Forecasting data revisions of GDP: a mixed frequency approach |
| |
Authors: | Jens Hogrefe |
| |
Affiliation: | (1) Institut für Weltwirtschaft, Düsternbrooker Weg 120, 24105 Kiel, Germany |
| |
Abstract: | Releases of GDP data undergo a series of revisions over time. These revisions have an impact on the results of macroeconometric models documented by the growing literature on real-time data applications. Revisions of U.S. GDP data can be explained and are partly predictable according to Faust et al. (J. Money Credit Bank. 37(3):403–419, 2005) or Fixler and Grimm (J. Product. Anal. 25:213–229, 2006). This analysis proposes the inclusion of mixed frequency data for forecasting GDP revisions. Thereby, the information set available around the first data vintage can be better exploited than the pure quarterly data. In-sample and out-of-sample results suggest that forecasts of GDP revisions can be improved by using mixed frequency data. |
| |
Keywords: | Data revisions GDP Real-time data Mixed frequency Factor model |
本文献已被 SpringerLink 等数据库收录! |
|