________________________________ Da: alexios <[email protected]> Cc: "[email protected]" <[email protected]> Inviato: Venerdì 27 Gennaio 2012 17:23 Oggetto: Re: [R-SIG-Finance] Negative Estimate of EGARCH model Dear Alexios, Please, thanks for the quick reply however, quoting your statement, "Correcting" the possibility of negative parameter estimates implies an error in the model." implies that it is normal to get negative estimates in the model. "Correcting" the possibility of negative parameter estimates implies an error in the model. This is not so, and I suggest you try to read some of the references cited in the package about how GARCH models work. The point of eGARCH is that we work with the log of the variance so we do not have to worry about positivity. The parameters therefore, for the intercept (\omega), shock (\alpha) and asymmetry (\gamma) are 'unrestricted' while the parameter on the lagged variance (\beta) is constrained (for stationarity) to be less than 1 and positive. Thus, it is NOT a correction you are seeking but a customization to suit your own beliefs on the model. You are free to go into the code and change these restrictions to suit your purpose. The restrictions may be found in the "rugarch-startpars.R" file in the source installation and the function is ".egarchstart". Regards, Alexios On 27/01/12 16:05, Papa Senyo wrote: > Dear All, > Using the rugarch package, how can you correct the possibility of negative > estimates for instance in EGARCH parameters being negative? > Is it possible to put a restriction on it in order to get positive estimates > especially for the ARCH AND GARCH PARAMETER > > Kind regards, > Papa > > [[alternative HTML version deleted]] > > _______________________________________________ > [email protected] mailing list > https://stat.ethz.ch/mailman/listinfo/r-sig-finance > -- Subscriber-posting only. If you want to post, subscribe first. > -- Also note that this is not the r-help list where general R questions > should go. [[alternative HTML version deleted]]
_______________________________________________ [email protected] mailing list https://stat.ethz.ch/mailman/listinfo/r-sig-finance -- Subscriber-posting only. If you want to post, subscribe first. -- Also note that this is not the r-help list where general R questions should go.
