With the current version of Dynare, is it possible to compute optimal simple rules using a measure of welfare as the objective function or is the only allowed objective function still a weighted sum of the variances of some variables of the model?
Many thanks.