PRA Pillar 2A Review Phase 1: The 1 January 2027 Reset

On 28 May 2026 the Prudential Regulation Authority closed the first phase of its PRA Pillar 2A review with policy statement PS15/26. The changes it confirms come into force on Friday 1 January 2027, the same day the PRA switches on the Basel 3.1 standards, and they land in the part of the capital stack that firms build themselves: the Pillar 2A add-on and the Internal Capital Adequacy Assessment Process (ICAAP) that supports it.

If your project plan still has this dated to the middle of 2026, it is working from an earlier draft. Two elements of the original consultation, pension risk and market risk, were once pencilled in for a 2026 start. The final policy pulls every piece of Phase 1 onto one date. Get that wrong and a capital team either rebuilds its add-on models too early against rules that later moved, or signs off a 2027 ICAAP on the old benchmarking approach that no longer exists.

This is a change note for capital and prudential reporting teams at UK banks, building societies and PRA-designated investment firms. It walks through what PS15/26 settles across credit, operational, pension and market risk, the two new systematic credit-risk add-ons, the reporting items that go, and the ICAAP sign-off timing that decides which version of the rules your next assessment has to reflect.

Related reading: our guide to the ICAAP and ILAAP.

Where PS15/26 sits, and the date that actually matters

Pillar 2A capital exists to cover risks that Pillar 1 either misses or measures poorly. Interest rate risk in the banking book, credit concentration risk and pension obligation risk are captured nowhere in Pillar 1; credit, market and operational risk are captured, but not always well enough. The PRA sets a firm-specific add-on for those gaps, and the firm must undertake its own independent bottom-up assessment in its ICAAP. SS31/15 is explicit that merely replicating the PRA’s methodologies does not constitute a firm-owned assessment. The methodology behind the PRA side of that exercise is set out in statement of policy SoP5/15, and the supervisory expectations for the ICAAP and the Supervisory Review and Evaluation Process (SREP) sit in supervisory statement SS31/15.

PS15/26 is the feedback statement and final policy responding to consultation paper CP12/25. It updates SoP5/15, updates SS31/15, amends the Reporting Pillar 2 Part of the PRA Rulebook, and reissues the small-firm equivalents SoP5/25 and SS4/25 for Small Domestic Deposit Takers (SDDTs), plus the Pillar 2 reporting statement SS32/15, the reporting schedule and the FSA data items. The PRA received 20 responses to CP12/25, most of them on the credit risk proposals, and made a series of clarifying changes without abandoning the core design.

The single most important fact for planning is the commencement date. The amended Rulebook Part, the reporting templates and instructions, the schedule, and the updated SSs and SoPs all come into force on Friday 1 January 2027. The PRA aligned Phase 1 with its implementation of the Basel 3.1 standards deliberately, because much of the credit-risk change only makes sense once the revised standardised approach is live. For the mechanics of that wider Basel 3.1 go-live on the trading-book side, see our note on the PRA’s Basel 3.1 market risk and IMA adjustments.

The PRA Pillar 2A review Phase 1 calendar

Phase 1 has a short set of operative dates. Pin these into the capital plan before touching a model.

• 28 May 2026: PS15/26 published, confirming the final Phase 1 policy and the responses to CP12/25.
• Earlier proposal, now superseded: pension risk, market risk and counterparty credit risk changes were first proposed for 2 March 2026, then moved to 1 July 2026 when the consultation ran longer. Neither date survives.
• Friday 1 January 2027: single implementation date for every element of Phase 1. The amended Rulebook Part, SoP5/15, SoP5/25, SS31/15, SS4/25, SS32/15, the reporting schedule and the amended Pillar 2 data items (FSA072 to FSA076 and FSA081, revised by PS15/26) all come into force.
• 1 January 2027: the refined methodology to Pillar 2A is retired, as confirmed separately in PS2/26, so its associated reporting stops.
• During 2027: the PRA has said it will publish a further consultation paper opening Phase 2, a deeper review of individual Pillar 2A methodologies.

