On May 7th, the Network for Greening the Financial System (NGFS)—a coalition of central banks and financial supervisors—released the results of four short-term climate change scenarios. These scenarios aim to illustrate possible global conditions five years from now, based on the realization of distinct climate and policy pathways.
Each scenario is built on assumptions regarding transitional policies, actions taken during the period, and the occurrence of extreme climate events such as storms, droughts, and floods.
The NGFS integrates these assumptions into a climate change modelling framework, producing projections for macroeconomic, macro-financial, and sector-specific indicators—such as inflation, GDP, productivity, unemployment, cost of capital, credit spreads, and the risk-free rate. This is visualized in the accompanying graph, where gray boxes represent scenario inputs, outputs, variables, and assumptions, while blue boxes denote the models used.
Each scenario includes a set of quantitative characteristics, accompanied by a high-level narrative to convey its defining features:
This scenario aligns with the goals of the Paris Agreement, limiting global warming to no more than 1.5°C above pre-industrial levels. Green technologies drive economic growth, and revenues from carbon taxes are reinvested into green public projects. Consumer and investor preferences shift toward low-carbon sectors, while carbon-intensive industries face rising credit risks and financing costs.
In this scenario, climate risks are largely ignored until a sudden, disruptive event forces the rapid introduction of climate policy. The energy sector, heavily reliant on fossil fuels, faces a sharp rise in prices. This triggers a drop in demand, a fall in risk-free rates, and a crisis of confidence. Investment declines, and financial institutions experience elevated credit, market, and liquidity risks.
The world remains dependent on fossil fuels, and climate policy remains stagnant. Severe disasters in 2026 and 2027 lead to asset destruction and productivity losses in some regions, with ripple effects across the global economy.
Advanced economies (e.g. North America, Europe, Oceania, and parts of Asia) commit to net-zero greenhouse gas emissions. Meanwhile, other regions are hit by frequent and intense climate events. These shocks propagate globally through trade disruptions and financial contagion, creating an uneven global transition.
The process begins by feeding a selected scenario—with its specific assumptions on physical climate events and transition policies—into the GEM-E3 model (General Equilibrium Model for Energy, Economy, and Environment). This is a computable general equilibrium model that represents the entire economy through a system of equations describing the behavior and interactions of all economic agents. It is calibrated using empirical data.
The model assumes neoclassical behavior: firms aim to maximize profit, and households aim to maximize welfare. GEM-E3 generates projections including:
Full Input-Output tables by country/region
National accounts
Employment by sector
Unemployment rate
Public finance and revenues
Household consumption
Energy production and use
Greenhouse gas emissions and atmospheric pollutants
GEM-E3 is developed by E3 Modeling, a firm based in Athens.
Some outputs from GEM-E3 are then used as inputs to the CLIMACRED model, which is based on the 2023 paper “Climate Credit Risk and Corporate Valuation” by Stefano Battiston, Antoine Mandel, Irene Monasterolo, and Alan Roncoroni.
CLIMACRED is used to assess the financial valuation of corporate bonds and equity, including:
Probability of default
Cost of capital
Valuation of financial instruments
This enables an estimation of climate-related credit risks and market-based pricing effects on a firm's equity.
The EIRIN model takes projected carbon prices from GEM-E3 as input. Unlike GEM-E3, EIRIN is a Stock-Flow Consistent (SFC) macroeconomic model that does not assume equilibrium. Developed in the paper “The EIRIN Flow-of-funds Behavioural Model of Green Fiscal Policies and Green Sovereign Bonds”, it represents an economy through a finite set of agents linked by monetary flows between their balance sheets.
EIRIN incorporates bounded rationality and imperfect information, with agents such as:
Wage-earning households
The consumption goods and service sectors
High-carbon capital goods producers
Utility companies
Central and commercial banks
Fiscal regulators
The structure of this scenario-modelling framework highlights its complexity and multi-layered nature. Naturally, several important questions arise:
Why is the GEM-E3 model run twice?
How are differing assumptions across models reconciled?
For instance, GEM-E3 assumes rational optimizing agents, while EIRIN assumes bounded rationality. Does this pose a theoretical inconsistency?
How are overlapping outputs managed?
For example, the risk-free rate is determined in both GEM-E3 and EIRIN. If different values arise, which is used?
You can read the full report on the NGFS short-term climate scenarios here:
🌍 NGFS Short-term Climate Scenarios for Central Banks and Supervisors