Forecasting in Factor Augmented Regressions under Structural Change

24 Pages Posted: 2 Nov 2021 Last revised: 19 Dec 2022

Date Written: December 15, 2022

Abstract

Factor augmented regressions are widely used to produce out-of-sample forecasts of macroeconomic and financial time series. However, these series are subject to occasional breaks. We study the effect of neglected structural instability on the forecasts produced by factor augmented regressions when the latent factors are estimated by cross-sectional averages from a large panel of variables. Our results show that neglecting structural instability can be very costly in terms of forecasting performance. We derive analytical results to show that both instability in the factor model and in the forecasting equation have an impact on the produced forecasts. We further provide numerical results showing that conditioning upon the most recent break tends to produce more accurate forecasts than unconditional estimation methods based on expanding or rolling windows, although the actual gain depends on the location and the magnitude of the breaks. Finally, an application to out-of-sample stock return forecasting using liquidity proxies illustrates the empirical relevance of our results.

Keywords: Factor Augmented Regression, Structural Instability, Out-of-Sample Forecasts, Estimation Window, Cross-Sectional Averages.

JEL Classification: C13, C32, C38, C53.

Suggested Citation

Massacci, Daniele and Kapetanios, George, Forecasting in Factor Augmented Regressions under Structural Change (December 15, 2022). Available at SSRN: https://ssrn.com/abstract=3952012 or http://dx.doi.org/10.2139/ssrn.3952012

Daniele Massacci (Contact Author)

King's College London ( email )

United Kingdom

George Kapetanios

King's College, London ( email )

30 Aldwych
London, WC2B 4BG
United Kingdom
+44 20 78484951 (Phone)

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

Paper statistics

Downloads
117
Abstract Views
671
Rank
507,721
PlumX Metrics