Bayes academic comments on Bank of England inflation forecasting
Professor Giovanni Urga discusses the BoE's failure to accurately predict the rise and stubbornness of inflation in the UK
The Bank of England (BoE) has launched a review of how it conducts its economic forecasting after being criticised for its failure to accurately predict the rise and stubbornness of inflation in the UK.
Following the economic shock of the pandemic, inflation has continued to increase and interest rates have risen from historic lows. Due to inflation remaining high, the money markets are now also pricing the possibility of interest rates reaching as high as 5.75% in February 2024.
Speaking about the BoE’s issues around accurately forecasting inflation, Giovanni Urga, Professor of Econometrics and Finance and Director of the Centre for Econometric Analysis (CEA) at Bayes Business School (formerly Cass), said:
“The issue is relevant at any level, and it is at the core of the macro-financial debate, with a particular focus on the relationship between interest rates and inflation, given the impact on both cost-of-living and mortgage rates.
“It is a fact that over the last year or so the Bank of England has systematically underestimated the persistence of inflation. The Bank has also just launched a review of how to make and use economic forecasts.
“In particular, it would be interesting to understand what drove the Bank’s inflation forecasts. In a recent speech by Jonathan Haskel, a member of the Bank of England’s Monetary Policy Committee, at the Peterson Institute for International Economics in Washington DC there was some insightful indication on ‘What’s driving inflation: wages, profits, or energy prices?’ A crucial issue is how to model the contribution of these variables.
"At the Centre for Econometric Analysis of Bayes Business School, we have leading research output showing that it is crucial to model time variation (via either Markov-Switching models and/or time-varying factor models and/or the dummy saturation approach) between macroeconomic variables. This is particularly true when you consider inflation and its main drivers, and our Research Fellow Lars Spreng has also shown in his PhD thesis that considering factors such as instabilities/time variation greatly improves out-of-sample forecasting accuracy.”