I have 3 variables. 2 are I(1) and 1 is I(0). I specify 3 ARDL's with first difference of each as dependent variables and conduct the bounds test on each ARDL. I found cointegration in only 1 of the ARDL's. (1) Can I conclude that there is 1 cointegrating equation like in the johansen frame work? (2) Should I then estimate a VECM or a single equation ECM? (3) What should be the next step if from bounds testing, I had found cointegration in 2 of the ARDL's?