Portal de Periódicos da CAPES

Seasonal Adjustment Based on ARIMA Model Decomposition: TRAMO-SEATS

2016; Springer Nature; Linguagem: Inglês

10.1007/978-3-319-31822-6_5

2199-7365

Estela Bee Dagum, Silvia Bianconcini,

Forecasting Techniques and Applications

TRAMO-SEATS is a seasonal adjustment method based on ARIMA modeling. TRAMO estimates via regression dummy variables and direct ARIMA modeling (regARIMA) the deterministic components, trading days, moving holidays, and outliers, which are later removed from the input data. In a second round, SEATS estimates the stochastic components, seasonality, and trend-cycle, from an ARIMA model fitted to the data where the deterministic components are removed. SEATS uses the filters derived from the ARIMA model that describes the behavior of the time series. By imposing certain conditions, a unique canonical decomposition is performed to obtain the ARIMA models for each component. This chapter discusses with details the estimation methods used by TRAMO and SEATS as well as the basic assumptions on the derivation of ARIMA models for each component. An illustrative example of the seasonal adjustment with the TRAMO-SEATS software default option is shown with the US New Orders for Durable Goods series. The illustrative example concentrates on the most important tables of this software that enable to assess the quality of the seasonal adjustment.