PRA CP16/26 O-SII Buffer: Release in Stress, Rebuild Over Years
On 18 November 2026 the Prudential Regulation Authority published PRA CP16/26, proposing to write into its statement of policy SoP4/16 when it would lower the O-SII buffer, including to zero, in a shock, and how it would rebuild the buffer afterwards. The O-SII buffer is the UK’s systemic CET1 buffer for ring-fenced banks, large building societies and large deposit-takers without material trading activity. Responses are due by Thursday 18 February 2027.
The CP leaves normal-times calibration, the Financial Policy Committee (FPC) framework, the annual rate-setting process and the scope of firms unchanged. What changes is the stress case in a capital plan: the conditions for a release, an indicative period during which no rate increase would be expected, and a rebuild in stages with 12 months, in general, to meet each new rate.
Related reading: PRA Pillar 2A Review Phase 1: The 1 January 2027 Reset
CP16/26 dates and documents
- 7 July 2026: the PRA’s statement on buffer usability and releasability, and the FPC’s Financial Stability in Focus on the bank capital framework, announce an O-SII consultation for 2026 H2.
- 18 November 2026: CP16/26 published, with Appendix 1 containing the draft amendments to SoP4/16.
- By 15 December each year: the PRA expects to announce O-SII buffer rates. Under SoP4/16 paragraph 5.1, rates announced in December 2026 take effect from 1 January 2028, and the PRA may adapt that timeline.
- 18 February 2027: consultation closes. Responses go to CP16_26@bankofengland.co.uk.
- Final policy: the amended SoP would take effect on publication of the final policy, for which the CP gives no date.
What CP16/26 adds to SoP4/16, and what it leaves alone
The consultation edits one document. Appendix 1 inserts a new paragraph 4.7 into SoP4/16 and adds Annex 1, “Release and rebuild scenarios in a systemic stress”. In the July 2026 version of the SoP, paragraph 4.7 is marked deleted, so the new text fills an empty slot.
The power behind it already exists. Regulation 18 of the Capital Buffers and Macro-prudential Measures Regulations 2025 (SI 2025/653) lets the PRA, on supervisory judgment, set a rate that differs from the framework rate or set no rate at all for a relevant O-SII. Paragraph 4.6 of the SoP limits that judgment to exceptional cases, and its examples concern the methodology misfiring: a firm deliberately manipulating its systemic score, or a leverage exposure measure that has changed materially between year-end and rate-setting. The draft Annex 1 adds economic shocks to the exceptional circumstances.
Scope is narrower than the O-SII list, a point SoP4/16 makes itself. The buffer applies to groups containing ring-fenced bodies, to building societies holding more than GBP 35 billion in deposits and shares as the SoP defines them, and to other UK deposit-takers with more than GBP 35 billion in core deposits and trading assets below 10% of Tier 1 capital. An O-SII outside those three groups has no O-SII buffer to release. CP16/26 does not change those eligibility tests. CP13/26 proposes indexation of other fixed regulatory thresholds but expressly excludes thresholds within the FPC’s O-SII buffer framework; the separate consultation is covered in our note on PRA threshold indexation under CP13/26.
The two conditions behind an O-SII buffer release
Draft paragraph 4.7 sets a two-part expectation. The PRA expects to lower O-SII rates if it anticipates that firms face the prospect of losses that could lead them to restrict lending to defend capital positions by more than the economic outlook warrants, and that an FPC release of the countercyclical capital buffer (CCyB) alone would be insufficient to absorb those losses. The PRA would make that judgement in engagement with the FPC.
Annex 1 lists eight factors, including the shock’s nature, severity and expected duration, the probability of another shock, market functioning and firms’ capacity to absorb losses, and states that they do not constitute a mechanical test.
The annex also names cases where a release would be unlikely: a mild downturn, a stress the CCyB release is judged sufficient to absorb, or market stress in which O-SII buffer firms are not expected to incur losses large enough to cut lending. Shocks to non-bank financial institutions are only likely to justify a cut if they significantly threaten those firms’ ability to keep supplying credit.
