It is read with SoP5/25, the PRA’s statement of policy on Pillar 2 methods for SDDTs (SS4/25 paragraph 1.6), and both were last updated in July 2026. Until 1 January 2027 an SDDT works to SS31/15 for everything except the update cycle (PS4/26 paragraph 1.16).
How does SS4/25 differ from SS31/15?
It keeps the shape and much of the wording of SS31/15, and changes the update cycle, the stress benchmark, the buffer, the Pillar 2A methods and the document. The PRA published a marked-up comparison of its draft with CP7/24 in September 2024; the draft has moved since, so the table compares the July 2026 texts.
| Topic | SS31/15 (firms outside the regime) | SS4/25 (SDDTs) |
|---|---|---|
| ICAAP document | Documented and updated annually, or more often if the business changes (2.1) | Updated in full at least every two years; annually for new and growing banks; sooner after a material change (2.2, 2.3) |
| Stress tests and scenario analysis | At least annually (3.8) | At least every two years; annually for new and growing banks (3.8) |
| Reverse stress test | At least annually (ICAA rule 15.4) | At least every two years, and may be qualitative (ICAA rule 15.4; SS4/25 4.1, 4.3) |
| Benchmark scenario | Bank Capital Stress Test scenario or the PRA’s published scenario (3.16 to 3.19) | Two non-cyclical SDDT scenarios a year; the firm’s own at least as severe (3.16 to 3.18) |
| Pillar 2B | PRA buffer, alongside the combined buffer (5.20, 5.21) | Single Capital Buffer of no less than 3.5% of RWAs before any governance scalar, met in CET1 (5.18, 5.20) |
| Pillar 2A methods | SoP5/15 (1.7) | SoP5/25 (1.6) |
| Liquidity risk | Assessed in the ICAAP (2.4) | Can draw on the ILAAP, and builds on it only where the ILAAP shows concerns (2.6, 2.7) |
| Document structure | No suggested structure | Optional structure in Annex 1 |
Source: PRA SS31/15 and SS4/25, July 2026 versions, paragraph numbers in brackets; PRA Rulebook, Internal Capital Adequacy Assessment Part, rule 15.4.
The last row is easy to skip. Annex 1 sets out an optional structure, from executive summary to reverse stress test, and lets an SDDT cross-refer to its ILAAP and recovery plan instead of repeating them. It also asks for the minutes of the committees that recommended or approved the ICAAP, with a “detailed description of challenges raised”. The PRA calls the structure “entirely optional”, but it shows what a supervisor expects to find.
How often must an SDDT update its ICAAP?
In full at least every two years, or annually if the SDDT is a new and growing bank (SS4/25 paragraph 2.2). The cycle has applied since 20 January 2026, together with the amended rule that an SDDT’s reverse stress test is documented, reviewed and approved at least every two years (PS4/26 paragraphs 1.15 and 1.16; ICAA rule 15.4).
Two years is a minimum. Paragraph 2.3 expects an earlier full update after a material change, naming material balance sheet growth, a change in business model such as a move to higher loan-to-value lending, and changes in market conditions or interest rates. The PRA can also ask for an annual update where an ICAAP is of poor quality. Stress tests follow the same two-year minimum (paragraph 3.8).
The PRA will not tie its review to the firm’s cycle. It declined to align ICAAP cycles with the C-SREP and said C-SREPs “can be conducted based on prior year ICAAP documents” (PS20/25 paragraph 5.12). The document the supervisor reads may be well over a year old, so it has to stand up on its own.
How is Pillar 2A set under SoP5/25?
The PRA sets it using its own SDDT methods in SoP5/25, with the firm’s ICAAP as an input where it judges that appropriate (SS4/25 paragraph 5.10). SoP5/25 replaces SoP5/15 for SDDTs (SoP5/25 paragraph 1.1), and four points shape the ICAAP.
- Credit risk. In most cases the PRA does not expect extra Pillar 2A capital for credit risk (SoP5/25 paragraph 2.4). A detailed assessment is expected from new and growing banks, mainly unsecured retail lenders and firms with bespoke or non-standard lending, such as sub-prime, shared equity or lifetime mortgages (SS4/25 paragraphs 2.10 and 2.11). The method is the firm’s own credit scenarios over 12 months, more severe than the PRA’s SDDT scenarios (paragraph 2.15).
