2nd order approximation and differences in means


I am trying to simulate the 2nd order approximation for my model and I keep getting ‘NaN’ for my moments of simulated variables. However, I am able to simulate under the 1st order approximation. Additionally, is it strange that under the 1st order approximation, I keep getting different means?

Thank you so much

  1. Without pruning that can happen.
  2. What do you mean with

Simulated moments will depend on the random numbers used.

Thank you! I managed to fix that.

With regards to the second question about the changing mean, everytime I perform a simulation – I keep getting different simulated moments. I wanted to find out if there was any way I could avoid this as I am doing a policy analysis?

Why can’t you use theoretical moments?

Thank you for your response. I wanted to find out how I can get these theoretical moments?

You need to set periods=0

Thank you, I have done this and I keep getting:

Error using stoch_simul (line 72)
STOCH_SIMUL error: number of periods for the simulation isn’t specified

Error in allinvesting (line 632)
info = stoch_simul(var_list_);

Error in dynare (line 102)
evalin(‘base’,fname) ;

I would appreciate any assistance with this please

Please provide the mod-file.

b.mod (10.4 KB)

benchmark_params.mat is missing.

There are still various dependencies you did not provide.

broyden.m (3.2 KB) broydeni.m (2.2 KB) broydenx.m (1.9 KB) dpcheck.m (2.8 KB) dprod.m (1.4 KB) dpsimul.m (2.1 KB) dpsolve.m (12.3 KB) dpss.m (971 Bytes) dpssres.m (1.0 KB)

Please let me know if it works now.

No, optget is missing. Why can’t you simply provide the mat-file?

Dear Professor,

I managed to fix it - thanks so much!