PS9/24 – Implementation of the Basel 3.1 standards near-final part 2

Published on 12 September 2024

1: Overview

1.1 This Prudential Regulation Authority (PRA) near-final policy statement (PS) provides feedback to responses to the following chapters of consultation paper (CP) 16/22 – Implementation of the Basel 3.1 standards:

  • Chapter 1 – Overview
  • Chapter 3 – Credit risk – standardised approach
  • Chapter 4 – Credit risk – internal ratings based approach
  • Chapter 5 – Credit risk mitigation
  • Chapter 9 – Output floor
  • Chapter 11 – Disclosure (Pillar 3)
  • Chapter 12 – Reporting

1.2 This near-final PS also contains feedback to responses on the parts of Pillar 2 relating to the Pillar 2A credit risk methodology, use of internal ratings based (IRB) approach benchmarks, and the interaction with the output floor.

1.3 This PS is the second of two near-final PSs related to Basel 3.1. On 12 December 2023, the PRA published the first PS – Near-final PS17/23 – Implementation of the Basel 3.1 standards near-final part 1. It provided feedback to responses to the following policy areas covered in CP16/22: scope and levels of application, market risk, credit valuation adjustment (CVA) and counterparty credit risk, operational risk, and currency redenomination.footnote [1] It also included feedback to responses on Pillar 2 relevant to those areas. Following that publication, the PRA has made some minor clarifications and corrections to the near-final rules published within PS17/23. These have been incorporated into the following parts of the comprehensive Near-final PRA Rulebook instrument, covering all of the Basel 3.1 standards, included in Appendix 2: (i) the Market Risk Parts (Trading Book (CRR) Part, Market Risk: Advanced Standardised Approach (CRR) Part, Market Risk: Internal Model Approach (CRR) Part), (ii) the Credit Valuation Adjustment Risk Part, and (iii) the Counterparty Credit Risk (CRR) Part.

1.4 The Appendices to this PS contain the PRA’s additional near-final policy material relevant to the above chapters, as follows:

  • near-final PRA Rulebook: CRR Firms: (CRR) Instrument [2024] (Appendix 2);
  • summary of the purpose of the rules – Basel 3.1 (Appendix 3);
  • corresponding CRR rules – Basel 3.1 (Appendix 4);
  • CRR restatement provisions – Basel 3.1 (Appendix 5);
  • near-final amendments to supervisory statement (SS) 10/13 – Credit risk standardised approach (Appendix 6);footnote [2]
  • near-final amendments to SS13/16 – Underwriting standards for buy-to-let mortgage contracts (see footnote 2) (Appendix 7);
  • near-final SS3/24 – Credit risk definition of default (Appendix 8);footnote [3]
  • near-final SS4/24 – Credit risk internal ratings based approach (see footnote 3) (Appendix 9);
  • near-final amendments to SS17/13 – Credit risk mitigation (see footnote 2) (Appendix 10);
  • near-final amendments to SS12/13 – Counterparty credit risk (Appendix 11);
  • near-final updated disclosure templates and instructions (Appendix 12);
  • near-final mapping of reporting template codes (Appendix 13);
  • near-final updated reporting templates and instructions (Appendix 14);
  • near-final SS34/15 – guidelines for completing regulatory reports (Appendix 15);footnote [4]
  • near-final PRA Standards Instrument: Technical Standards (Economic Downturn) Revocation Instrument [2024] (Appendix 16);
  • near-final PRA Rulebook: CRR Firms: SDDT Regime (Interim Capital Regime) Instrument [2024] (Appendix 17);
  • corresponding CRR rules – Interim Capital Regime (Appendix 18); and
  • CRR restatement provisions – Interim Capital Regime (Appendix 19).

