My goal is to show that t-GAS is better than t-GARCH, so I am trying to analyze some data about Crude Oil volatility comparing these models with functions in R packages (GAS and rugarch). After estimation, values of AIC, BIC, LL are better in GARCH. How can I resolve it? Thank you!

  • $\begingroup$ Please clarify your specific problem or provide additional details to highlight exactly what you need. As it's currently written, it's hard to tell exactly what you're asking. $\endgroup$
    – Community Bot
    Nov 9, 2022 at 8:47
  • $\begingroup$ What do you think about my answer? If it is helpful and clear, you may accept it by clicking on the tick mark to the left. Otherwise, you may ask for further clarification. This is how Cross Validated works. $\endgroup$ Dec 7, 2022 at 9:04

1 Answer 1


Your goal is to show a result you think is true before seeing the data. That rings some bells... (Why consult data if you already know the answer?) And the data is not willing to cooperate with you. You may question the data, the implementation of the models or... your preconception.

  • Might it not be that GARCH beats GAS in this instance?

Of course, the data and the implementation may also be faulty.

  • Are you sure AIC, BIC and LL are comparable across these model classes? Statistical software often report incorrect values of AIC, BIC and LL by omitting some constants from the calculations. This still allows comparisons within model classes (such as GARCH(1,1) vs. GARCH(2,2)) but not necessarily across them (GARCH vs. GAS). Also, e.g. the rugarch package in R used to, and perhaps still continues, to report AIC and BIC per observation, thus divided by the sample size, while most other packages report raw AIC and BIC.

Your Answer

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

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