# Can I use a VAR in first differences despite having co-integrated data?

I have two variables. Both are I(1), so non-stationary in levels but stationary in first differences.

However, having run some tests, I find that both are co-integrated. Based on my statistics classes, I'm led to believe I should then use an ECM. However, I want to instead use first differences and estimate a VAR so that I can estimate IRFs.

Is this okay? Or should I not do this given the co-integrated nature of the variables? Or is it possible to somehow run an IRF from the ECM?

Hi: ( Note that below considers the bivariate case but same type of discussion holds in general for $n > 2$ ).