1
$\begingroup$

I just started using the "fGarch" package in R and until now everything has been going fine.

I have estimated two successful models from two different datasets that have been estimated correctly using a normal distribution.

I am now trying to estimate a GARCH model with a fitted $t$-distriubtion but cannot get it to work.

One suggested model is:

      myGarch=garchFit(formula=~arma(0,1)+garch(1,1),include.shape=TRUE,cond.dist="std", data = log_returns)

Then I use this to look at the shape paramter:

garchModel@fit$params$shape

This returns 4, which is the standard degrees of freedom for the $t$-distribution indicating that they have not been estimated? This holds true even if I change the other model parameters or try with a different dataset. I'm sure I'm missing something obvious.

$\endgroup$
1
$\begingroup$

Turns out garchModel@fit$params$shape will only report the standard shape for some reason and the coef function is what one would want to use to find the shape parameter. I didn't realize it would be reported there. If I'm wrong in this, do tell; but I think that's the solution.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.