I'm working on a differentiated seasonal time series with 2 breaks and non-zero mean. So, besides the constant i've got 2 dummies for breaks correction.

Question: when performing LB test, if my model is (2,1,1)(1,0,1) should i consider only p+q+P+Q coefs for subtracting degrees of freedom (meaning -5 degrees) or should i include all the constants (meaning -8 degrees)?

  • $\begingroup$ use all the implied coefficients .. less.8 in your case $\endgroup$ – IrishStat Jul 29 at 20:59
  • $\begingroup$ @IrishStat alright. But i just read that LB test does not perform well for ARIMAX model so i'm switching to BG test. Since bgtest() doesn't have a fitfd argument, should i subtract -8 from lag argument? $\endgroup$ – Alberson Miranda Jul 29 at 21:59
  • $\begingroup$ OK ............. $\endgroup$ – IrishStat Jul 29 at 22:29

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.