Premium
Revisions of time varying seasonal filters
Author(s) -
Dagum Estela Bee
Publication year - 1982
Publication title -
journal of forecasting
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.543
H-Index - 59
eISSN - 1099-131X
pISSN - 0277-6693
DOI - 10.1002/for.3980010204
Subject(s) - seasonal adjustment , autoregressive integrated moving average , seasonality , series (stratigraphy) , hodrick–prescott filter , environmental science , mathematics , statistics , econometrics , filter (signal processing) , magnitude (astronomy) , climatology , meteorology , time series , economics , geography , computer science , physics , geology , business cycle , mathematical analysis , paleontology , variable (mathematics) , astronomy , keynesian economics , computer vision
The main purpose of this study is to analyse the magnitude and the nature of the revisions that the time varying seasonal filters of the X‐II and X‐II‐ARIMA methods introduce in the current seasonally adjusted series The total revision is measured by the mean absolute difference of the transfer functions corresponding to the forecasting and the concurrent seasonal filters with respect to the central‘final’seasonal filter. To take into consideration the fact that the spectrum of a typical economic time series peaks at the low and seasonal frequencies, the revision measures are calculated for selected frequency intervals associated to the trend‐cycle, seasonal variations and the irregular component.