Dinar simulations verification

Dear members,

One question was put to me lately about my simulations on Dynare and I could not answer it. The question is the following: how do we know for sure that the final results for the economy from Dynare’s simulations and estimations are correct? How could we assess its robustness and how could we verify it?
One parameters, lets say independent from the others (not that it matters much), could be wrong in its calibration and thus distort the final outcome. Or the equations could be misspecified and lead to a different outcome than the one we should expect.

Thank you in advance,


You will never know for sure, because you don’t know the true data generating process. But you can always check the sensitivity of your results with respect to sensible variations in parameters and do cross-validation by looking at untargeted moments/statistics or the model’s out of sample predictions.