The ICAAP timing is a second calendar layered on top, and it catches teams out. The PRA confirmed in January 2026 that ICAAPs signed off by boards during 2026 should include an impact assessment of the Basel 3.1 standards. ICAAPs signed off by boards from 1 January 2027 should be prepared on a Basel 3.1 basis and reflect the final policy changes in PS15/26. So the assessment your board approves in early 2027 is the first one that has to be built entirely on the new footing.

Who the policy applies to

Scope here is specific, and it is worth stating precisely instead of reaching for “all firms”. PS15/26 is relevant to PRA-regulated banks, building societies, PRA-designated investment firms and all PRA-approved or PRA-designated holding companies, including those inside the SDDT regime. It is not relevant to credit unions.

The SDDT split matters for the credit-risk chapter in particular. The two new systematic credit-risk methodologies apply to banks, building societies, designated investment firms and their holding companies, and do not apply to firms that have opted into the SDDT regime, aside from one consequential Basel 3.1 update on the eligibility of guarantees and credit derivatives as credit risk mitigation. The operational risk changes, by contrast, were deliberately aligned across SDDT and non-SDDT policy materials so the methodologies read consistently. If your firm is an SDDT, read the credit-risk sections below as background and work from SoP5/25 and SS4/25 for your own obligations.

Credit risk: the benchmarking methodology is being removed

The headline credit-risk change is the removal of the benchmarking methodology, including the internal ratings based (IRB) benchmarks, from SoP5/15. That approach compared a firm’s standardised approach (SA) risk weights against IRB reference points to flag possible underestimation. The PRA’s view is that once Basel 3.1 sharpens the risk sensitivity of the SA, IRB risk weights stop being a reliable comparator, and maintaining the benchmarks becomes both less meaningful and harder to sustain.

Eight respondents pushed back, saying the benchmarks were useful inputs to their own ICAAP assessments and worrying that gaps between IRB and SA risk weights could persist after Basel 3.1 and raise competition concerns. The PRA held its position and will not publish replacement benchmarks as a reference tool, on the basis that they would add reporting and operational burden for limited benefit. One practical consequence follows immediately: any ICAAP credit-risk narrative that leans on the IRB benchmark comparison needs a new evidential spine for the 2027 assessment. The benchmark tables are being retired.

Two systematic credit-risk add-ons replace the benchmarks

In place of the benchmarks, the PRA introduces two systematic methodologies that feed the Pillar 2A credit-risk add-on where Pillar 1 SA risk weights understate the risk. Both are Pillar 2A tools. Both sit on top of the Pillar 1 SA risk weights, applying a charge where the SA outcome is judged too low.

The first targets relevant exposures to non-UK central governments and central banks (CG/CBs) and to regional governments and local authorities (RG/LAs). For the framework scheduled to apply from 1 January 2027, the July 2026 PRA materials use the PRA Rulebook together with regulations 6(1) and 6(3) of the Overseas Prudential Requirements Regime statutory instrument, rather than CRR Articles 114(7) and 115(4). The PRA made those rules on the understanding that the statutory instrument would be made and in force before 1 January 2027. The PRA sets minimum effective risk weights by credit quality step (CQS), and calculates a systematic add-on for the shortfall. UK central government, UK central bank and UK devolved administration exposures are outside this methodology. The minimum effective risk weights confirmed in Phase 1 are:

Exposure	CQS 1 / MEIP 0 to 1	CQS 2 to 3 / MEIP 2 to 3	CQS 4 to 6 / MEIP 4 to 7 / unrated
CG/CBs (excluding the UK)	No minimum effective risk weight	5%	20%
RG/LAs (excluding UK devolved administrations)	5%	20%	100%

RG/LA minimum effective risk weights apply on the basis of CQS only. The PRA kept the 5% floor for CQS2/3 CG/CB exposures despite requests to halve it to 2.5%, drawing on external default data over horizons longer than one year and on the Bank of England’s December 2025 Financial Stability in Focus assessment of bank capital requirements. It made one scope change from the consultation: exposures secured by collateral recognised through the financial collateral simple method are now excluded, because combining the Pillar 1 collateral haircut with a Pillar 2A add-on would be too conservative. An RG/LA exposure carrying an eligible CQS1 central government or central bank guarantee does not attract an additional requirement under the systematic approach.

