Skip to content

Content Library

Back to Articles

Article October 10, 2025

Climate Stress Tests Are Broken - Banks Must Build Their Own

Climate Stress Tests Are Broken - Banks Must Build Their Own

Lara Hall, Event Producer

Climate scenarios built for science don’t translate cleanly into finance. Central bankers are adding short-term adversity and sector detail, while banks are engineering in-house models, validation methods and narratives to turn climate risk from awareness

• Early climate scenarios lack short-term adversity and sector detail
• Central bankers are adding policy and market shocks with PD and spread impacts
• Banks are building agent-based models and SSP – RCP methods to own assumptions
• Qualitative 10-year narratives guide strategy but need quantitative bridges
• Validation shifts to plausibility, traceability and explainability
• Data consistency and geolocation are prerequisites for credible analysis
• Nature risk demands localized data and supply-chain transmission mapping
• Goal is full integration of climate risk into standard stress testing
• Internal models improve ownership, transparency and decision usefulness
• The focus moves from awareness to action and integration

Financial institutions are learning a hard truth about climate risk – models designed for scientific assessment rarely work as plug-and-play tools for balance-sheet decisions. 

Central banking officials acknowledged that early scenario vintages lacked short-term adversity and sector granularity, prompting a pivot to more finance-ready narratives that include policy shocks, market stress and disruptive physical events with impacts on credit spreads and default probabilities. 

Speaking at a recent CeFPro conference on climate risk, the head of climate economics at a central European bank described the agenda bluntly.

For him, there is a clear need for more granular sector pathways using input-output tables, tougher damage functions to capture nonlinear losses, and dedicated short-term scenarios that look and feel like macro stress tests rather than academic pathways. 

The aim, he said, is consistency of data and methods – and integration of climate alongside other risks in standard supervisory exercises. 

Banks are not waiting. In the same session, a managing director and head of risk analytics outlined a two-track build – an agent-based macro model calibrated to internal assumptions, and a rapid “SSP – RCP” pathway method that derives deflationary effects from temperature and inflationary effects from carbon prices, then propagates them through sectors to credit. 

The driver is ownership and explainability, not black-box dependencies. 

A head of climate risk analytics added that qualitative scenario narratives are also proving decisive: decade-long storylines that translate physical hazards and transition uncertainty into concrete strategic choices for sectors and portfolios. 

Those narratives, he claims, underpin internal macro scenarios, credit discussions and climate scorecards – but still need quantitative bridges to external frameworks to flow into capital and pricing. 

His institution’s validation teams now anchor reviews on plausibility, traceability and explainability rather than traditional back-tests. 

That shift matters because governance is the bottleneck. Traditional validators lack climate-finance depth, and many empirical links – such as sector value added to credit – remain weak. 

It became clear in the session that shared standards and transparent assumptions are emerging as the pragmatic compromise: rigorous enough to satisfy oversight, flexible enough to keep innovation moving. 

Events you may be interested in

View all of our upcoming events

Sign Up to Our Newsletter

Sign up

Connect Plus

Unlock more with Connect Plus membership

This content is available to Connect Plus members.