Unmaterialized News Shock clarification

Stochastic simulations
1

Hi all,

I have a question on the application of the anticipated news shock. (e.g., https://github.com/JohannesPfeifer/DSGE_mod/blob/master/RBC_news_shock_model/RBC_news_shock_model.mod)

The shocks are applied as follows:

//8. exogenous TFP process
z=rhoz*z(-1)+eps_z_surprise + eps_z_news(-8);

shocks;
    var eps_z_news=1; //8 period anticipated TFP news shock
    var eps_z_surprise=1; //TFP surprise shock
end;

However, I don’t understand how these shocks offset each other given they both have positive signs (z=rhoz*z(-1)+eps_z_surprise + eps_z_news(-8))

Shouldn’t it be:

//8. exogenous TFP process
z = rhoz*z(-1) - eps_z_surprise + eps_z_news(-8);

shocks;
    var eps_z_news=1; //8 period anticipated TFP news shock
    var eps_z_surprise=1; //TFP surprise shock
end;

or

//8. exogenous TFP process
z= rhoz*z(-1) + eps_z_surprise + eps_z_news(-8);

shocks;
    var eps_z_news = 1; 
    var eps_z_surprise = -1 ; 
end;

And if these alternatives don’t work, what part of the original code implies the shocks are actually offsetting each other rather than going in the same direction?

For example, I was thinking of changing the shocks block so that it accounts for exactly offsetting (corr = -1) shocks:

shocks;
var eps_z_news = 1; 
var eps_z_surprise = 1 ;
corr eps_z_news, eps_z_surprise = -1;
end;

Thanks,

Camilo

2

Given that the shocks themselves can be positive or negative, the use of a plus is simply a convention. Also note that negative variance in the shocks do not make sense.

What is important is that in the simult_-function used to simulate the model the second (surprise) shock indeed has a negative sign.

3

Thank you, Johannes-

In your github those lines looked commented so I failed to see the options used in simult.

Now everything’s clear.

Best,

Camilo

4

Yes. Github’s syntax highlighting does not work well for Dynare code.

5

when replicating the code, the following errors showed up:

Error using simult_
Too many input arguments.

Error in RBC_news_shock_jpfeifer (line 239)
y2 = simult_(M_,options_,initial_condition_states,oo_.dr,shock_matrix,1);

Error in dynare (line 223)
evalin('base',fname) ;

how to fix that?

6

Is it only me who met such an issue? Or it’s sth related with my dynare/matlab version?

7

The codes are exactly the same from
https://github.com/JohannesPfeifer/DSGE_mod/blob/master/RBC_news_shock_model/RBC_news_shock_model.mod

8

I fixed it by deleting the first two inputs:

y2 = simult_(initial_condition_states,oo_.dr,shock_matrix,1);

I referred to BreakingFeaturesIn4.6 · Wiki · Dynare / dynare · GitLab
This forum is really helpful! Thank you all guys!

9

Please use the most recent Dynare version.