1.5 The PRA has not made final rule instruments at this stage because HM Treasury (HMT) is required to first revoke the relevant parts of the Capital Requirements Regulation (CRR)footnote [5]. HMT will make commencement regulationsfootnote [6] to revoke the relevant provisions of the CRR before the PRA can replace them in PRA rules. Once the commencement regulations have been made, the PRA intends to make all the final policy materials, rules, and technical standards (covering the areas in this near-final PS and those in PS17/23) in a single, final PS.

1.6 The near-final rules included in Appendix 2 are relevant to PRA-authorised banks, building societies, PRA-designated investment firms, and PRA-approved or PRA-designated financial holding companies or mixed financial holding companies (firms). As set out in Chapter 9 – Interim Capital Regime,footnote [7] they do not apply to UK banks and building societies that meet the Small Domestic Deposit Taker (SDDT)footnote [8] criteria and choose to be subject to the Interim Capital Regime (ICR) (see Appendix 17 for the near-final rules applicable to such firms). As set out in CP16/22 and described in CP7/24 – The Strong and Simple Framework: The simplified capital regime for Small Domestic Deposit Takers (SDDTs), the PRA intends to use the Basel 3.1 rules as the basis for the SDDT capital regime.

Background

1.7 In CP16/22, the PRA set out its proposals to implement the parts of the Basel III standards that remain to be implemented in the UK. The PRA refers to them as the ‘Basel 3.1 standards’. The proposals addressed the final elements of the Basel III standards concerning the measurement of risk-weighted assets (RWAs) – the denominator of risk-based capital ratios.footnote [9] They aimed to address concerns over the credibility of RWA calculations and firms’ reported capital ratios by improving the measurement of risk in both the internal models (IM) approaches and standardised approaches (SAs), and the comparability of risk measurement across firms. These goals link closely to the PRA’s primary objective by promoting safety and soundness and support its secondary objective to facilitate effective competition by increasing the alignment of modelled and standardised approaches.

1.8 In developing its proposals, the PRA targeted an implementation that was aligned with international standards, with adjustments which are tailored to the UK financial system and advance the PRA’s objectives while supporting key considerations including growth and competitiveness.

1.9 For the CP16/22 chapters relevant to this near-final PS, the PRA proposed the following:

  • Credit risk – standardised approach: a more risk-sensitive approach to risk weighting residential mortgage exposures, changes to the valuation of real estate, revisions to the risk weights for corporate exposures including to small and medium-sized enterprises (SMEs), the introduction of specific treatments for ‘specialised lending’ exposures, removal of assumptions of implicit sovereign support for exposures to banks, changes to the risk weights for equity exposures, changes to off-balance sheet conversion factors (CFs), and proposed due diligence requirements.
  • Credit risk – internal ratings based approach: restrictions on the use of the IRB approach for equities and low default portfolios, such as exposures to banks and other financial institutions, large corporates, and sovereign exposures. Other proposals included revisions to the approaches for corporate exposures including to SMEs, amendments to the treatment of ‘specialised lending’ exposures, changes to the risk parameters used in IRB modelling, including new input floors for probability of default (PD), loss given default (LGD) and exposure at default (EAD), and greater specification of parameter estimation practices to reduce variability in RWAs for portfolios where the IRB approach remains available.
  • Credit risk mitigation: changes to the treatment of both funded and unfunded credit protection and amendments to the PRA’s expectations with respect to credit risk mitigation (CRM) to introduce greater clarity regarding the application of the framework.
  • Output floor: the implementation of an output floor with respect to firms’ calculation of RWAs that would limit the aggregate RWA reductions available to firms through their application of IMs.
  • Disclosure (Pillar 3): updates to the UK Pillar 3 disclosure requirements to reflect the proposals set out in CP16/22 and align them to the Basel 3.1 standards.
  • Reporting: alignment of the PRA supervisory reporting requirements with the other proposals set out in CP16/22.

1.10 The PRA did not propose any policy changes with respect to its Pillar 2 framework in CP16/22, but it provided a high-level description of the implications for Pillar 2 of the changes proposed to the Pillar 1 framework, including aspects related to the Pillar 2A credit risk methodology, use of IRB benchmarks, and the interaction with the output floor.

