Stoch_simul after estimation, Smets-Wouters 2007

Hi everyone,

I’m attempting to simulate some data from Johannes’ Dynare code for Smets-Wouters 2007, but I’m slightly confused by the output.

I’ve added the stoch_simul command after estimation, and the Dynare code runs fine, but the units of the observations are a little worrisome. For example, looking at the output series ‘y’ yields both positive and negative values, but are these in log units? They seem too large to be the case.

But the growth rate equation specifying output growth is the difference, which leads me to believe that y is in log form.

Any clarification would be appreciated.


It’s a linearized model, so y is output deviations from steady state and as the data are scaled by 100 measured in percent, i.e. 1 is 1 percent.