Adjustment of projections to recent observations

In economic forecasting with a dynamic model it is important to take proper account of recent statistical information. From such current indicators, non-zero residuals in the behavioural equations of the model can be estimated, and these will serve to improve the forecasts. Minimum norm estimates of...

Full description

Saved in:
Bibliographic Details
Published in:European economic review 1984-01, Vol.26 (1-2), p.153-166
Main Authors: Sandef, J, Don, F.J.H, van den Berg, P.J.C.M
Format: Article
Language:English
Subjects:
Algorithms
Economic forecasting
Economic indicators
Economic models
Forecasting techniques
Mathematical models
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:In economic forecasting with a dynamic model it is important to take proper account of recent statistical information. From such current indicators, non-zero residuals in the behavioural equations of the model can be estimated, and these will serve to improve the forecasts. Minimum norm estimates of the residuals given the observations on current indicators have proved to work quite well. We discuss a numerical example based on Central Planning Bureau forecasting practice, and conclude with thoughts on extensions and an account of some related work. An appendix discusses the conjugate gradient algorithm used to compute the minimum norm residuals.
ISSN:0014-2921
1873-572X
DOI:10.1016/0014-2921(84)90027-8