Summary of responses

1.11 The PRA received formal written responses to CP16/22 from 126 respondents. In addition, the PRA received responses through various other channels including holding over 70 meetings with stakeholders to discuss their views.

1.12 Respondents were generally supportive of the overall approach the PRA set out in CP16/22; however, they raised questions and concerns about a range of issues in the policy areas addressed in this near-final PS. These included requests for additional guidance and clarification on certain proposals and adjustments to specific proposals in favour of treatments that respondents considered would be more flexible or proportionate, less operationally burdensome, or which would reduce the capital impact on firms in order to reflect what respondents considered to be the prudential risk. Some responses focused on the interaction of the package of Basel 3.1 proposals with the PRA’s objectives and the matters to which it must have regard when making policy and rules (‘have regards’). Some provided new evidence for the PRA to take into account when drafting the near-final rules.

1.13 The substantive issues raised in the responses are addressed in detail in the relevant chapters of this near-final PS. Where the PRA has made material changes to its draft policies, it has updated the relevant objectives and have regards analysis in the relevant chapters.

Changes to draft policy

1.14 When making rules, the PRA is required by the Financial Services and Markets Act (FSMA) 2000 to have regard to representations made to it, and to publish an account, in general terms, of its feedback to them. Where the final rules differ from the draft in the CP in a way that the PRA considers is significant, FSMA 2000footnote [10] requires the PRA to publish:

  • details of the difference together with a cost benefit analysis (CBA); and
  • a statement setting out in the PRA’s opinion whether or not the impact of the final rule on mutuals is significantly different to: (i) the impact that the draft rule would have had on mutuals; and (ii) the impact that the final rule will have on other PRA-authorised firms, and if so, details of the difference.

1.15 The publication requirement however does not apply if the PRA considers that (making the appropriate comparison), there will be no increase in costs or there will be an increase in costs, but that increase will be of minimal significance.footnote [11]

1.16 Taking into account the responses to CP16/22, the PRA has made a number of adjustments and corrections to the draft policy where it considers this to be appropriate. These changes reflect the evidence and arguments stakeholders have provided in their responses and in meetings with the PRA. The PRA welcomes the high level of engagement with stakeholders that has informed and enabled these changes, in line with the PRA’s objectives. The most material changes to the policy proposed in CP16/22 include:

  • Amendments to the treatment of SME lending to reduce the operational burden for firms and lower capital requirements. The near-final policy includes: (i) a new firm-specific structural adjustment to Pillar 2A (the ‘SME lending adjustment’) to ensure that the removal of the SME support factor under Pillar 1 does not result in an increase in overall capital requirements for SME exposures; (ii) a new, simplified definition of SME to reduce the operational burden for firms applying the definition and broaden the scope of exposures which qualify for preferential treatment as an SME; and (iii) the removal of the requirement that SA risk weights for commercial real estate (CRE) exposures should be no lower than 100% for exposures where repayment is not materially dependent on cashflows from the property, resulting in lower risk weights.
  • Amendments to the treatment of infrastructure lending to lower capital requirements. The near-final policy includes: (i) a new firm-specific structural adjustment to Pillar 2A (the ‘infrastructure lending adjustment’) to ensure that the removal of the infrastructure support factor under Pillar 1 does not result in an increase in overall capital requirements for infrastructure exposures; and (ii) a new lower risk weight of 50% for ‘substantially stronger’ project finance exposures under the slotting approach, which is a reduction from the CP16/22-proposed risk weight of 70% for qualifying exposures.
  • Lower, more risk-sensitive CFs for off-balance sheet items to better reflect empirical data provided by respondents. The near-final rules: (i) assign a 20% CF for ‘transaction-related contingent items’, a reduction from the 50% CF proposed in CP16/22; and (ii) assign a 40% CF for ‘other commitments’ (except for UK residential mortgage commitments), a reduction from the 50% CF proposed in CP16/22.
  • A more risk-sensitive and operationally simpler approach to the valuation of residential real estate under the credit risk SA. The near-final rules: (i) include a ‘backstop’ revaluation event requiring firms to obtain a new valuation once five years (or three years in certain cases) has passed since the last revaluation event, which will prevent products without natural refinancing events (eg lifetime mortgages) from being unjustifiably disadvantaged; (ii) remove the requirement for firms to adjust a valuation to reflect the value of the property that would be sustainable over the life of the loan; and (iii) clarify downward revaluation requirements, making them more mechanistic, and therefore simpler, by requiring firms to revalue properties if they estimate (through their regular monitoring) that the market value of the property has decreased by more than 10% relative to the value recorded at the last valuation event.
  • An approach to calculating the output floor which enhances consistency between SA approaches and the output floor. The near-final rules adjust the output floor calculation to reflect the different treatment of accounting provisions under the SA and the IRB approach.

