Forecasting in dynamic factor models subject to structural instability

Abstract

This chapter assesses forecasts constructed using dynamic factor models for their reliability in the face of structural breaks. Dynamic factor models have had notable empirical forecasting successes, but there has been little work to date on the performance of factor-based macroeconomic forecasts under structural instability. In factor models, even if factor loadings are unstable, if the instability is sufficiently independent across series then the estimated factors could be well estimated even if individual relations between the observable series and the factors are unstable. This chapter first lays out the implications for forecasting of different types of structural instability in dynamic factor models, provides a new empirical investigation (using US data for 144 quarterly macroeconomic time series) of factor-based forecasting with potential instability, and investigates separately the effects of structural change on the estimation of the factors and on the use of those factors for forecasting.