Re: [Gretl-users] Dummy variable in cointegration analysis....
On Tue, 4 Oct 2011, mariusz doszyń wrote:
I'd like to include a dummy variable to simple system of
equations (with
cointegration):1) delta_y[t]=a0+a1*y[t-1]+a2*x[t-1]+u[t], 2)
delta_x[t]=b0+b1*y[t-1]+b2*x[t-1]+e[t]. The dummy variable
d[t]=[1,0,1,1,1,0,0,...1,1,0,1,0,1], so it somethimes equals 1, sometimes 0,
and generally it increases the level of y[t] variable in some periods (when
d[t]=1). While using vecm command should I add d[t] or rather delta_d[t]
(which is differenced d[t])? My intuition is that delta_d[t] should be added
to avoid some kind of trend.
What you need to know about gretl's "vecm" command is that it will
not automatically difference the dummy for you. Beyond that the
model specification is up to you.
Dummy variable should be added as a xlist or rxlist (where dummy
is in a cointegration space)?
If you want to confine the dummy to the cointegration space (in
which case it will not generate a trend, even if it's in levels
form), then use rxlist, otherwise use xlist. From "help vecm":
"The optional lists xlist and rxlist allow you to specify sets of
exogenous variables which enter the model either unrestrictedly
(xlist) or restricted to the cointegration space (rxlist). These
lists are separated from ylist and from each other by semicolons."
Allin Cottrell