Are Disaggregate Data Useful for Factor Analysis in Forecasting French GDP?

30 Pages Posted: 14 Sep 2010

See all articles by Barhoumi Karim

Barhoumi Karim

Banque de France - Economic Study and Research Division

Olivier Darné

University of Nantes - Faculty of Business and Economics

Laurent Ferrara

Banque de France

Date Written: January 1, 2009

Abstract

This paper compares the GDP forecasting performance of alternative factor models based on monthly time series for the French economy. These models are based on static and dynamic principal components. The dynamic principal components are obtained using time and frequency domain methods. The forecasting accuracy is evaluated in two ways for GDP growth. First, we question whether it is more appropriate to use aggregate or disaggregate data (with three disaggregating levels) to extract the factors. Second, we focus on the determination of the number of factors obtained either from various criteria or from a fixed choice.

Keywords: GDP Forecasting, Factor Models, Data Aggregation

JEL Classification: C13, C52, C53, F47

Suggested Citation

karim, barhoumi and Darné, Olivier and Ferrara, Laurent, Are Disaggregate Data Useful for Factor Analysis in Forecasting French GDP? (January 1, 2009). Banque de France Working Paper No. 232, Available at SSRN: https://ssrn.com/abstract=1676797 or http://dx.doi.org/10.2139/ssrn.1676797

Barhoumi Karim (Contact Author)

Banque de France - Economic Study and Research Division ( email )

31, rue Croix des Petits Champs
75049 Paris Cedex 01
FRANCE

Olivier Darné

University of Nantes - Faculty of Business and Economics ( email )

Nantes
France

Laurent Ferrara

Banque de France ( email )

Paris
France

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Downloads
58
Abstract Views
864
Rank
692,737
PlumX Metrics