Randomness of simulated data using 'stoch_simul'

Stochastic simulations
1

Dear Johannes,
Thank you very much for your helpful guidance, I am grateful. I have some questions about the command ‘stoch_simul’.

  1. When we run the model each time and simulate observable variables using ‘stoch_simul’, we will get different simulated observable variables due to randomness involved right?
  2. The more simulated sample size we have when we simulate observable variables using ‘stoch_simul’, the more likely we get better statistics from observable variables such that they look more similar to that of actual data right?
    Thank you very much and look forward to hearing from you.
    Best regards,
    Jesse
2
  1. In principle yes, but Dynare sets a seed to get reproducible results across runs of Dynare.
  2. It depends on what you are after. With more draws, the simulated moments should better approximate the theoretical/asymptotical ones. But sometimes you are interested in small sample properties.