[Gretl-users] Re: multi-step-ahead-forecasting
Am 07.01.2021 um 21:50 schrieb Brian Revell:
I have a different but similar problem where 2 series appear to be
recursively linked Yt=f(Xt ...) ; Xt=g(Yt-1....) . How to generate
dynamic roll-forward forecasts from the two separate equations when both
Tt and Xt are of interest.
Need to specify a (bivariate) system.
-s