1Explain Pillar 2: Pillar 2A (the P2R) as a binding requirement, and the PRA buffer and P2G as guidance above the buffers.
2Describe the SREP: its four elements, the 1 to 4 scores and what they lead to.
3Build a bank's overall capital stack and find its MDA line: 10.73% for RiskAtoms Bank plc, £573.5m below its CET1.
4Read the same bank in the UK and EU terms: TCR and PRA buffer against TSCR, OCR and P2G.
Why it matters
RiskAtoms Bank plc reports a CET1 ratio of 13.24%. Pillar 1 asks for 4.5%, so it looks three times safe. It is not. Its supervisor adds capital for risks the formulas miss, the buffers sit on top, and a further buffer is sized from a stress test. Once everything is stacked, the bank's dividends are restricted below 10.73%, and the supervisor would like to see more than that. The gap that matters is £573.5m, not 8.7 percentage points.
Illustrative dataRiskAtoms Bank plc under supervisory review
The same made-up UK bank as Levels 3 and 4: £67bn of assets at 31 December 2025, in £ millions. Its booked capital comes from Module 4.1 and stands: CET1 of £3,019.9m against risk-weighted assets of £22,807.7m (13.24%), Pillar 2A of 2.5% of RWA, the 2.5% conservation buffer and the 2% UK countercyclical buffer. Module 4.9 adds what a supervisor would add: a PRA buffer and EU guidance, SREP scores, economic capital by risk and a stress scenario. Every figure it adds is an assumption of the module, labelled illustrative; none is any real bank's data.
What Pillar 1 cannot see
L4.1.01: Capital ratios and the capital stack set out the three pillars. Pillar 1 is a formula: the same risk weights for every bank. It has no charge for interest rate risk in the banking book, none for concentration, none for a pension scheme, and it measures operational risk from income, not from the bank's own losses. Pillar 2Pillar 2: the supervisory review pillar of Basel. The bank assesses its own capital needs (the ICAAP) and the supervisor reviews that assessment and the bank as a whole (the SREP), adding capital where Pillar 1 falls short. fills the gap in two steps. The bank assesses its own capital needs in its ICAAP (L4.9.02: The ICAAP). The supervisor then reviews the bank and its assessment in the SREPSREP (supervisory review and evaluation process): the supervisor's regular review of a bank's business model, governance, capital and liquidity, which ends in scores and in firm-specific requirements and guidance., and sets capital on top of Pillar 1. The Basel Committee's four principles say both halves: banks should have a process for assessing their capital adequacy in relation to their risk profile (SRP20.1), and supervisors “should expect banks to operate above the minimum regulatory capital ratios and should have the ability to require banks to hold capital in excess of the minimum” (SRP20.3).
Pillar 2A: a requirement
The PRA splits the result into two parts (SS31/15 5.10). Pillar 2APillar 2A (UK), or Pillar 2 requirement, P2R (EU): a binding, firm-specific capital requirement for risks Pillar 1 misses or underestimates. It sits with Pillar 1 below the buffers. covers risks Pillar 1 does not capture or does not fully capture, such as IRRBB and concentration. It is binding, set firm by firm, and usually expressed as a share of RWA (5.14, 5.16). With Pillar 1 it makes the total capital requirementTotal capital requirement (TCR, UK) or total SREP capital requirement (TSCR, EU): Pillar 1 plus Pillar 2A. Falling below it is a breach of a minimum requirement.. At least 56.25% of it must be CET1 and at least 75% Tier 1, the same proportions as Pillar 1 (SS31/15 5.16; CRD Art 104a(4) in the EU). RiskAtoms Bank plc's Pillar 2A is the 2.5% that M4.1: The Basel framework assumed illustrative, so its TCR is 10.5% of RWA, £2,394.8m.
The second part looks forward: what could happen to the bank's capital in a downturn. In the UK this is the PRA bufferPRA buffer (UK): a firm-specific amount of CET1 the PRA expects a bank to hold above its TCR and the CRD buffers, so that it can still meet its requirements in a severe stress. Confidential.. It is held “over and above” the TCR and the CRD buffers, so that the bank can keep meeting its requirements in adverse circumstances (SS31/15 5.20). It is all CET1 (5.23), set from the bank's results under severe but plausible stress, among other factors (5.21; SoP5/15), and it can be raised by up to 40% of the CET1 TCR where risk management and governance are significantly weak (5.22). In the EU the matching tool is Pillar 2 guidance (P2G)Pillar 2 guidance, P2G (EU): the ECB's expectation of CET1 above the OCR, set mainly from the stress test. Not binding and outside the MDA.. Both sit outside the MDA: missing P2G “shall not trigger the restrictions referred to in Article 141 or 141b” (CRD Art 104b(6)). Neither is a minimum; a bank that dips into its PRA buffer talks to the PRA and agrees a plan.