The narrow point for credit risk teams sits in paragraph 1.5 of the annex. The PRA would distinguish defensive lending cuts from an appropriate tightening of lending conditions that reflects borrowers’ creditworthiness and the economic outlook. Paragraph 1.7 adds that a reduction would not represent an assessment that a firm had become less systemically important.
Where a released O-SII buffer shows up in the capital stack
The direct effect is on the combined buffer. A lower O-SII rate lowers the capital level at which automatic restrictions on distributions apply. The consultation paper then describes three linked items, and they carry different weights.
| Linked item | What CP16/26 says about a release | Strength of the statement |
|---|---|---|
| Additional leverage ratio buffer (ALRB) | Calibrated by reference to the O-SII buffer, so a reduction flows through automatically | Automatic (CP paragraph 3.16) |
| PRA buffer group risk add-on and leverage ratio group add-on | The PRA would consider corresponding changes, where applicable, to reflect release at group level | Considered, no commitment (CP paragraph 3.17) |
| O-SII hurdle rates in future Bank Capital Stress Tests | Intended to reflect prevailing O-SII rates, so the test does not require a faster rebuild than the authorities intend | Stated intention of the FPC and PRA (CP paragraph 3.18) |
All three statements sit in the consultation paper’s text. The draft SoP amendments in Appendix 1 do not repeat them, so they would not become part of SoP4/16 as drafted.
The leverage side behaves differently on distributions. The July 2026 Financial Stability in Focus (FSiF) states that, at present, the ALRB is applied at 35% of the equivalent risk-based buffer, and PS11/26 records that capital distribution constraints in the UK are not triggered by leverage ratio metrics. The distribution-threshold effect described in CP16/26 operates through the risk-weighted combined buffer. An O-SII rate reduction also lowers the ALRB automatically, but does not reduce the separate 3.25% minimum UK leverage ratio requirement. How the UK leverage framework compares with the Basel minimum is covered in our leverage ratio RCAP 2026 note.
The indicative period and the path back to normal-times rates
At release, the PRA would communicate the revised rates, the date from which they apply and an indicative period during which no increase would be expected. In setting that period it would take account of the nature and expected duration of the stress and relevant FPC actions, including any indicative period for the UK CCyB.
Treat the end of the indicative period as a review date. Annex 1 says the period “would not create an automatic timetable for rebuilding”, and the PRA could extend it if conditions do not support rebuilding, including where impairments emerge with a lag. At the end of the period the PRA would assess whether firms could start rebuilding while continuing to lend, looking among other things at whether firms incurred the losses anticipated at release and how much of the released capital they used.
Rebuilding would typically take several years. When the PRA raises O-SII rates in stages, firms would in general have 12 months before each stage’s rate must be met, and the PRA may set a shorter implementation period in exceptional circumstances. The annex sketches three paths: a relatively faster rebuild after a strong recovery with limited losses, a gradual rebuild that allows firms to meet higher rates from a share of retained earnings, and a longer period or extended indicative period after a weak recovery. Its Chart 1 is qualitative, with no values on the axes.
Reporting and capital planning before and after a release
The PRA states that the proposals would not introduce new reporting requirements, require changes to firms’ systems or require any action in normal times. No PRA rules change, and the proposals fall below the CBA Panel’s materiality threshold, so the Panel was not consulted.
The work sits in planning. Paragraph 2.10 of the CP says firms could apply the lower rates in their capital planning. My working assumption is that a stress scenario assuming a release needs four linked inputs: the lower distribution threshold, the automatic ALRB reduction, a group add-on assumption flagged as uncertain, and a stepped rebuild with 12-month lead times.
For Pillar 3 disclosure, PS11/26 amends the UK CC1 instructions from 1 January 2027, with a first reference date of 31 December 2026. Under the instructions for row 64, an institution should provide an accompanying qualitative narrative describing the applicable capital distribution constraints only where its CET1 ratio is below the CET1 overall capital requirement reported in that row. A lower O-SII buffer rate would reduce the combined-buffer component of that threshold, but a release would not, by itself, trigger the narrative. CP16/26 proposes no change to UK CC1.
CP16/26 inside the wider buffer reform
The July 2026 FSiF sets a long-term aim of a simpler framework centred on a single buffer that is releasable in stress and usable without automatic distribution restrictions, to be pursued with international authorities. The PRA calls CP16/26 an earlier domestic step towards that aim.
