Third order perturbation

Stochastic simulations
14

To get them less wiggly, you would need more replications (50000 or so), which is typically impractical.
There are essentially two ways to proceed:

  1. Use the Andreasen et al toolkit, which produces analytical GIRFs; it is not yet fully implemented into Dynare from our side, but we are working on it.
  2. Use IRFs around the stochastic steady state/EMAS. An example is at https://github.com/JohannesPfeifer/DSGE_mod/blob/master/Basu_Bundick_2017/Basu_Bundick_2017.mod
1 Like