Forecasting substantial data revisions in the presence of model uncertainty

Release date
01/03/2006
Reference
DP2006/02
Authors
Anthony Garratt; Gary Koop; Shaun P. Vahey
Published as
Garratt, Anthony, Gary Koop and Shaun Vahey (2008). ‘Forecasting substantial data revisions in the presence of model uncertainty’, The Economic Journal, Wiley, Volume 118(530), Pages 1128-1144, DOI: https://doi.org/10.1111/j.1468-0297.2008.02163.x.
A recent revision to the preliminary measurement of GDP(E) growth for 2003 Q2 caused considerable press attention, provoked a public enquiry and prompted a number of reforms to UK statistical reporting procedures. In this paper, we compute the probability of “substantial revisions” that are greater (in absolute value) than the controversial 2003 revision. The predictive densities are derived from Bayesian model averaging over a wide set of forecasting models including linear, structural break and regime-switching models with and without heteroskedasticity. Ignoring the nonlinearities and model uncertainty yields misleading predictives and obscures the improvement in the quality of preliminary UK macroeconomic measurements relative to the early 1990s.