelihood is minimized ... 

hope this helps,
Yohan

-- 
PhD student
Swiss Federal Institute of Technology
Zurich

www.ethz.ch