The second methodology addresses retail unconditionally cancellable commitments (UCCs). The Pillar 1 SA applies a 10% conversion factor to retail UCCs, which the PRA’s analysis found sits well below the equivalent for IRB firms, leaving the exposure undercapitalised. The systematic methodology sets a conversion-factor reference point of 20%, and produces an add-on where a firm’s conversion factor is below that reference. A firm may substitute its own conversion factor in place of the 20% reference where the PRA considers it robustly substantiated, but that factor is floored at the Pillar 1 conversion factor of 10%. Here the PRA made a notable carve-out. Its calibration data covered UCCs to individuals but not to retail SMEs, and it accepted that the methodology may not fit SME facilities, so it removed SMEs from scope entirely. A common misreading is already circulating that every retail commitment is caught. Retail UCCs to SMEs are not in the systematic methodology, though the PRA has said it may revisit this if evidence of systematic undercapitalisation emerges.

For both methodologies the arithmetic that turns a risk-weight or conversion-factor gap into capital is worth stating plainly, because the PRA clarified it in response to a query. The shortfall is expressed as an underestimation of risk-weighted assets, and that figure is multiplied by 8% to give the capital underestimation that the add-on covers. The systematic outputs, where they apply, form the baseline of a firm’s Pillar 2A credit-risk add-on. They cannot be reduced through the idiosyncratic assessment. A firm’s Pillar 2A credit-risk requirement is the systematic components plus any additional idiosyncratic amount. The UK standardised-approach credit-risk changes that underpin these add-ons are set out in PS1/26 (Implementation of Basel 3.1: Final rules, published 20 January 2026) and the PRA Rulebook. For the parallel EU framework, see our note on CRR3 standardised credit risk and ECAI due diligence, which governs EU institutions under a separate regime.

Idiosyncratic credit risk moves from prescribed scenarios to a firm-owned choice

CP12/25 proposed to require firms to run credit scenarios of a prescribed severity to assess idiosyncratic credit risk across their SA book. Sixteen respondents replied, and the message was consistent: credit scenarios are resource-heavy, applying them to an entire portfolio is disproportionate, and the guidance was too high level on severity and design. Some argued the whole point of Pillar 2 is that firms own their capital adequacy judgement.

The PRA agreed and changed the policy. Under the final SS31/15, firms have flexibility to choose their method, whether credit scenarios, proxy IRB approaches or another technique, provided it is sound and proportionate and captures high-severity tail events over a 12-month horizon. Detailed assessment is expected only for the smaller subset of SA exposures more likely to carry idiosyncratic risk that Pillar 1, and any systematic add-on, does not already capture. The PRA gives examples of lending that may warrant closer work, such as niche portfolios, without turning the list into a closed set. Firms can still net overcapitalisation against undercapitalisation across credit portfolios where each has had a detailed assessment, but not by discounting the systematic baseline.

This is a genuine softening, and it is easy to over-read in the other direction. That flexibility is genuine but bounded. A firm that concludes Pillar 1 is sufficient for a portfolio still has to be able to show the assessment behind that conclusion, and where it uses a proxy IRB approach it has to give the PRA enough detail to follow the modelling methodology and assumptions. For a separate EU framework, see our note on the EBA’s revised SREP guidelines on ICAAP, ILAAP and Pillar 2 capital.

Operational risk: AMA disappears, the Pillar 2A mechanic stays

Basel 3.1 removes AMA permissions from the Pillar 1 operational-risk framework. PS15/26 updates the Pillar 2A operational-risk framework by removing obsolete AMA-specific wording, adding transparency and guidance, and onboarding selected former AMA requirements and guidelines as good practices; the underlying methodology is unchanged. FSA072 to FSA075 remain mandatory for significant firms and for any non-significant firm that had PRA permission to use AMA as at 31 December 2026, unless the required data have already been reported to the PRA by other means. The instructions are clarified for the removal of AMA permissions and the interaction between Pillar 1 and Pillar 2 reporting.

