Obtain GARCH series from an e-GARCH model
Posted: Wed Jan 28, 2015 6:02 am
Hi, I'm using Eviews 8 and obtained a GARCH variance series using the following commands:
equation onraf.arch(1,1,asy=1,egarch,tdist) d(sonia) d(sonia(-1)) d(sonia(-2)) c dbru dbre dbru(-1) dbre(-1) dbru(-2) dbre(-2) (sonia(-1)-br(-1)) me qe ye pre_bh bh endmp @ me qe ye pre_bh bh endmp
onraf.makegarch onrafgarch
The problem is I am unable to replicate the estimated GARCH series in Excel. I tried with a simpler GARCH structure:
equation onraf.arch(1,1) d(sonia) d(sonia(-1)) d(sonia(-2)) c dbru dbre dbru(-1) dbre(-1) dbru(-2) dbre(-2) (sonia(-1)-br(-1)) me qe ye pre_bh bh endmp @ me qe ye pre_bh bh endmp
and got the exact same GARCH series in Excel. So it seems to be a problem with the e-garch structure.
What could be going wrong? In Excel, I obtain the log(garch) using the same equation that appears in the results window (+ regressors and residuals imported from Eviews), and then take the exponential to obtain the GARCH term. The results are v different though seem to be highly positively correlated with the Eviews GARCH series.
Thanks
equation onraf.arch(1,1,asy=1,egarch,tdist) d(sonia) d(sonia(-1)) d(sonia(-2)) c dbru dbre dbru(-1) dbre(-1) dbru(-2) dbre(-2) (sonia(-1)-br(-1)) me qe ye pre_bh bh endmp @ me qe ye pre_bh bh endmp
onraf.makegarch onrafgarch
The problem is I am unable to replicate the estimated GARCH series in Excel. I tried with a simpler GARCH structure:
equation onraf.arch(1,1) d(sonia) d(sonia(-1)) d(sonia(-2)) c dbru dbre dbru(-1) dbre(-1) dbru(-2) dbre(-2) (sonia(-1)-br(-1)) me qe ye pre_bh bh endmp @ me qe ye pre_bh bh endmp
and got the exact same GARCH series in Excel. So it seems to be a problem with the e-garch structure.
What could be going wrong? In Excel, I obtain the log(garch) using the same equation that appears in the results window (+ regressors and residuals imported from Eviews), and then take the exponential to obtain the GARCH term. The results are v different though seem to be highly positively correlated with the Eviews GARCH series.
Thanks