I am analyzing the results of my VAR model, where the original non-stationary variables are presented as growth rates, and after differentiation they are interpreted as change of growth rates and used in the model.

Question: Which impulse-response function (irf) is appropriate for me to present, the original impulse response (which presents the increase of one unit of the differentiated variable) or the cumulative irf (which presents the increase of one unit of the growth rate, it escapes the differentiation, as I understand)?

For me it is logical to use cumulative irf, but I have not seen it used in any paper.

  • $\begingroup$ It looks like a subject-matter question rather than a statistical one. As long as both techniques are statistically sound (which they are), choose whichever makes more sense in terms of subject-matter interpretation. $\endgroup$ – Richard Hardy Mar 23 '17 at 12:53
  • 1
    $\begingroup$ Note for future reference when you are writing papers for publication: the English pronoun "I" is always capitalized, even when written in the middle of a sentence. $\endgroup$ – Chill2Macht Mar 23 '17 at 17:54

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.