Implementation issues on shrinkage estimators for seasonal factors within the X-11 seasonal adjustment method
This commentary on Miller and Williams [Intl. J. Forecast. 20 (2004)S29-49] discusses how shrinkage can be implemented within X12-ARIMA. We discuss how the seasonal factors are estimated in X12-ARIMA, how shrinkage can be translated into a moving average, if this is compatible with the philosophy behind the X12-ARIMA method, and suggest possible improvements.