I am trying to estimate ARMA-GARCH model for my stock returns time series.

I have estimated ARMA model for my series, and found that there exists ARCH, so added GARCH(1,1) term. However I now find previously significant ARMA coefficients being insignificant.

In this case, should I remove the insignificant ARMA terms? I am reluctant in doing so as I read from a book that it is not wise to remove ARMA terms judging from their significance.

Also, as I am not performing any predictions with this model (only using it as a normal return model for event study), I thought it would be unnecessary to do so?

  • $\begingroup$ Could you give some more detail in how you are going to use the model? $\endgroup$ Feb 21, 2017 at 7:49
  • $\begingroup$ I specify my base model usimg arma-garch, then add date dummies to check for significance. If dummies are significant, that means there is an abnormal return. And no abnormal returns when dummies are insignificant $\endgroup$
    – R.Lee
    Feb 21, 2017 at 10:31
  • $\begingroup$ What do you think about my answer? $\endgroup$ Mar 8, 2017 at 19:40
  • $\begingroup$ Please beware that statistical significance is not a black or white problem. The addition of extra terms increases estimation errors which themselves reduce significance. So I wouldn't worry too much if a 0.03 p-value suddenly becomes a 0.07 (I will assume 5% thresholds, but this applies to any significance level) This phenomenon is normal and there is nothing "magical" about 5%. I would be concerned, though, if a $10^{-8}$ p-value suddenly became a $0.38$, though. Which of these is your case? $\endgroup$
    – David
    Jun 28, 2019 at 8:38

1 Answer 1


This boils down to the general question of model selection with the goal of explanatory modelling and statistical inference. Indeed, statistical significance may not be a sound approach to feature selection as per Rob J. Hyndman's blog post "Statistical tests for variable selection" and Frank Harrell's "Regression Modelling Strategies" referenced therein (although note that Hyndman is more interested in forecasting than explanatory modelling).

In practice, I would select the model from a reasonable pool of candidates (here probably comprising the model with the insignificant coefficient, the model without it, and perhaps some other models) using information criteria such as AIC or BIC. But I would keep in mind that inference from this model will be conditional on the selected model being the "true" model, which is an uncomfortable qualification (see answers -- if there are any -- to my question "Statistical inference under model misspecification" and check out threads referenced therein; this is a basic and ubiquitous problem in applied modelling that I do not know a good solution for).


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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