We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Forecasting with measurement errors in dynamic models.
- Authors
Harrison, Richard; Kapetanios, George; Yates, Tony
- Abstract
This article explores a trade-off that confronts forecasters and monetary policy makers when they use data that are measured with error. The researchers begin by illustrating and proving how, faced with a choice between either using or not using the most recent observations in forecasting, if measurement error is sufficiently large it can be optimal not to use them. They move on to consider a case where measurement error is larger the more recent the data observation. Finally, They derive the optimal forecasting model from the class of linear autoregressive models. These forecasts are optimal subject to the constraint that the forecaster uses as many lags to forecast as are present in the data-generating process itself.
- Subjects
ECONOMIC policy; FISCAL policy; MONETARY policy; MONEY supply; STATISTICS; MONETARY theory
- Publication
Bank of England Quarterly Bulletin, 2004, Vol 44, Issue 4, p461
- ISSN
0005-5166
- Publication type
Article