|
Activity Number:
|
320
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Tuesday, July 31, 2007 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Business and Economics Statistics Section
|
| Abstract - #309204 |
|
Title:
|
VAR Estimation and Forecasting When Data Are Subject to Revision
|
|
Author(s):
|
Evan Koenig*+ and N. Kundan Kishor
|
|
Companies:
|
Federal Reserve Bank of Dallas and Texas Tech University
|
|
Address:
|
Research Department, Dallas, TX, 75201,
|
|
Keywords:
|
data revisions ; forecasting ; Kalman filter ; vector autoregression
|
|
Abstract:
|
Conventional VAR estimation and forecasting ignores the fact that economic data are often subject to revision many months or years after their initial release. This paper shows how VAR analysis can be modified to account for such revisions. The proposed approach assumes that government statistical releases are efficient with a finite lag. It takes no stand on whether earlier revisions are the result of "news" or of a reduction in "noise." The technique is illustrated using several forecasting models of real economic activity. In each case, the proposed procedure outperforms conventional VAR analysis and the more-restrictive methods for handling the data-revision problem that are found in the existing literature. It yields forecasts of GDP growth and the unemployment rate that are more accurate than those from the Blue Chip newsletter and the Survey of Professional Forecasters.
|