Leverage buffers move on a separate track. The same FSiF sets out an intention, subject to consultation, to bring the ALRB calibration for firms with systemic buffers into line with international standards and to make the ALRB for domestically systemic firms releasable in stress. CP16/26 does not implement either change.
For the EU comparison, our guide to macroprudential buffer stacking in the EU covers how CRD buffers combine, and our note on the EBA O-SII buffer opinion covers an EU case where combined systemic buffers passed 5 percent.
Frequently Asked Questions
If the FPC cuts the UK CCyB, does our O-SII rate fall with it?
The draft annex says a CCyB release would not automatically lead to an O-SII release. In its escalating-stress scenario, the PRA might first judge the CCyB release sufficient and reconsider if the stress intensified, weighing the marginal benefit of a further release against the temporary loss of structural resilience.
Could the PRA cut the rate part-way rather than to zero?
The CP refers to lowering rates, “including to zero”. Under regulation 18(3) of SI 2025/653, a rate the PRA sets by supervisory judgment must be 1%, 1.5%, 2%, 2.5% or 3%, and a full release corresponds to setting no rate. On my reading, a partial release would land on one of those steps; the CP does not discuss step sizes.
Our ring-fenced sub-group carries an O-SII buffer the consolidated group does not match. What happens at group level?
SoP4/16 already treats this as group risk, reflected in Pillar 2 group capital where the consolidated group has no G-SII buffer or a lower G-SII rate. On a release, CP16/26 commits only to considering corresponding changes to the PRA buffer group risk add-on and leverage ratio group add-on, with no automatic pass-through.
Related Articles
- PRA Pillar 2A Review Phase 1: The 1 January 2027 Reset: PS15/26 and the Pillar 2A and ICAAP changes from 1 January 2027.
- Macroprudential Buffer Stacking: How EU Buffers Combine in Practice: how the five CRD buffers combine and where COREP carries them.
- EBA O-SII Buffer Opinion: When Combined Systemic Buffers Exceed 5 Percent: an EU case of combined systemic buffers above 5 percent.
- Leverage Ratio RCAP 2026: Where Six Jurisdictions Depart from Basel: the Basel leverage ratio assessments, including the UK.
- PRA Threshold Indexation: How CP13/26 Moves 128 Thresholds: tying fixed Rulebook amounts to nominal GDP.
Sources and References
- PRA, CP16/26: Clarifying the PRA’s approach to the implementation of the O-SII buffer (18 November 2026): bankofengland.co.uk
- PRA, CP16/26 Appendix 1: Draft amendments to SoP4/16 (PDF): bankofengland.co.uk
- PRA, SoP4/16: The PRA’s approach to the implementation of the O-SII buffer (July 2026, PDF): bankofengland.co.uk
- The Capital Buffers and Macro-prudential Measures Regulations 2025 (SI 2025/653), Part 3: legislation.gov.uk
- PRA statement on enhancing the usability and releasability of capital buffers (7 July 2026): bankofengland.co.uk
- Bank of England, Financial Stability in Focus: The bank capital framework (7 July 2026): bankofengland.co.uk
- Bank of England, Financial Stability in Focus: The FPC’s assessment of bank capital requirements (December 2025, PDF): bankofengland.co.uk
- PRA, PS11/26: Disclosure: resolvability resources, capital distribution constraints and the basis for firm Pillar 3 disclosure (26 March 2026): bankofengland.co.uk
- PRA, PS11/26 Appendix 6: Instructions for own funds disclosure, UK CC1 row 64 (PDF, effective 1 January 2027): bankofengland.co.uk
- PRA, CP13/26 (7 October 2026), paragraph 3.17 on thresholds outside the proposed indexation framework: bankofengland.co.uk
Responding to CP16/26 before 18 February 2027
The points I would test in a response are practical: whether an indicative period of the kind described gives enough certainty to use released capital, whether 12 months per rebuild stage fits a retained-earnings plan, and whether the group add-on and stress-test hurdle statements belong in the SoP text. Responses go to CP16_26@bankofengland.co.uk by 18 February 2027, stating whether the PRA may publish the respondent’s name.
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.