Requirement or guidance?
The SREP
In the EU, the EBA's guidelines set one SREP for every supervisor. It has four scored elements: business model analysis, internal governance and institution-wide controls, risks to capital, and risks to liquidity and funding (EBA/GL/2022/03, para 16). Each gets a scoreSREP score: 1 (low risk) to 4 (high risk) for each element and overall, with F for a bank that is failing or likely to fail. from 1 (low risk) to 4 (high risk), or F for a failing bank (paras 34, 44), and the scores lead to the measures: P2R, P2G and qualitative actions. From the 2026 cycle the ECB maps each score straight to a P2R range on a grid, then sets a figure within it by constrained judgement, using the bank's ICAAP figures where they are reliable; requirements set this way apply from 1 January 2027 (ECB Pillar 2 methodology page) as at 7 Nov 2026. The PRA's SREP covers much the same ground: the bank's arrangements and governance, the risks to capital, and liquidity, with a frequency and intensity scaled to the firm (SS31/15 5.1–5.8A).
RiskAtoms Bank plc's SREP scores: 1 low risk to 4 high risk. Illustrative scores of the Module 4.9 data.
Element
Score
Why (illustrative)
Business model assessment
2
Profitable retail and commercial model; NII highly rate-sensitive.
Consumer credit and CRE concentration; drawdown of about 4pp in the bank's adverse run.
Risks to liquidity and funding
2
LCR 157%, NSFR 147%, 47 survival days.
Overall
2
Sort itWhich SREP element?
Sort each finding about RiskAtoms Bank plc into the element where the supervisor would score it.
A 200bp fall in Bank Rate would cut next year's NII by 12%
Customer redress and complaints after a mis-selling review hit fee income and reputation
The board has not reviewed the trading limits framework since it was set
The risk report that feeds the CET1 ratio takes 25 working days to produce
The EU NII outlier test is breached: 6.99% of Tier 1 against 5%
Corporate lending is concentrated in real estate and in one country
The combined liquidity stress leaves 47 days of survival
The bank's stack
RWA £22,807.7m. Buffer rates as at 31 December 2025 (Module 4.1).
Layer
% of RWA
£m
Met in
Pillar 1
8.0%
1,824.6
4.5% CET1, 1.5% AT1, 2% Tier 2
Pillar 2A
2.50%
570.2
1.406% / 0.469% / 0.625%
= TCR
10.50%
2,394.8
Conservation buffer
2.5%
570.2
CET1
UK countercyclical buffer
2.0%
456.2
CET1
O-SII buffer
0%
0.0
below the lowest threshold
PRA buffer illustrative
1.62%
368.5
CET1
All layers
16.62%
3,789.6
The PRA buffer is confidential, so the Module 4.9 data sizes one by a simple rule: the bank's own adverse drawdown, 4.12% (L4.9.03: Designing a supervisory stress test), less the conservation buffer, 1.62% illustrative. SoP5/15's worked example is stricter in one way and kinder in another. It measures the largest fall in the CET1 held above a CET1 hurdle, taken here as 5.906% of RWA (4.5% plus the CET1 part of Pillar 2A), which for the bank is £788.8m in year 1, and takes off the conservation buffer and the “relevant” countercyclical buffer, here the full 2% UK rate: £1,026.3m at today's RWA. On that basis the bank's PRA buffer would be nil: its countercyclical buffer already covers the stress. SoP5/15 does not say which countercyclical rate is relevant, and for banks outside the Bank of England's stress test it sets the hurdle “equal to” the TCR without saying how that converts to CET1, so the nil is illustrative too illustrative. The Try it lets you switch between the two.
The formal version
Requirements, buffers and the MDA line
With P2A as a share of RWA, the combined buffer rate CBR=CCB+CCyB+O-SII, AT1 and Tier 2 held, and shares from SS31/15 5.16 and CRD Art 104a(4):
rCET1GAT1GT2MDA lineMDA factor=4.5%+0.5625P2A,rT1=6%+0.75P2A,TCR=8%+P2A=max(0,(rT1−rCET1)RWA−AT1)=max(0,(TCR−rT1)RWA−T2−max(0,AT1−(rT1−rCET1)RWA))=rCET1+CBR+RWAGAT1+GT2,H=CET1−MDA line×RWA={10.6,0.4,0.2,0H≥0CET1 in the 4th, 3rd, 2nd, 1st quartile of the CBR
The PRA buffer as SoP5/15's example computes it, with h the CET1 hurdle (taken here as 4.5% plus the CET1 part of Pillar 2A) and Et the CET1 above it in year t of the stress:
In the EU (EBA/GL/2022/03, 1.2), with the same P2R shares: TSCR=Pillar 1+P2R, OCR=TSCR+CBR, and P2G above the OCR.