The detail that catches teams is the loss-data basis. Pillar 2 operational-risk reporting is built on gross losses net of direct recoveries, and it excludes insurance recoveries. Pillar 1 under Basel 3.1 includes insurance recoveries. Loss data prepared for the ICAAP therefore has to be assembled on the Pillar 2 basis to match the templates, and it cannot be lifted from the Pillar 1 figures. The PRA also confirmed that boundary credit-related operational-risk events should be excluded from FSA072 and FSA073, and that reporting should map to the Basel event types in Annex 2 of the Operational Risk Part of the Rulebook, using internal taxonomies only where they map cleanly.

On the capital number itself, firms asked whether Pillar 2A operational-risk requirements would fall as Pillar 1 rises under the new standardised approach. The PRA repeated the position from PS17/23: it will mechanically adjust firms’ Pillar 2A operational-risk requirements in line with Basel 3.1 changes to Pillar 1 risk-weighted assets, so that total nominal operational-risk capital stays broadly unchanged for most firms. That off-cycle review of firm-specific requirements is already under way, and it happens through supervisory channels and the firm does not calculate it directly.

Pension and market risk: lighter assessment, one harmonised date

Pension obligation risk is the clearest simplification in the package. The PRA removes its prescribed stress scenarios and lightens the Pillar 2A pension-risk assessment and its reporting for certain schemes. Where a scheme is fully bought-in or has a funding ratio of at least 130% on the firm’s accounting basis, a firm need not provide a full submission, with a minor modification to the FSA081 data item after respondents flagged how residual surplus and insured annuities interact with the funding-ratio definition. Firms still have to hold the arrangements to produce a full dataset if the position changes.

The PRA implemented the additional disclosures concerning its market-risk and counterparty-credit-risk methodologies broadly as proposed, with minor clarifications. It separately reconfirmed that market-risk add-ons and CVA-risk add-ons would be adjusted through the off-cycle review to address improved Pillar 1 capture; it did not state that all counterparty-credit-risk Pillar 2A requirements would be reduced. The bigger point for pension and market risk is timing. Both were once scheduled for a 2026 start, and both now move to 1 January 2027 so the whole of Phase 1 lands together. That harmonisation is the reason the “September 2026” framing that attaches to this topic does not match the final policy.

Reporting: FSA076 slims, FSA077 and FSA082 go

The reporting changes track the methodology changes. Removing the benchmarking approach lets the PRA simplify the FSA076 data item and decommission FSA077 and FSA082, which five respondents supported as a genuine reduction in burden. The refined methodology reporting also stops from 1 January 2027, in line with PS2/26. The net effect is fewer data points, but the items that remain change shape.

Two FSA076 clarifications are worth flagging for the mapping exercise. The July 2026 future FSA076 instructions identify relevant central-government and central-bank exposures by reference to regulation 6(1) of the OPRR statutory instrument and relevant regional-government and local-authority exposures by reference to regulation 6(3). They also make clear that retail SME commitments sit outside the UCC systematic add-on. It also confirmed, across SS31/15, SoP5/15 and the FSA076 instructions, that all banking-book and trading-book exposures assigned a risk weight under the credit-risk SA are within scope of the systematic methodologies. Teams that read the CP and assumed trading-book SA exposures were out will need to correct that assumption before building the return.

What to change in the ICAAP now

The work splits into three streams. First, the credit-risk methodology: retire the benchmark comparison from the ICAAP credit-risk section, build the two systematic add-ons for non-UK sovereign, sub-sovereign and retail UCC exposures, and decide the firm-owned approach to idiosyncratic risk for the exposures that need detailed assessment. Second, the loss and data plumbing: separate the Pillar 2 operational-risk loss basis from the Pillar 1 figures, and re-map FSA076 and FSA072 to FSA075 to the amended instructions. Third, the governance calendar: make sure the ICAAP the board signs off from 1 January 2027 is prepared on a Basel 3.1 basis and reflects PS15/26, while a 2026 sign-off still carries a Basel 3.1 impact assessment.

One first-hand observation from running these mapping exercises: the change that generates the most rework is never the new formula, it is the scope wording. A single renamed FSA076 item, like the retail SME UCC carve-out, quietly re-cuts which exposures feed the add-on, and a model built to the consultation text rather than the final instructions produces a number that reconciles to nothing. Read the final SoP5/15 and SS31/15 wording before writing the specification, not after.

