Skip to main content
EU Science Hub

Global Multi-country model

Image of a globe with graphics representing a global economic model

The GM model complements the QUEST III model, for macroeconomic surveillance, monitoring, forecasting and research.

Since 2014, the JRC and DG ECFIN develop the GM model (the EC Global Multi-country macro-economic model), to complement the QUEST III model for macroeconomic surveillance, monitoring, forecasting and research.

Download the GM model

The GM model belongs to the class of New-Keynesian Dynamic Stochastic General Equilibrium (DSGE) models. These models are core workhorses in international institutions and central banks as well as academic research.

The GM model offers different multi-country configurations: the GM2 version features the euro are (EA) and the rest-of-the-world (RoW); GM3-EMU versions focus on an individual Euro Area country together with the rest-of-the-EA (REA) and RoW; the GM3 version covers the EA, US and the RoW.

The GM model is estimated using Bayesian techniques, which allows combining a micro-founded model structure useful for policy analysis with a good probabilistic description of the observed data and forecasting performance. Recent extensions focus on macroeconomic nonlinearities, energy price dynamics, and the economic impact of the pandemic.

  • Since 2015, the GM model is regularly applied at all stages of DG ECFIN's Economic forecast. For example, it assesses revisions of external assumptions (such as oil prices) or provides alternative projections such as upside and downside scenarios during the pandemic. It also provides projections for macroeconomic aggregates and fiscal variables.
  • The GM model is used for shock decompositions, for example, to assess the main drivers of growth, inflation, and imbalances in historical data and current economic forecasts.
  • GM-based results have been published in peer-reviewed academic journals, fostering links to academia and the research community.

Further info and references on GM model

To find out more about the JRC's work on similar topics, explore the related JRC portfolios: