by jpfeifer » Thu Feb 05, 2015 7:44 am
As the above posts say, only the mean and standard deviation of the shocks matter, not the distribution itself. The decision rules will be correct for any distribution with the same mean and standard deviation as the one you provide. For simulations, things are more complicated as Dynare uses normally distributed shocks. In this case, if you want to generate data simulated with shocks from this distribution, you have to do that manually (but it is straightforward to do)