0

I'm trying to build an ARIMA model and choose the appropriate order of AR(p) and MA(d). I'm using ACF and PACF plots. They look very similar, very same (please take a look).

ACF and PACF plots

Could you please give me some hint on why is that and whether I can build a corrected model (if it's ok)? Based on the plots I think that ARIMA (2, 1, 2) can be an option.

Thank you

Claudia
  • 1
  • The correlations are very weak and most are not significant. I do not think you can build an ARIMA model with this data. – Arne Decker May 11 '22 at 06:29
  • As Arne said, it doesn't appear to be an ARMA model. You could try bootstrap methods to be sure (tsboot in R), but your data does not appear to have autorregresive components. – Jesús A. Piñera Jan 17 '23 at 19:24

0 Answers0