- Credit concentration. A wholesale add-on of 3.5% of relevant RWAs, covering single-name, sector and geographic concentration, and a retail add-on of 1% of relevant RWAs for geographic concentration (SoP5/25 paragraphs 4.3 and 4.4). The PRA engages with any SDDT whose large exposures add up to more than 200% of Tier 1 capital (paragraph 4.8).
- Operational risk. The PRA considers an add-on for every SDDT (paragraph 3.4), using the method for firms outside the regime since PS15/26 aligned the two (PS15/26 paragraph 3.30).
- Quality of capital. Pillar 2A is met with at least 56.25% CET1, no more than 43.75% additional Tier 1 and no more than 25% Tier 2 (SS4/25 paragraph 5.13).
The wholesale add-on can be reproduced before the PRA sends it. It is built from the firm’s own Pillar 1 reporting (PS20/25 paragraph 3.44), and SoP5/25 paragraph 4.5 names the cells: row 0040 of template SC 02.00 less ten others. Our article on how the PRA sets Pillar 2A covers the method for firms outside the regime.
How does the Single Capital Buffer enter the capital plan?
As a floor the plan has to clear in the base case, sized above that floor by the firm’s own stress test. The PRA sets the SCB at no lower than 3.5% of RWAs before any risk management and governance scalar, “even where firm-specific stress testing results suggest a buffer lower than 3.5%” (SoP5/25 paragraph 12.11).
Above the floor, the number comes from the ICAAP: the fall in the firm’s CET1 surplus over its total capital requirement under the stress, after the PRA has tested the projections and management actions against its own models and peers (paragraphs 12.8 to 12.10). The PRA’s worked example shows the arithmetic. A stress that uses £136 million of CET1 surplus is 3% of RWAs of £4,533 million, so the SCB is set at the 3.5% floor; the same £136 million against RWAs of £3,400 million is 4%, and the SCB is 4% (paragraphs 12.22 and 12.23).
For the capital plan, four expectations follow.
- It is met with 100% CET1, none of it CET1 already held for total capital requirement (SS4/25 paragraph 5.20).
- The base case should meet it; if not, expect a request for a revised capital plan (SoP5/25 paragraph 12.31).
- It is for use in a stress, not in the normal course of business or the base plan (paragraph 12.43).
- A firm that expects to draw on it must tell the PRA early, with a capital restoration plan, and faces no automatic restrictions on distributions (paragraphs 12.44 and 12.46).
Significant risk management or governance weaknesses can add a scalar of up to 40% of the CET1 needed for variable total capital requirement; in the PRA’s example a 20% scalar adds 1.125% of RWAs (paragraphs 12.25 and 12.26). The SCB is reassessed with the SREP, ordinarily every two to four years (paragraph 12.6). The capital plan should show the SCB as a range: the floor, the firm’s own stress result and, where weaknesses are still open, the effect of a scalar. Our article on the PRA buffer sets out what it replaces.
What does stress testing look like for an SDDT?
Firm-run, against a benchmark that does not move with the cycle. The PRA publishes two SDDT scenarios each year, and an SDDT’s own scenarios for testing whether it keeps minimum capital should be at least as severe (SS4/25 paragraphs 3.16 and 3.18).
The 2026 scenarios were published on 14 July 2026, alongside the ICAAP scenarios for firms outside the regime, and the Bank has issued supplementary guidance on the difference. The SDDT scenarios leave out the countercyclical adjustments in the ICAAP scenarios, so the buffer they produce stays broadly constant through the cycle. In 2026 that makes them harder on unemployment and commercial property.
| Variable | SDDT scenarios, July 2026 | ICAAP scenarios, July 2026 |
|---|---|---|
| Unemployment | +4.5 percentage points | +3.3 percentage points |
| Real GDP | -5% | -5% |
| House prices | -28% | -28% |
| Commercial real estate prices | -45% | -35% |
Source: Bank of England, supplementary stress testing guidance for SDDT capital regime ICAAPs (2026), table of shock sizes in the July 2026 scenarios.
The timing matters. The Bank says the SDDT scenarios are for ICAAPs prepared under the SDDT capital regime, and that for ICAAPs “signed off before 01 January 2027, all firms should use the ICAAP scenarios”. An SDDT approving its ICAAP this autumn runs the ICAAP scenarios; its first ICAAP on the SDDT scenarios is the next one signed after 1 January 2027. The guidance also says the PRA may adjust an SDDT’s SCB for the effect of IFRS 9, still no lower than 3.5% of RWAs.