Frequently Asked Questions

Does the PRA Pillar 2A review Phase 1 take effect in September 2026?

No. The final policy and rules in PS15/26 come into force on Friday 1 January 2027. Some elements were once proposed for a 2026 start, including a 1 July 2026 date for pension and market risk, but the PRA harmonised the whole of Phase 1 onto the January 2027 date to align with the Basel 3.1 standards.

Do the new minimum effective risk weights apply to UK gilts and UK local authority exposures?

UK central government and Bank of England exposures are excluded. UK devolved administration exposures are also excluded from the regional-government and local-authority methodology. Other UK local-authority exposures are not categorically excluded and must be assessed against the applicable SoP5/15 and FSA076 criteria.

Are retail SME commitments caught by the retail UCC add-on?

No. The PRA removed SMEs from the scope of the systematic methodology for retail UCCs, because its calibration data covered UCCs to individuals rather than SMEs. The FSA076 UCC table is being renamed to make the carve-out explicit. The PRA has said it may revisit this if evidence emerges that retail SME UCCs are systematically undercapitalised.

Can a firm reduce the systematic add-on through its idiosyncratic assessment?

No. Where the systematic methodologies apply, their output is the baseline of the Pillar 2A credit-risk add-on and cannot be netted down through the idiosyncratic assessment. Netting of overcapitalisation against undercapitalisation is still allowed across credit portfolios that have each had a detailed assessment, but the systematic baseline stands.

Does removing the AMA increase Pillar 2A operational-risk capital?

Not by design. The PRA will mechanically adjust Pillar 2A operational-risk requirements in line with the Basel 3.1 changes to Pillar 1 risk-weighted assets, so total nominal operational-risk capital stays broadly unchanged for most firms, consistent with PS17/23. The firm-specific off-cycle review handling this is under way through supervisory channels.

Which ICAAP has to be built on the new basis first?

ICAAPs signed off by boards from 1 January 2027 should be prepared on a Basel 3.1 basis and reflect the PS15/26 changes. ICAAPs signed off during 2026 should include a Basel 3.1 impact assessment. The first fully new-basis assessment is therefore the one a board approves in early 2027.

What happens to the benchmarking and refined-methodology reporting?

FSA076 is simplified and FSA077 and FSA082 are decommissioned from 1 January 2027. The refined methodology to Pillar 2A is retired on the same date under PS2/26, so its associated reporting stops. The items revised by PS15/26 are FSA072 to FSA076 and FSA081. The full Pillar 2 reporting schedule for non-SDDTs is broader: it also covers FSA071 (firm information and Pillar 2 summary), concentration-risk items FSA078 and FSA079, market-risk item FSA080, and PRA111 for firms with assets of at least £5 billion at the relevant level of consolidation used as the basis of their ICAAP. SDDTs and SDDT consolidation entities submit PRA119; they submit FSA081 only where they have defined benefit pension schemes, and PRA111 only where they meet the same £5 billion asset threshold. FSA077 and FSA082 are decommissioned by PS15/26.

When does Phase 2 arrive?

The PRA has said it will publish a further consultation paper during 2027 to open Phase 2, a more detailed review of individual Pillar 2A methodologies. Phase 2 is separate from the strong and simple work for SDDTs and from the retirement of the refined methodology, both of which have their own policy statements.

Related Articles

• PRA Basel 3.1 Market Risk and IMA Adjustments – How the PRA’s Basel 3.1 market-risk framework starts on 1 January 2027, while the internal model approach for market risk is the subject of further proposals in CP9/26.
• ICAAP and ILAAP Explained – What the internal capital and liquidity adequacy assessments cover and how supervisors use them.
• EBA Revised SREP Guidelines: ICAAP, ILAAP and Pillar 2 Capital – The EU treatment of Pillar 2 capital, a useful contrast for dual-reporting groups.
• CRR3 Standardised Credit Risk and ECAI Due Diligence – The EU standardised-approach credit-risk overhaul under CRR3: a separate EU regime from the UK Basel 3.1 changes in PS1/26, relevant to dual-reporting groups.
• CRR3 Output Floor Phase-In – How the Basel 3.1 output floor interacts with the rest of the capital stack.

