Regression with ARIMA errors

[soudoubaiqi]

Hi folks,

I'm trying to use a non-stationary time series X to explain another non-stationary time series Y
Y=bX+e

where e is a non-stationary process and its first order difference d(e) can be modeled as a MA(1)

Would anyone help me to figue out how to specify the estimation in the Eviews. Thanks!

[startz]

Maybe

Code: Select all

ls d(Y) d(x) ma(1)