• I have two series, price and sales.
  • Sales is mean-reverting stationary but price is stationary only after controlling for an intercept break.
  • I want to set up a 2-equation VAR model and the research interest is to estimate the cumulative effect of price on sales through impulse response function.

My question: Is the impulse-response function (irf) of a price shock on sales still biased even after I include the break dummy as a regressor in the two equations? Say the VAR model is: \begin{aligned} Sales_t &= \beta_{10} + \beta_{11}Sales_{t-1} + \beta_{12}Sales_{t-2} + \beta_{13}Price_{t-1} + \beta_{14}Price_{t-2} + \beta_{15}D_t + e_t \\ Price_t &= \beta_{20} + \beta_{21}Sales_{t-1} + \beta_{22}Sales_{t-2} + \beta_{23}Price_{t-1} + \beta_{24}Price_{t-2} + \beta_{25}D_t + e_t \end{aligned}

My answer is yes, irf is still biased because the regressors $Price_{t-1}$ and $Price_{t-2}$ are still nonstationary.

My solution: include both equations the interaction terms: $\beta_{6}Price_{t-1}\times D_t$ and $\beta_{7}Price_{t-2}\times D_t$.

Would you please assess the above my question, my answer, and my solution?


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.