Combining Disaggregate Forecasts or Combining Disaggregate Information to Forecast an Aggregate

34 Pages Posted: 10 Mar 2010

See all articles by David F. Hendry

David F. Hendry

University of Oxford - Department of Economics

Kirstin Hubrich

Board of Governors of the Federal Reserve System

Date Written: July 23, 2009

Abstract

To forecast an aggregate, we propose adding disaggregate variables, instead of combining forecasts of those disaggregates or forecasting by a univariate aggregate model. New analytical results show the effects of changing coefficients, mis-specification, estimation uncertainty and mis-measurement error. Forecastorigin shifts in parameters affect absolute, but not relative, forecast accuracies; mis-specification and estimation uncertainty induce forecast-error differences, which variable-selection procedures or dimension reductions can mitigate. In Monte Carlo simulations, different stochastic structures and interdependencies between disaggregates imply that including disaggregate information in the aggregate model improves forecast accuracy. Our theoretical predictions and simulations are corroborated when forecasting aggregate US inflation pre- and post 1984 using disaggregate sectoral data.

Keywords: Aggregate forecasts, disaggregate information, forecast combination, inflation

JEL Classification: C51, C53, E31

Suggested Citation

Hendry, David F. and Hubrich, Kirstin, Combining Disaggregate Forecasts or Combining Disaggregate Information to Forecast an Aggregate (July 23, 2009). ECB Working Paper No. 1155. Available at SSRN: https://ssrn.com/abstract=1551223

David F. Hendry (Contact Author)

University of Oxford - Department of Economics ( email )

Manor Road Building
Manor Road
Oxford, OX1 3BJ
United Kingdom
+44 1865 278544 (Phone)
+44 1865 278557 (Fax)

Kirstin Hubrich

Board of Governors of the Federal Reserve System ( email )

20th Street and Constitution Avenue NW
Washington, DC 20551
United States

Register to save articles to
your library

Register

Paper statistics

Downloads
55
Abstract Views
406
rank
370,031
PlumX Metrics