Dear Stata users,

after checking the stationarity of my panel data regression model: Y = f (X1, X2, X3, X4), it comes out that just one independent variable is non-stationary (precisely, I(1)), whereas all the other variables (including the dependent variable) are stationary (I(0)).

Due to this, I was wondering whether conducting a cointegration test does even make sense, since the latter implies that all the variables shall be I(1) ?

Besides, to my understanding, the standard cointegration tests (such as the Johansen test, for example) rely on this condition as well..

Thank you.

Jack