After I identified the order of my time series, I got ARMA(3,3). I then estimated the parameters, and found that coefficients AR(3) and MA(3) are essentially opposite of each other (0.9926 vs. -0.9713). Other coefficients are essentially zero.
Does this indicate that the underlying process is likely white noise and ARMA model is not suitable?
