seasonally adjusted series

[newsagent]

Quite a few times we download data from various organizations (IMF, WB, etc) where the series are already seasonally and calendar adjusted. In this case, can we still use SARIMA or just ARIMA with PDQ(0,0,0)?

I am saying this because I have applied SARIMA to seasonally adjusted series and the ARIMA function suggested to include SAR and SMA terms. What should I do in this case?

[robjhyndman]

Seasonal adjustment is meant to remove the seasonality. If the ARIMA() function is suggesting the inclusion of seasonal components, that suggests that the seasonal adjustment has not completely removed the seasonality in the original data. It's up to you whether you want to try to model the remaining seasonality, or just use the ARIMA model to handle the non-seasonal dynamics in the data.