A. Levin's code Vs. OLR,what's the difference?


In Dynare, optimal policy rule under commitment (Ramsey problem) can be solved using OLR. In the forum, I also found A. Levin code to compute Ramsey policy. What’s the difference between these two?

Thank you.


OLR should only be used with linear models and not with linearized models. Otherwise the result is wrong.
If you have a nonlinear model and that you want to get a linear (or second order) approximation of the optimal policy, use A. Levin procedure. I will soon release a way of doing it directly in Dynare version 4.