Try it
Start from the booked stack. Raise Pillar 2A to 4%: the MDA line rises faster than the CET1 requirement, because the AT1 and Tier 2 slices grow and CET1 has to fill them. Then take a loss of £600m and watch the bank cross the line and its payout cap fall. Switch the PRA buffer between the module's figure and the SoP5/15 method.
Try itBuild the stack, then take a loss
Read the stack as
PRA buffer:
Pillar 1 CET1 4.5% 4.50%Pillar 2A, CET1 part 1.41%CET1 filling AT1 and Tier 2 gaps 0.32%Conservation buffer 2.50%UK countercyclical buffer 2.00%PRA buffer 1.62%MDA line 10.73%Bank's CET1 13.24%
CET1 ratio13.24%
Where it sitsabove buffers
Total capital requirement (TCR)10.50%
CET1 needed for P1 + P2A5.906%
MDA restriction line (CET1)10.73%
Line with the PRA buffer12.34%
Headroom to the MDA line£573.5m (2.51pp)
Headroom over the PRA buffer too£205.1m
Share of profits it may pay out (MDA factor)no regulatory cap
AT1 and Tier 2 gaps CET1 must fill£73.0m
Illustrative, £ millions at 31 December 2025, RWA held at £22,807.7m. AT1 is £400m and Tier 2 £574.8m; when Pillar 2A rises, their slices grow and CET1 fills the gap first, so the MDA line climbs faster than the CET1 requirement. The PRA buffer sits on top: falling into it means a conversation with the supervisor, not an automatic cap on distributions.
SRP20.1–20.4 the four principles of supervisory review; CRD Arts 104a, 104b, 141 the SREP, P2R, P2G and the MDA. Statements on where rules stand are as at 7 November 2026 as at 7 Nov 2026.
In the UK
SS31/15 sets the PRA's SREP, Pillar 2A and the PRA buffer; SoP5/15 sets its methods. The versions in force are those of October 2025, effective since 1 July 2026. Both have new versions effective 1 January 2027, the day the Basel 3.1 rules take effect (each last updated in July 2026; they take in PS15/26 of May 2026, and SoP5/15 also PS1/26 on Basel 3.1). SS31/15 keeps the 56.25% CET1 share of Pillar 2A (5.16).
In the EU
On 18 November 2025 the ECB published its 2025 SREP for 105 banks: CET1 P2R broadly stable at 1.2%, CET1 P2G down from 1.3% to 1.1% after the 2025 stress test, overall CET1 requirements and guidance of 11.2% for 2026, and an average score of 2.5 (from 2.6), with a quarter of banks still scored 3 or 4. On 26 June 2026 the EBA published revised SREP guidelines that replace EBA/GL/2022/03 from 1 January 2027; they keep the structure and give more weight to ICT, ESG and CSRBB.
In the US
There is no firm-specific Pillar 2 capital add-on of the European kind. The capital rules require a bank to hold capital “commensurate with the level and nature of all risks” and to have a process for assessing its capital adequacy (12 CFR 217.10(e)). For the largest banks the stress test plays the Pillar 2B role: the stress capital buffer replaces the conservation buffer (L4.9.06: Federal Reserve stress tests).
Spark Atoms
✦Spark AtomA buffer for bad governance+
Where the PRA finds risk management and governance significantly weak, it may raise the PRA buffer by a scalar of up to 40% of the bank's CET1 TCR, and take it off once the weaknesses are fixed (SS31/15 5.22). For RiskAtoms Bank plc that would be up to £538.8m. Bad data and weak controls cost capital, not just a letter.
✦Spark AtomScores cluster in the middle+
In the ECB's 2025 SREP the average score improved only from 2.6 to 2.5, and a quarter of banks stayed in the two weakest categories. The ECB's new methodology ties P2R to those scores more directly from 2027, so a one-notch move in a score can change a bank's requirement.
Key takeaways
Pillar 2 is the bank's own assessment (the ICAAP) and the supervisor's review (the SREP); supervisors expect banks to operate above the minimum and can require more (SRP20).
