IRF in a two-steps procedure

[unknown]

Hi,

I am estimating an error-correction model in a multivariate framework. Since the standard procedure (Johansen ML) leads to not satisfaying results, I use the two-steps Engle-Granger approach. However, my interest lies in analysing the impulse responses functions in the VECM. So I would like to know if it is possible to compute these function from the two-steps method in Eviews and if so, how to do this?

Thanks for your help.

[unknown]

I updated this old post with the hope that soemone will bring an idea.
My approach consists in estimating the long run equilibrium with the DOLS and FOLMS procedures. Then I use the residuals of the first step to estimate the short run equilibrium.
Based on that second step, I would like to plot the impulse response functions. So my question is is it possible to do that using EViews or does the only way to get the innovations is to use the standard VAR/VECM procedure?

Many thanks in adavance for your help.