I was running some likelihood ratio tests (using the current version of
tseries) and found a different value for the log-likelihood from what I
was getting using other software. I've traced the problem to R's GARCH
using the conditional standard deviations instead of the conditional
variances. This amounts to a factor of 2 once logs are taken, and could
easily be a case of bad parentheses.
It doesn't appear to affect the estimated coefficients, however.
I'm not a good fortran programmer, so I'll just toss this up to the
group
for comment/investigation.
Thanks,
Elliot.
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !) To: r-help-request at
stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
