I have the results from a standard VAR model with Monte Carlo simulated confidence bands. I have the graphs for the impulse responses as well and I know that the shock size is one standard deviation.

As I need the size of the shock in terms of the impulse variable, I want to know if there is a way to approximate the standard deviation from the graphs?

  • $\begingroup$ The impulse response of variable x in period one will be the shock to variable x itself, won't it? Then you could just look at the first period's impulse response of the relevant variable. Or is the shock supposed to take place in period zero? Then still it should be quite easy to work out analytically the effect of the shock in period one, and you could solve it "backwards". $\endgroup$ – Richard Hardy Sep 14 '15 at 9:43
  • $\begingroup$ But when I have only the IRF graph from y's response to a shock in x in period zero? $\endgroup$ – jeffrey Sep 14 '15 at 12:58
  • $\begingroup$ If you have the estimated VAR model, you should be able to analytically derive what the impulse response will be. It should be quite easy to do that one or two periods ahead. Given that and given the observed outcome, you should also be able to derive the thing backwards: given the response, obtain the size of the shock that caused it. $\endgroup$ – Richard Hardy Sep 14 '15 at 13:23

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.