How to interpret IRFs from shocks

Hello everyone,

I wanted to ask how to read IRFs interpretation from shocking the steady-state. Do I read that there is a 15 bps shock as I am multiplying by the standard error?
Is reading “stderr 1” when “sig_log_pi = 1” as 1% shock and
reading “stderr 1” when “sig_log_pi = 0.0015” as 15 bps shock correct?

Furthermore, sig_log_pi=1 and var e_log_pi; stderr 0.0025 equivalent to sig_log_pi=0.0025 and var e_log_pi; stderr 1? Both are interpreted as 100 bps shock?

var log_pi;
parameters sig_log_pi;
varexo e_log_pi;

sig_log_pi = 0.0015; rho=0.95;

model;
log_pi = rho*log_pi(-1) + sig_log_pi * e_log_pi;
end;

shocks;
var e_log_pi; stderr 1;

Yes, stderr 1 sets the standard error to 1 and Dynare by default uses a one-standard deviation shock. But your shock-term is sig_log_pi * e_log_pi, so its standard deviation is sig_log_pi *std(e_log_pi)=sig_log_pi. As you set sig_log_pi = 0.0015 and log_pi is measured in percent, this is a 15 basis point shock.
You will get the same if you use e_log_pi as the shock term with stderr 0.0015

Thank you for your reply and clarification. If my model is at quarterly level and I would like to see an annual 100 bps shock, would my standard error be stderr 0.0025 or stderr 1?

That depends on which of the two equivalent versions you use.