Climate risk after the retreat: why the prudential case outlasts the pledges

Two years ago, climate strategy in banking was framed by public pledges: net-zero targets, alliance memberships and sector decarbonisation pathways. That framing has largely collapsed. What remains is less visible but more durable: supervisors in Europe and the UK treating climate as a driver of credit, market and operational risk, and expecting banks to manage it with the same rigour and data as any other.

The retreat from voluntary commitments

The Net-Zero Banking Alliance voted on 3 October 2025 to cease operations and move to a guidance-based model, after a wave of departures that began with the six largest US banks in December 2024 and January 2025 and spread to Canadian, UK and Swiss institutions; its target-setting guidance remains publicly available¹. Two weeks later, the Federal Reserve, FDIC and OCC withdrew their 2023 principles for climate-related financial risk management for institutions with more than $100 billion in assets, saying existing safety and soundness standards already require banks to address all material risks².

Global standard setters have also softened their ambitions. In June 2025 the Basel Committee published its climate-related financial risk disclosure framework as a voluntary framework, leaving jurisdictions to decide whether to implement it, and acknowledging that climate data quality is still evolving³. In the EU, the Omnibus I directive, in force since March 2026, narrowed the Corporate Sustainability Reporting Directive to companies with more than 1,000 employees and over €450 million turnover, and removed the due-diligence directive's requirement for companies to adopt climate transition plans⁴. For banks, that means fewer borrowers reporting the emissions and transition data that lenders rely on.

Prudential expectations have hardened

European and UK supervisors have gone the other way. The ECB has used legally binding decisions and periodic penalty payments to enforce its climate expectations: in November 2025 it fined a Spanish bank €187,650 for 65 days of failing to assess the materiality of its climate and environmental risks⁵, and in February 2026 it imposed €7.55 million on a large French banking group for missing the same kind of deadline by 75 days⁶. The EBA's guidelines on managing ESG risks, including prudential plans with quantifiable targets and milestones to address the financial risks of transition, apply to larger EU banks from 11 January 2026 and to small and non-complex institutions by January 2027⁷.

In the UK, the PRA published an updated supervisory statement on managing climate-related risks in December 2025, replacing its 2019 expectations and covering governance, risk management, scenario analysis, data and disclosure; firms were given six months to review gaps and plan remediation⁸.

Transition finance: growing, but not at the pace needed

The commercial opportunity is real. BloombergNEF estimates banks arranged $2.3 trillion of energy supply financing in 2025, up 15%, including a five-year high of $1.15 trillion in low-carbon financing, lifted by grid investment. But fossil-fuel financing also rose, to $1.19 trillion, leaving the global energy supply banking ratio at 0.97:1, only slightly above 0.95:1 in 2024 and far from the average of about 4:1 BloombergNEF estimates is needed this decade⁹. The regional divergence is stark: Europe at 2.5:1, China at 1.6:1 and North America at 0.5:1⁹.

For banks, the lesson is that transition finance is shifting from a reputational commitment to a growth line judged on returns. Grid, storage, electrified industry and building retrofits require long-dated, structured financing that plays to bank strengths. The winners will be the banks that can assess transition risk in a borrower's plan as rigorously as its cash flows, and price accordingly.

The data gap is the binding constraint

Every prudential expectation, from materiality assessments to transition plans and scenario analysis, rests on counterparty-level data about emissions, transition plans and physical exposure. That data is patchy, and financed-emissions figures in particular are hard to compare. A Bank of England staff analysis found that four global systemically important banks of similar size and business model reported financed emissions ranging from 4 to 115 MtCO2e. Swapping emissions data providers moved estimates by only around 10%; widening the boundary from a minimal set of loans to the PCAF scope added about 50%, and extending it to all activities, including underwriting and asset management, roughly another 50%¹⁰.

The PCAF standard, updated in December 2025, requires banks to disclose a weighted data-quality score from 1 (verified, reported emissions) to 5 (sector-average proxies)¹¹. For most loan books outside large listed corporates, proxies dominate, and the narrowing of EU corporate reporting will make direct data scarcer, not richer.

• Define the boundary once. Agree group-wide which activities count, and disclose it; comparability starts with scope, not vendors.
• Collect at origination. Capture emissions, energy performance and transition-plan data in credit applications and annual reviews, not in year-end reporting sprints.
• Track data quality as a KPI. Move exposures up the PCAF data-quality ladder, prioritising high-emitting sectors and large exposures.
• Link climate to credit. Feed transition and physical risk into ratings, limits, pricing and ICAAP, as supervisors now expect.

The policy cycle will turn again; climate commitments have risen and fallen before. What will not change is that transition and physical risks affect collateral values, borrower cash flows and sector default rates. Banks that build the data and risk capabilities now, on prudential rather than reputational grounds, will be ready whichever way the politics move, and better placed to finance the transition profitably.