Steady state from initval and endval


I have a question about using the perfect foresight solver with initval and endval.

I have a medium-sized model. I have two steady states and want to study the deterministic path between both using the perfect foresight solver after a permanent shock. For this, I have solved analytically and numerically for both steady states beforehand.

After I give Dynare my own numerical steady state values in the initval and endval blocks, it seems that Dynare calculates the steady state again but using its numerical solver. Thus, the values in oo_.steady_state are different from what I introduced in initval and endval, which results in the residuals of the model being too large.

Is there a way of telling Dynare to always use my own given values for initval and endval without going through its own numerical solver?

Thank you in advance.

Yes, there is. But the problem you describe indicates that the steady state you computed does not match the equations you entered. Hence, overriding Dynare’s steady state finding would not solve the issue. Rather, you would force wrong simulations.