For a better user experience please update your browser or use Chrome or Firefox browser.
Utilities
Publication detail
HomeFinite approximations to linear filters and the monitoring of revisions in seasonally adjusted series
Publication metadata
Statistical offices involved in the production of seasonally adjusted series face the problem of revising preliminary figures. The process of revising until historical estimators are obtained can last relatively long, and, in general, the number of periods needed to obtain a final estimator cannot be controlled by the analyst. The length of the revision period is closely related to the estimation method; we shall focus on the model-based signal extraction approach (see for example Box, Hillmer and Tiao, 1978, and Burman, 1980). In this paper, we show how infinite seasonal adjustment filters can be optimally approximated by finite ones, and we apply this result to the problem of controlling the length of the revision period. We also show how considering finite versions of the signal extraction filters improves the interpretation of the X11 filters in the model-based framework.
- Published:
- Corporate author(s): Eurostat (European Commission)
- Personal author(s): Depoutot, Raoul ; Planas, Christophe
- Themes: Economy and finance
- Subject: statistical method
|
PDF
|
|||
| ISSN 1725-0595 | ISBN 92-894-5348-6 | DOI | Catalogue number KS-AN-03-020-EN-N |
- Released on EU Publications:
Document viewer
Note:
The publication cannot be displayed in the document viewer. Open web version
You may access it from: Download and languages