Page 1 of 1

ECM for FMOLS and DOLS

Posted: Fri Aug 28, 2009 10:58 am
by Nadiarock1
Hi,
Is there an Error Correction Model procedure or an example from a textbook for a Fully Modified OLS and a Dynamic OLS with RATS?

Alose, I have been asked to graph the Impulse Fonction Reactions for the ECM. I don't think I can do it because IFR are associated with VAR methodology. Do I am right?

Thanks a lot,
Nadia

Re: ECM for FMOLS and DOLS

Posted: Fri Aug 28, 2009 11:40 am
by TomDoan
The King, Plosser, Stock, Watson replication file (kpswaer1991) has quite a few examples of the use of the SWDOLS procedure.

FMOLS and DOLS are designed to estimate the cointegrating vector only, not the full VECM. However, given the estimate of the cointegrating vector, you can define a VAR with the ECT similar to the ECT.RPF example. (The JOHMLE procedure used in that is another way to estimate the cointegrating vector).