# How to change standard deviations of shocks？

Dear all,
I have a question concerning the size of standard deviations of shock .Impulse response function is the impact of 1 unit standard deviation by default.
In a linear model I have four shocks: monetary policy shock, technology shock, cost-push shock and housing preference shock. If I want to change housing preference shock to 5 unit standard deviation, the rest donot change. Is there a way to do that?
Thank you！

I am not sure whether I have gotten your question correctly or not, however I think You should set shocks block as follows

shocks;
var v; // defined shock as varexo
stderr 0.1; // standard error of shock
[and similarly for other shoks]
end;

then you can easily change stderr to have desirable simulations.

At first order, the IRFs are linear. Thus, the IRF to a 5 standard deviation shock will simply be five times the IRF to a 1 standard deviation shock. IF you want to simulate You could use the `simult_`-function to generate any shock combination you want. See

for an example.

jpfeifer，for example,I have four shocks: monetary policy shock, technology shock, cost-push shock and housing preference shock. I want to amplify the impact of monetary policy shock, I want monetary policy shock is 10 or 100 times larger than the other, how should I do？
should I change “var eps_r=0.01” to “var eps_r=0.1”?
“shocks;
var eps_r=0.1;
var eps_a=0.01;
var eps_pi=0.01;
var eps_h=0.01;
estimated_params;
stderr eps_r,inv_gamma_pdf,0.1,2;
stderr eps_a,inv_gamma_pdf,0.01,2;
stderr eps_pi,inv_gamma_pdf,0.01,2;
stderr eps_h,inv_gamma_pdf,0.01,2;

What exactly are you trying to do? If you are estimating your model, that will determine the appropriate standard devation.

jpfeifer，in my linear(ized) model, I want to change the size of 1 unit standard deviation about monetary policy shock， whether could I change the size of 5 unit standard deviation？
I’m sorry I don’t know how to change size of shocks.

Again, for which purpose do you want to change the shock size? Only for displaying IRFs? Because in the above code you posted it, it was about estimation.