Key Takeaways

• PS15/26, published 28 May 2026, finalises Phase 1 of the PRA Pillar 2A review and sets a single implementation date of Friday 1 January 2027, aligned with the Basel 3.1 standards.
• The IRB benchmarking methodology is removed from SoP5/15 and will not be replaced with reference benchmarks.
• Two systematic Pillar 2A credit-risk add-ons take its place: minimum effective risk weights for non-UK CG/CB and RG/LA exposures, and a conversion-factor reference point of 20% for retail UCCs (floored at the Pillar 1 10%), with retail SMEs carved out.
• Systematic add-on outputs are the baseline of the Pillar 2A credit-risk requirement and cannot be reduced by the idiosyncratic assessment; the shortfall in risk-weighted assets is converted to capital at 8%.
• Idiosyncratic credit risk moves from prescribed credit scenarios to a firm-owned choice of method, with detailed assessment expected only for higher-risk SA exposures.
• Basel 3.1 removes AMA from Pillar 1, but PS15/26 does not remove the Pillar 2A operational-risk methodology. The PRA is separately adjusting firm-specific Pillar 2A operational-risk requirements through its off-cycle review so that total nominal Pillar 1 plus Pillar 2A operational-risk capital remains broadly unchanged for most firms; Pillar 2 loss data excludes insurance recoveries.
• FSA076 is simplified, FSA077 and FSA082 are decommissioned, and ICAAPs signed off from 1 January 2027 must be prepared on a Basel 3.1 basis reflecting PS15/26.

Sources and References

• Prudential Regulation Authority, PS15/26 – Pillar 2A review – Phase 1, published 28 May 2026: https://www.bankofengland.co.uk/prudential-regulation/publication/2026/may/pillar-2a-review-phase-1-policy-statement (including the updated SoP5/15, SS31/15, SoP5/25, SS4/25, SS32/15, the reporting schedule and FSA data items in the appendices).
• PRA, CP12/25 – Pillar 2A review – Phase 1 (the consultation to which PS15/26 responds).
• Bank of England, Financial Stability in Focus: The FPC’s assessment of bank capital requirements, December 2025: https://www.bankofengland.co.uk/-/media/boe/files/financial-stability-in-focus/2025/fsif-the-fpcs-assessment-of-bank-capital-requirements.pdf.
• PRA, PS1/26 – Implementation of Basel 3.1: Final rules (published 20 January 2026).
• PRA, PS17/23 – Implementation of the Basel 3.1 standards near-final part 1 (mechanical adjustment of Pillar 2A operational risk).
• PRA, PS9/24 – Implementation of the Basel 3.1 standards near-final part 2 (market and counterparty credit risk Pillar 2A treatment).
• PRA, PS2/26 – Retiring the refined methodology to Pillar 2A – final.
• PRA, PS4/26 – The Strong and Simple Framework: the simplified capital regime for SDDTs – final.

Building the 2027 ICAAP against final text, not the consultation

Phase 1 is a rebuild of the firm-owned parts of the capital stack, timed to arrive with Basel 3.1 on 1 January 2027. The methodology moves are real but bounded: benchmarks out, two systematic add-ons in, prescribed credit scenarios replaced by a firm-chosen assessment, AMA permissions removed from Pillar 1 by Basel 3.1 with the Pillar 2A operational-risk methodology unchanged and total nominal operational-risk capital held broadly stable through the off-cycle adjustment, and a lighter pension-risk assessment. The trap is drafting a specification from CP12/25 when SoP5/15 and SS31/15 have already moved on scope, on the SME carve-out, and on the single go-live date. Work from the July 2026 future versions of SoP5/15, SS31/15 and SS32/15 and the July 2026 FSA076 instructions, which incorporate PS15/26 and PS16/26 and are scheduled to apply from 1 January 2027; do not build from the May 2026 PS15/26 appendices alone.

Last updated: July 2026

Disclaimer: The information on RegReportingDesk.com is for educational and informational purposes only. It does not constitute legal, regulatory, tax, or compliance advice. Always consult your compliance officer, legal counsel, or the relevant supervisory authority for guidance specific to your institution.