The RATS Software Forum

Hey Tom,

I have used RWZ approach to identify a VAR model using sign restrictions. However, I am confused regrading the interpretation of the impulse responses. The impulse responses are scaled to have unit variances. Do I need to re-scale them in terms of standard deviations as advised in programming guide by Walter Enders?

Thanks in advance

I'm not sure to what you are referring in the Enders' Programming Manual. The sign-restriction approach generates "impulse vectors" which are unit orthogonalized shocks, but they are natural scale in the actual variables.