PRA Publishes Thematic Feedback on IFRS 9 Credit Loss Accounting, Flags Data Governance and Climate Risk Gaps at Major UK Banks

The Prudential Regulation Authority (PRA) published on 30 September 2026 a letter from Executive Director David Bailey to chief financial officers of selected PRA-regulated deposit-takers, setting out thematic feedback from its review of written auditor reports covering IFRS 9 (International Financial Reporting Standard 9) expected credit loss (ECL) accounting. ECL accounting requires banks to recognise likely future loan losses early, before they actually default, rather than waiting for defaults to occur. The PRA's review focused on three areas. First, data governance: while auditors broadly found effective controls, the maturity and consistency of firms' frameworks varied. Better-practice firms used sensitivity analysis to identify critical data elements and applied quantitative thresholds to assess potential materiality. Second, model risk: firms continue to redevelop models and improve monitoring tools, but implementation has been uneven, and the PRA wants to see more responsive identification of emerging risks and more complete post-model adjustments. Third, climate risks: the PRA acknowledged continued progress in firms' ability to assess climate-related credit risks in line with its supervisory expectations, but called for more granular assessments, stronger links between scenario analysis and ECL judgements, and better underlying data. The PRA noted that aggregate ECL coverage across firms and asset classes is at its lowest since before Covid. It stressed this does not in itself indicate under-provisioning, but said that economic and geopolitical uncertainty, combined with limited recent default experience for some portfolios, makes timely incorporation of new risks into ECL models especially important. In a forward-looking addition to this year's review, the PRA has also asked auditors to assess firms' processes for identifying and monitoring credit risks in , reflecting concerns about the complexity, leverage and interconnectedness of those risks.