Engle Granger

[sweivE]

Can someone please tell me what is the difference between these two procedures:

1. Two-step EG, where I first estimate the static long run relation, get the residuals and then estimate the error correction model.
2. Estimating a system of two equations - one for the long run and the error correction. In the EC, I replace the estimated residual with the actual expression for it (the dependent variable minus the RHS variables)

And one other thing - How can I preform a forecast with the first procedure? My problem is the residual in the out of sample period. I rather not use the VECM methodology.

(I use Eviews 7)