Jennifer L. Castle, Jurgen A. Doornik and David F. Hendry

Climate Econometrics, Nuffield College, University of Oxford, New Road, OX11NF, UK.

A sequence of increasingly large same-sign 1-step-ahead forecast errors are most likely due to a sudden unexpected shift. Large forecast errors can be expensive, but also contain valuable information that has not been utilised to date.

Impulse indicators can act as intercept corrections (ICs) to set forecasts back on track at the forecast origin. These ICs can be quickly tested to determine whether they can be replaced by a location shift or broken trend, greatly improving forecast accuracy.

The analysis is applied to forecasting the log of the UK’s annual consumer price inflation (CPIH), which rose rapidly from mid-2021 to over 9% in 2022 after a series of four essentially unpredictable shocks, and led to large forecast errors by the Bank of England. Each trend shift was followed by significant forecast errors, but our approach, using the information in those forecast errors, experienced only seven large errors overall by rapidly detecting breaks, avoiding long periods of systematic forecast failure.

The data and PcGive batch file as well as the resulting output can be downloaded here.