The European Central Bank published a blog post explaining how it is reshaping its economic modelling after existing tools underestimated the post-pandemic inflation surge even when actual commodity prices, interest rates and exchange rates were fed back into the models. The main message is that the ECB is moving beyond more aggregated, linear frameworks toward a toolkit that is more granular, state-dependent and better able to handle large shocks, uncertainty and changing transmission mechanisms. The blog identifies two main weaknesses in the earlier framework: linear models did not capture the stronger pass-through from large energy shocks to inflation, and they understated indirect and second-round effects such as firms passing costs through more aggressively and wages catching up faster. In response, the ECB described new bottom-up energy models that split pre-tax energy prices into fuel, gas and electricity, use techniques for outliers and extreme volatility, and help assess risks from structural changes in energy markets, climate policies and extreme weather. It has also added new data sources and supply-constraint indicators to satellite models, re-estimated core projection models including ECB-BASE and NAWM II to improve treatment of commodity shocks and monetary policy transmission, and expanded the use of higher-frequency indicators, machine-learning-based inflation density forecasts and heterogeneous-agent models to assess distributional effects and interactions between monetary and fiscal policy. The blog says these changes are already feeding into projections and scenario analysis, with the March 2026 exercise adjusting pass-through elasticities to reflect the severity of a Middle East conflict scenario rather than relying on a fixed linear response. It also notes that work is continuing on models linking long-term structural issues with macro-financial dynamics, including defence spending, competitiveness policies, geopolitical developments, artificial intelligence, demographic change and long-term fiscal challenges.
European Central Bank2026-07-10
European Central Bank blog outlines upgraded modelling toolkit to capture non-linear inflation shocks
The European Central Bank said in a blog post that it is upgrading its modelling toolkit after older models understated the post-pandemic inflation surge. The revised framework adds more granular energy and supply-side modelling, reworked core projection models, higher-frequency data and machine-learning and heterogeneous-agent approaches. These tools are being used in projections and scenario analysis to better capture non-linear shocks and changing inflation transmission.