This might be more of a question for R-SIG-Finance and followup should probably be there, but you might get a start with the rugarch package.
Michael On Wed, May 2, 2012 at 4:13 AM, Ivette <iva_mihayl...@mail.ru> wrote: > I have done the usual estimation of GARCH models, applied to my historical > dataset (commodities futures) with a maximum likelihood function and > selected the best model on the basis of information criteria such as Akaike > and Bayes. > > Can somebody explain me please the calibration scheme for a GARCH model? > > I was not able to find a paper, dealing with exactly this algorithm for my > case. I only understood that I have to compare the performance of the best > GARCH model (from the estimation step), fitted to my historical dataset and > a GARCH simulation (let's abbreviate this Squared Error difference to "E2"). > However, it is not clear to me: > - with what parameters' values to start this simulation, > - how many times it is normal to perform it, and > - what to compare via E2 (maximum likelihood values, or parameter values) > - how to construct&assess E2 for the GARCH case. > > Thank you in advance for your suggestions. > > Ivette > > -- > View this message in context: > http://r.789695.n4.nabble.com/calibration-of-Garch-models-to-historical-data-tp4602606.html > Sent from the R help mailing list archive at Nabble.com. > > ______________________________________________ > R-help@r-project.org mailing list > https://stat.ethz.ch/mailman/listinfo/r-help > PLEASE do read the posting guide http://www.R-project.org/posting-guide.html > and provide commented, minimal, self-contained, reproducible code. ______________________________________________ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
