Magyar nyelven nem elérhető
Alessandro Giammaria
- 12 June 2025
- WORKING PAPER SERIES - No. 3062Details
- Abstract
- Energy inflation is a major source of headline inflation volatility and forecast errors, therefore it is critical to model it accurately. This paper introduces a novel suite of Bayesian VAR models for euro area HICP energy inflation, which adopts a granular, bottom-up approach – disaggregating energy into subcomponents, such as fuels, gas, and electricity. The suite incorporates key features for energy prices: stochastic volatility, outlier correction, high-frequency indicators, and pre-tax price modelling. These characteristics enhance both in-sample explanatory power and forecast accuracy. Compared to standard benchmarks and official projections, our BVARs achieve better forecasting performance, particularly beyond the very short term. The suite also captures a sizable variation in the impact of commodity price shocks, pointing to higher elasticities at higher levels of commodity prices. Beyond forecasting, our framework is also useful for scenario and sensitivity analysis as an effective tool to gauge risks, which is especially relevant amid ongoing energy market transformations.
- JEL Code
- C32 : Mathematical and Quantitative Methods→Multiple or Simultaneous Equation Models, Multiple Variables→Time-Series Models, Dynamic Quantile Regressions, Dynamic Treatment Effect Models, Diffusion Processes
C53 : Mathematical and Quantitative Methods→Econometric Modeling→Forecasting and Prediction Methods, Simulation Methods
E31 : Macroeconomics and Monetary Economics→Prices, Business Fluctuations, and Cycles→Price Level, Inflation, Deflation
E37 : Macroeconomics and Monetary Economics→Prices, Business Fluctuations, and Cycles→Forecasting and Simulation: Models and Applications
- 15 March 2024
- OCCASIONAL PAPER SERIES - No. 344Details
- Abstract
- This paper takes stock of the ECB’s macroeconometric modelling strategy by focusing on the models and applications used in the Forecasting and Policy Modelling Division. We focus on the guiding principles underpinning the current portfolio of the main macroeconomic models and illustrate how they can in principle be used for economic forecasting, scenario and risk analyses. We also discuss the modelling agenda which is currently under development, focusing notably on heterogeneity, machine learning, expectation formation and climate change. The paper makes it clear that the large macroeconometric models typically developed in central banks remain stylised descriptions of our modern economies and can fail to predict or assess the nature of economic events (especially when big crises arise). But even in highly uncertain economic conditions, they can still provide a meaningful contribution to policy preparation. We conclude the paper with a roadmap which will allow the ECB and the Eurosystem to exploit technological advances and cooperation across institutions as a useful means of ensuring that the modelling framework is not only resilient to disruptive events but also innovative.
- JEL Code
- C30 : Mathematical and Quantitative Methods→Multiple or Simultaneous Equation Models, Multiple Variables→General
C53 : Mathematical and Quantitative Methods→Econometric Modeling→Forecasting and Prediction Methods, Simulation Methods
C54 : Mathematical and Quantitative Methods→Econometric Modeling→Quantitative Policy Modeling
E52 : Macroeconomics and Monetary Economics→Monetary Policy, Central Banking, and the Supply of Money and Credit→Monetary Policy