Page 1 of 1

How to estimate a structural break when using FMOLS?

Posted: Wed Jul 10, 2013 6:29 pm
by boksua
Hello, I am working on estimating wealth effect(or consumption function?) with Eviews8.

My question is, when a structural break is likely to be around the late 1999 in regime (I mean, a shift in regime - according to Gregory-Hansen), is it OK to just enter elements such attached?

(consum= consumption
income=income
wealth=wealth
dum=0 before 1999q2 and 1 since 1999q2)

and when the estimation out says dummy variables are all significantly different from 0, can we suggest there may have been a structural break?
(I know that Gregory-Hansen test assumes there is only one structural break in a sample period)

Thanks!