When running varselect in R, I usually get a few different models to choose from based on different statistics. I know of:

  1. Akaike information criterion (AIC)
  2. Bayesian information criterion (BIC) or Schwarz criterion (SC)
  3. Final Prediction Error (FPE) criterion
  4. Hannan–Quinn information criterion (HQC)

What is the practical difference between these tests and under what circumstances should I prefer one over the other?


1 Answer 1


Monte Carlo evidence suggests that the BIC penalty function works well in designs calibrated to macroeconomic data (Stock and Watson, 2011). It is common to run them all for comparison purposes.

  • $\begingroup$ Welcome to CV. Please add full reference in case your link dies in the future. Thanks $\endgroup$
    – Antoine
    Jul 5, 2021 at 15:15

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.