I(2) JJ Cointegration Test

[chris651]

Hi @ll,

I am working with a 5 variable time series dataset for spain and it seems that all but one variables are I(2).

My questions:

(i) Can I use the implemented JJ cointegration test for I(2) variables as well, or do i have to make some adjustments before i use it because they are not I(1)? (In the case that I only include the four I(2) variables)

(ii) How could I proceed when I include the I(1) variable too? (4xI(2) and 1xI(1))

Thank you very much in advance

Chris