[R-SIG-Finance] garch model estimation

[Yohan Chalabi]

>>>> "ST" == ShyhWeir Tzang <swtzang at gmail.com>
>>>> on Thu, 22 Oct 2009 20:16:48 +0800

   ST> I used the data from the package fPortfolio like the following.
   ST> But the results seem not right as the estimated parameters
   ST> such as alpha and
   ST> beta didn't change at all. Can someone or the authors help me
   ST> out? Another
   ST> question is how to compute the standard errors and t values
   ST> numerically by
   ST> the R code? Any hint or help is highly appreciated.


Hi ShyhWeir,

Your data set is badly scaled for the optimization and you should first
check if there is any GARCH process in it.

try out with the data from package fGarch :

# after your code
library(fGarch)
data(dem2gbp)
x <- dem2gbp[,1]
init<-garchInit(x)
garchllFit(x, init[,1], init[,2],init[,3])

This gives me the same fitted parameters as on the presentation.

You might be interested by the paper of Zivot on "Practical Issues in
the Analysis of GARCH Models" which is available on his website.

HTH,
Yohan

-- 
PhD candidate
Swiss Federal Institute of Technology
Zurich

www.ethz.ch