Page 1 of 1

Dummy variable in cointegration equation

Posted: Sun Nov 27, 2011 12:02 am
by nic
Hi All,

I just wanted to know whether it is possible to use co-integration to examine the long-run relationship between X and Y in Eviews where Y is a dummy variable and X has a unit root. Alternatively, what is the option in Eviews to incorporate the dummy variable Y if I need to examine the long run relationship between X and Z (both X and Z are stationary). How do one interpret such dummy variable coefficient in the cointegrating equation?

Thanking you in advance for your help.

Re: Dummy variable in cointegration equation

Posted: Sun Nov 27, 2011 7:50 am
by startz
Stationary variables, including dummy variables, are never cointegrated.

Re: Dummy variable in cointegration equation

Posted: Mon Nov 28, 2011 9:18 pm
by nic
Stationary variables, including dummy variables, are never cointegrated.
Thank you Startz. You did partially solve my problem. So, you mean to say that it is not possible to examine the effect of say gender (dummy variable) on the long run relationship between two non-stationary variables (assuming that these two non-stationary variables are co-integrated). Put alternatively, what is the best way to analyze the effect using dummy variable (Y) such as gender in a model with non-stationary dependent variable (Z) and non-stationary independent variable (X) . Please recommend any tool in Eviews if it exists.

Thanking you in advance.