Mixing deterministic and stochastic shocks:interpretation of the HPD interval

Hi everyone,

I’m doing some experiments mixing deterministic and stochastic shocks in a estimated DSGE model (basically integrating stoch_simul and forecast). All is fine, and the results are sensible (at least in the scenario I’m carrying out).

What I cannot understand is the right interpretation of the HPD interval (related to the use of forecast command)

In the manual, you say: ‘HPDinf: Lower bound of a 90% HPD interval of forecast due to parameter uncertainty, but ignoring the effect of measurement error on observed variables.’

What do you mean by parameter uncertainty? You mean all the parameters (including the calibrated ones) or just the estimated ones? Or do you mean the estimated standard deviations? Or something else?

Thank you,

Only estimated parameters come with uncertainty as their value is not exactly known. Calibrated parameters have fixed values with no uncertainty attached. Estimated standard deviations belong to the estimated parameters.