The Decomposition Structures in Two Seasonal Adjustment Programs X-12-ARIMA and TRAMO-SEATS - With Some Empirical Tests to Compare Both Programs -(in Japanese)

Two Widely-used seasonal adjustment programs are U.S.Census Bureau's X-12-ARIMA and TRAMO-SEATS written b V.Gomez and A.Maravall. The program TRAMO-SEATS, in particular, has beenattracting a considerable attention of researchers concerned with seasonal adjustment in Europe. X-12-ARIMA and TRAMO-SEATS are based on different methods for seasonal adjustment. X-11 which is the decomposition part in X-12-ARIMA is a method that uses an existing set of moving average filters. SEATS which is the decomposition part in TRAMO-SEATS is a method that uses ARIMA-model-based signal extraction with filters derived from an ARIMA-type time series model that describes the behavior of the series. But, RegARIMA in X-12-ARIMA and TRAMO in TRAMO-SEATS are very similar in that both programs are used for estimating regression models for time series, for outlier detection are correction, and for testing and correcting calendar effects, and so on. So, in this paper, we will focus on the decomposition structure in two programs X-12-ARIMA and TARAMO-SEATS, namely X-11 and SEATS. And, we will present some features of X-11 and SEATS. In addition , we will also carry out some empirical tests for the comparison of seasonal adjustment outcomes from X-12-ARIMA and TRAMO-SEATS.