Hi, I am new to simulation and got this question about the post estimations. After feeding the model with data and running the simulation, you get the posteriors of the parameters. Would it be fine to simply recalibrate the theoretical model with the new estimated parameters and turn off the estimation for future uses? future use include Welfare analysis (order 2) and uncertainty (order 3), policy scenarios . I am saying this as keeping the simulation on will be quite time consuming.
You are right that once you have estimated the parameters, you don’t need to estimate the model again. The only issue with what you describe is the mismatch between the different orders of approximation. If you were to estimate your model at order 2 or 3, the estimates would usually be different than at first order. That being said, what you describe is pretty is common in the literature.
Thanks a lot Pfeifer!