The rest is familiar. Capital is projected over three to five years, before and after management actions the firm could and would take (SS4/25 paragraphs 3.9, 3.23 and 3.24), and SDDTs with total assets of £5 billion or more submit the PRA111 data item with the ICAAP (SoP5/25 paragraph 12.48). The reverse stress test may be qualitative (SS4/25 paragraph 4.3), but failure still means the point at which the market loses confidence, which “may be reached well before the firm’s financial resources are exhausted” (paragraph 4.2).
What should the board see in the off year?
A short paper showing that the ICAAP still holds, and a minute recording that decision. The overall financial adequacy rule applies at all times and the Pillar 2 assessments must be carried out regularly (ICAA rules 2.1 and 3.4), and SS4/25 expects resource adequacy to be reported to the management body “as often as is necessary” (paragraph 2.4).
In the year between full updates, the board pack needs four things:
- actual and projected capital against total capital requirement plus SCB, with the distance to the 3.5% floor;
- a check against each trigger for an early update in paragraph 2.3, from balance sheet growth to moves in rates;
- the SDDT scenarios published that July, set against the ones the last ICAAP used, with a view on whether the firm’s own stress would still be at least as severe;
- the calendar for the other documents: since 1 April 2026 the recovery plan is reviewed at least every two years, kept in step with the ICAAP and ILAAP (Recovery Plans Part rule 4.2; PS10/26 paragraph 2.9), while the solvent exit analysis runs at least every three years (rule 7.3).
If any of these points to a material change, the answer is a full update. Our guide to recovery plan indicators covers what the SCB changes in the recovery plan, our SDDT regime page and strong and simple guide cover the regime, and the SDDT eligibility checker tests the nine criteria. Our ICAAP service writes or reviews an ICAAP to SS4/25, with the RisKIT capital models running the projections in open Excel.
Questions readers ask
Does SS4/25 apply before 1 January 2027?
Only in part. Its expectations on how often an SDDT updates its ICAAP and reverse stress test have applied since 20 January 2026; the rest of SS4/25 takes effect on 1 January 2027, and until then SDDTs work to SS31/15 (PS4/26 paragraphs 1.15 and 1.16).
Can an SDDT’s reverse stress test be qualitative?
Yes. SS4/25 paragraph 4.3 allows a more qualitative reverse stress test that identifies the risks or events that could make the business plan unviable, assesses remedial actions and identifies mitigants. The PRA may still ask the SDDT to quantify the financial position that would mean business failure (paragraph 4.4).
Is the Single Capital Buffer disclosed?
Not publicly. SDDTs give the PRA’s SREP feedback letter, which sets the SCB, to their auditors, and publicly disclose the amount and quality of their total capital requirement at the highest level of UK consolidation. The PRA expects the SCB itself to be treated as confidential unless the law requires disclosure (SS4/25 paragraph 5.26).
Sources: PRA SS4/25 The ICAAP and the SREP for SDDTs (July 2026 version, effective 1 January 2027), paragraphs 1.6, 2.2 to 2.7, 2.10, 2.11, 2.15, 3.8, 3.9, 3.16, 3.18, 3.23, 3.24, 4.1 to 4.4, 5.10, 5.13, 5.18, 5.20 and 5.26 and Annex 1; PRA SoP5/25 Pillar 2 methodologies for SDDTs (July 2026 version), paragraphs 1.1, 2.4, 3.4, 4.3 to 4.5, 4.8, 12.6, 12.8 to 12.11, 12.22, 12.23, 12.25, 12.26, 12.31, 12.43, 12.44, 12.46 and 12.48; PRA SS31/15 The ICAAP and the SREP (July 2026 version), paragraphs 1.7, 2.1, 2.4, 3.8, 3.16 to 3.19, 5.20 and 5.21; PRA CP7/24 comparison of the draft SDDT statement with SS31/15 (September 2024); PRA Rulebook, Internal Capital Adequacy Assessment Part, rules 2.1, 3.4 and 15.4; PRA Rulebook, Recovery Plans Part, rules 4.2 and 7.3; PRA PS4/26 (20 January 2026), paragraphs 1.15 and 1.16; PRA PS20/25 (28 October 2025), paragraphs 3.44 and 5.12; PRA PS15/26 (28 May 2026), paragraph 3.30; PRA PS10/26 (26 March 2026), paragraph 2.9; Bank of England, Stress testing non-systemic banks and building societies (updated 15 July 2026) and supplementary stress testing guidance (2026). Accessed 10 October 2026.
How we research and check our articles: our editorial policy.
