Forecasting With Factor-Augmented Quantile Autoregressions: A Model Averaging Approach

Abstract

This paper considers forecasts of the growth and inflation distributions of the United Kingdom with factor-augmented quantile autoregressions under a model averaging framework. We investigate model combinations across models using weights that minimise the Akaike Information Criterion (AIC), the Bayesian Information Criterion (BIC), the Quantile Regression Information Criterion (QRIC) as well as the leave-one-out cross validation criterion. The unobserved factors are estimated by principal components of a large panel with N predictors over T periods under a recursive estimation scheme. We apply the aforementioned methods to the UK GDP growth and CPI inflation rate. We find that, on average, for GDP growth, in terms of coverage and final prediction error, the equal weights or the weights obtained by the AIC and BIC perform equally well but are outperformed by the QRIC and the Jackknife approach on the majority of the quantiles of interest. In contrast, the naive QAR(1) model of inflation outperforms all model averaging methodologies.