Pillar 2A, or P2R, is binding and sits with Pillar 1: at least 56.25% CET1 and 75% Tier 1. For the bank, 2.5%, a TCR of 10.5% (£2,394.8m).
The PRA buffer and P2G are guidance above the combined buffer, CET1 only, sized from stress; missing them does not trigger the MDA (CRD Art 104b(6)).
The SREP scores business model, governance, risks to capital and risks to liquidity from 1 to 4; from 2027 the ECB maps the scores to P2R ranges.
The MDA line includes the CET1 that fills AT1 and Tier 2 gaps: 10.73%, not 10.41%, leaving £573.5m of headroom.
Reaction
Check what stuck
Question 1 · objective 1
A UK bank's Pillar 2A is 3% of RWA, met in the minimum proportions. What CET1 ratio do Pillar 1 and Pillar 2A together require, in %?
%
Question 2 · objective 1
RiskAtoms Bank plc's CET1 falls from 13.24% to 11.5%: below its CET1 need including the PRA buffer, but above its MDA line. What happens?
Question 3 · objective 3
The bank holds £3,019.9m of CET1. Its CET1 requirement, its AT1 and Tier 2 gaps and its combined buffer take £2,446.4m. What is its MDA headroom, in £m?
£m
Question 4 · objective 2
Which of these is not one of the SREP's four scored elements in the EBA guidelines?
Question 5 · objective 4
In the EU view, with P2R of 2.5% (56.25% CET1), a combined buffer of 4.5% and P2G of 1.25%, what CET1 ratio does the supervisor expect, in %?
%
Glossary
Pillar 2
Pillar 2: the supervisory review pillar of Basel. The bank assesses its own capital needs (the ICAAP) and the supervisor reviews that assessment and the bank as a whole (the SREP), adding capital where Pillar 1 falls short.
SREP
SREP (supervisory review and evaluation process): the supervisor's regular review of a bank's business model, governance, capital and liquidity, which ends in scores and in firm-specific requirements and guidance.
Pillar 2A / P2R
Pillar 2A (UK), or Pillar 2 requirement, P2R (EU): a binding, firm-specific capital requirement for risks Pillar 1 misses or underestimates. It sits with Pillar 1 below the buffers.
TCR / TSCR
Total capital requirement (TCR, UK) or total SREP capital requirement (TSCR, EU): Pillar 1 plus Pillar 2A. Falling below it is a breach of a minimum requirement.
OCR
Overall capital requirement (OCR, EU): the TSCR plus the combined buffer requirement.
PRA buffer
PRA buffer (UK): a firm-specific amount of CET1 the PRA expects a bank to hold above its TCR and the CRD buffers, so that it can still meet its requirements in a severe stress. Confidential.
P2G
Pillar 2 guidance, P2G (EU): the ECB's expectation of CET1 above the OCR, set mainly from the stress test. Not binding and outside the MDA.
MDA
Maximum distributable amount (MDA): the cap on dividends, AT1 coupons and variable pay when CET1 falls inside the combined buffer (CRD Art 141).
SREP score
SREP score: 1 (low risk) to 4 (high risk) for each element and overall, with F for a bank that is failing or likely to fail.
PRA, SoP5/15 The PRA's methodologies for setting Pillar 2 capital, July 2026 update, effective 1 January 2027 (the PRA buffer worked example). Intermediate.
Basel Committee on Banking Supervision, Basel Framework SRP20 Four key principles, version effective 15 December 2019 (bis.org).
Prudential Regulation Authority, SS31/15 The ICAAP and the SREP, October 2025 version, in force from 1 July 2026 (paras 3.8, 5.1–5.38); July 2026 update, effective 1 January 2027.
Prudential Regulation Authority, SoP5/15 The PRA's methodologies for setting Pillar 2 capital, July 2026 update (updating January and May 2026), effective 1 January 2027.
Directive 2013/36/EU (CRD), Arts 104a, 104b and 141 (legislation.gov.uk copy of the EU text).
European Banking Authority, Guidelines on common procedures and methodologies for the SREP and supervisory stress testing, EBA/GL/2022/03, 18 March 2022; and press release of 26 June 2026 on the revised guidelines.
European Central Bank, press release on the 2025 SREP results, 18 November 2025; and Pillar 2 methodology page (bankingsupervision.europa.eu).
12 CFR 217.10(e), Board of Governors of the Federal Reserve System (Cornell LII copy).
Watch this lesson
Video version
Prefer to watch? The same lesson as a 12-minute video will sit here, made from this page's script.