Why do we need the dependent variable to be stationary when using ARMA/ARIMA? Aren't we already accounting for that autocorrelation by using ARMA terms? Isn't the whole point to use its auto-correlation to predict itself?

Plus, I have tried fitting ARMA/ARIMA models to stationary (differenced) and level series and found that the levels series were much more conducive to ARMA specifications.


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.