Abstract:We investigate whether the use of statistical learning techniques and big data can enhance the accuracy of inflation forecasts. We make use of a large dataset for the disaggregated prices of consumption goods and services, which we partially reconstruct, and a large suite of different statistical learning and traditional time-series models. The results suggest that the statistical learning models are able to compete with most benchmarks over medium to longer horizons, despite the fact that we only have a relatively small sample of available data. This may imply that the ability of statistical learning models to explain nonlinear relationships, or as an alternative, restrict the set of predictors to relevant information, is of importance. These characteristics of the statistical learning models may be particularly useful during periods of crisis, when deviations from the steady state are more persistent. We find that the accuracy of the central bank's near-term inflation forecasts compares favourably with those of other models, while the inclusion of off-model information, such as electricity tariff adjustments and other sources of within-month data, provides these models with a competitive advantage. Lastly, we also investigate the relative performance of the different models as we experienced the effects of the recent pandemic and identify the most important contributors to future inflationary pressure. Supplementary information: The online version contains supplementary material available at 10.1007/s00181-022-02329-y.

Forecasting UK inflation bottom up

Forecasting of Inflation Rate Contingent on Consumer Price Index: Machine Learning Approach

China's inflation forecasting in a data-rich environment: based on machine learning algorithms

Forecasting Inflation Using Economic Narratives

Forecasting in Data-Rich Environments: An Application to Inflation Rates

Real-time Inflation Forecasting Using Non-linear Dimension Reduction Techniques

Big data forecasting of South African inflation

Forecasting Inflation with Economic Narratives and Machine Learning

Assessing the Impact of Goods and Services Inflation on Consumer and Professional Forecaster Expectations: A Bayesian Shrinkage Approach

Forecasting Inflation in a Data-Rich Environment: The Benefits of Machine Learning Methods

Forecasting CPI Inflation Components with Hierarchical Recurrent Neural Networks

Forecasting euro area inflation using a huge panel of survey expectations

Inflation Forecasting from Cross-Sectional Stocks

Forecasting Inflation with the New Keynesian Phillips Curve: Frequencies Matter

An evaluation of the inflation forecasting performance of the European Central Bank, the Federal Reserve, and the Bank of England

Forecasting inflation using sentiment

Inflation Prediction Method Based on Deep Learning

Forecasting CPI inflation under economic policy and geopolitical uncertainties

Out-of-Sample Forecasts of China’s Economic Growth and Inflation Using Rolling Weighted Least Squares

Inflation forecasting in turbulent times

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