hello guys,

in dynare, how to define SSS variable, \sigma_x?

Please have a look at https://github.com/JohannesPfeifer/DSGE_mod/blob/master/Basu_Bundick_2017/Basu_Bundick_2017.mod

Hi, Pfeifer, sometimes, the response of var(x_t) to a shock is negative, then i can’t compute sigma_x?

so there is not measured uncertainty of x_t.

similar problem as 3rd order simulation negative conditional variance

Any insights?

thanks in advanced

Wenddy

I am not sure I understand the issue. Are you talking about IRFs? In that case, the IRF can be negative if it’s relative to a positive steady state.

not IRFs, Var(x_t) is negative, so \sigma_{x,t} is complex ,

the formula for x’s volatility: \overline{\sigma}_x=100*ln(\frac{\sigma_{x,t}}{\sigma_x}) can’t be computed.

I am afraid that this can happen in perturbation solutions. Does that happen for all values of parameters, particularly all values for the shock volatilities?

I didn’t try all values for parameters and shock. is there a way to find the parameters which play key role?

You need to experiment with different calibrations and see what happens.