Temporary Shocks

General DSGE Modeling
1

Hello Again,

I have a some clarifications regarding setting up the shock in DSGE:

  1. In Stochastic type of DSGE models a shock bloc as follows:

var epsi;
periods 1000:1002,1005;
values 0.20,0.1;

is a shock of magnitude 0.2 in the periods 1000:1002 and then 0.1 in period 1005 ?However, a shock in the following fashion:

shocks;
var psi = sigma^2;
end;

Is a shocks related to the variance of the variable, right?

  1. Suppose now I am interested in shocking the model first to induce recession per say and then shock it again using government expenditure/monetary policy response shocks e, then I write:

shocks;
var epsi;
periods 1000:1002;
values 0.20;
end;

followed by:

shocks;
var e;
periods 1003;
values 0.3;
end;

Am I right here or the two shocks must be in one bloc ?

Thanks a million for you help

2
  1. The syntax
var epsi;
periods 1000:1002,1005;
values 0.20,0.1;

is for perfect foresight simulations, while the syntax

shocks;
var psi = sigma^2;
end;

is for stochastic simulation (stoch_simul). For the latter, you cannot specify a shock trajectory, only a distribution of shocks. You can use the simult_-function for specific simulation paths in this case.

  1. No, multiple shocks-blocks in the perfect foresight syntax should not be concatenated.
3

I deeply value your comments