Kalman filter with GARCH errors
Posted: Sat Sep 15, 2018 5:46 am
Dear all,
I would like to ask about your opinion regarding the estimation of TVP model in Eview when the conditional variance of error term is time-varying. One way to go would be to use a Kalman filter with GARCH errors as in:
https://www.jstor.org/stable/1392543?se ... b_contents
However this procedure is not avaliable in Eviews (at least not without coding it yourself). Would you be willing to rely only Huber-White correction in this case? Or you think that a better approach would be to use an iterative procedure specified in this post:
viewtopic.php?t=7110
Looking forward to your comments.
I would like to ask about your opinion regarding the estimation of TVP model in Eview when the conditional variance of error term is time-varying. One way to go would be to use a Kalman filter with GARCH errors as in:
https://www.jstor.org/stable/1392543?se ... b_contents
However this procedure is not avaliable in Eviews (at least not without coding it yourself). Would you be willing to rely only Huber-White correction in this case? Or you think that a better approach would be to use an iterative procedure specified in this post:
viewtopic.php?t=7110
Looking forward to your comments.