Recalibrating Climate Risk

Recalibrating Climate Risk explains why economic models used by governments, central banks and investors are increasingly understating climate risks as the world moves towards 2°C. It shows how this can create a false sense of security – and why decision-makers should act now rather than wait for perfect models. 

Led by the University of Exeter (Green Futures Solutions), in partnership with Carbon Tracker, the report draws on structured expert judgement from climate scientists across 12 countries to clarify where today’s ‘damage models’ fall short and what decision-makers should do to manage investment risks under rising uncertainty.  

The report builds on earlier work challenging the under-pricing of climate damages in financial decision-making, including Carbon Tracker’s Loading the DICE Against Pensions (2023) and The Emperor’s New Climate Scenarios (IFoA/University of Exeter, 2023). 

Key findings 

• Physical climate damages are structural and compounding. At higher levels of warming, impacts are more likely to cascade across sectors and geographies, undermining the conditions economies rely on for stable growth. 
• Real-world losses are shaped by local and regional extremes (heatwaves, floods, droughts) that can be poorly captured by models focused on global average temperature. Many approaches still link damages to global mean temperature, even though disruption is often driven by local and regional extremes. 
• GDP-based metrics can undercount losses linked to mortality and morbidity, inequality, ecosystem degradation and social disruption, and can even appear to recover after disasters through reconstruction spending. GDP can mask welfare losses and distributional impacts, including when reconstruction spending raises GDP after disasters.  
• Uncertainty increases sharply towards and beyond 2°C. Tail risks and tipping points become more relevant, yet models often present precise-looking point estimates that may not be decision-useful in deep uncertainty. As warming rises, point estimates can look precise while becoming less reliable for decision-making under deep uncertainty. 
• The report argues against waiting for perfect modelling, and instead calls for governance, supervision and investment practice to be recalibrated towards precaution, robustness and transparency.