ARDL Bounds Testing With 0 lag?

[AndreaDay]

Hi! I'm doing a cointegration test using ARDL Bounds Testing approach. I'd tried to choose the optimal lag length using the lag criterion in Eviews and it suggested 0 lag.
Does this make sense for an ARDL model?
And when I proceed with the no lagged first difference variables model I found significant for my long-run coefficients using Wald Test. Can this result be sensible?

Many thanks,
Andrea

[ahmed1]

Can someone give Andrea an answer, I have similar challenge with Andrea. What if the optimum lag selection is zero, is it possible to have ARDL(0), that is with lag 0?