Johanson model of cointegration uses I(1) variables in a vector form which helps to find long run and short run coefficients also the direction of causality as in this model any variable can be endogenous. This manual also includes explanations to unit root test, variance decomposition test and impulse response function. The manual and graphical illustration to this model is provided in link below

Johanson Cointegration & VECM

### Like this:

Like Loading...

*Related*

My model includes 5 variables where the dependent and 2 independent variables are I(1) but the other 2 independent variables are I(0). Can i run a Johansen cointegration test or VECM here. If i can run VEC here then how to choose the number of cointegrating rank. please describe

first of all you cannot use mixed order of integration in VECM, even if you want to use then you have to mention that I(0) variables are exogenous in regression such that they will only appear in short run not in long run. for the cointegrating rank you have to use the rank test and it depends on the theory that if you want to ensure rank = 1 or any other.

Dear Noman Arshed, Thank you for your kind response. I think that the two I(0) independent variables would be exogenous and will appear for in long-run not in short-run and i will use then Impulse response function to show the effect of I(0) variables on the dependent variables.In Eviews, if the 2 I(0) variables are put as exogenous variables , these only appear in the Long-run estimation part not in the short-run estimation part.Would you please like to clarify it.Would you please show the command of VEC with exogenous variables.

Another fact is that, there is no theoretical restriction to make r=1, Can i use the Johansen cointegration test here?

Again , heartiest thank for your generous contribution to the dissemination of knowledge.