1.17 The PRA considers that the changes to the draft policy, including those listed above, are appropriate to reflect risks in a more proportionate manner, reduce operational burden on firms, enhance the relative standing of the UK and improve the clarity of rules in a manner that aligns with the PRA’s statutory objectives. The changes reflect the evidence and arguments that firms provided during the consultation period. Further details on all substantive issues raised in responses, and any related amendments to the draft policy, are set out in the relevant chapters of this near-final PS.

1.18 The PRA has also made a number of less substantive changes and clarifications to the draft policy which are not described in the chapters of this near-final PS. These are reflected in the near-final PRA Rulebook: CRR Firms: (CRR) Instrument [2024] in Appendix 2 and in the relevant near-final SSs. The document titled Comparison of Draft PRA Rulebook (CRR) Instrument [2023] against the near-final PRA Rulebook: CRR Firms (CRR) Instrument [2024]footnote [12] contains a comparison of the near-final rules with the draft rules as set out in CP16/22 for ease of identifying all of the rule-related changes made.

1.19 The PRA has also made a number of amendments and clarifications to reporting and disclosure templates not all of which are described in this near-final PS. Amendments and clarifications to the reporting templates and instructions are included in the near-final updated reporting templates and instructions (Annex 14).

1.20 The PRA has considered the aggregated CBA presented in CP16/22 in light of the adjustments to the proposals within this near-final PS and PS17/23 and finds it remains appropriate. In the PRA’s view, the adjustments would not significantly increase costs for firms relative to the costs implied by the proposals in CP16/22. The significant majority of the changes to CP16/22 are intended to result in a more proportionate treatment for firms and the PRA has adjusted some of its proposals to reduce operational burden on firms. Comparing the costs implied by the proposals in CP16/22 with the costs implied by both PS17/23 and this near-final PS, the PRA considers that that there will be minimal, if any, increase in overall costs, and that many firms should see a reduction in costs. Therefore, a fully revised CBA and mutuals statement has not been produced. Where amendments to individual policies have changed their impact and/or where the PRA has undertaken further analysis of their impact, the PRA has included analysis within the relevant chapter of this near-final PS.

1.21 The PRA does not consider that the impact of the near-final policy and rules in this near-final PS would have a significantly different impact on mutuals relative to the impact of the draft policy and rules on mutuals, or relative to the impact of the near-final policy and rules on other PRA-authorised firms.

Structure of the policy statement

1.22 This near-final PS is structured into the following chapters. The near-final rules and related policy material are included in the relevant appendices.

  • Chapter 2 – Credit risk – standardised approach
  • Chapter 3 – Credit risk – internal ratings based approach
  • Chapter 4 – Credit risk mitigation
  • Chapter 5 – Output floor
  • Chapter 6 – Pillar 2
  • Chapter 7 – Disclosure (Pillar 3)
  • Chapter 8 – Reporting
  • Chapter 9 – Interim Capital Regime

Accountability framework

1.23 The near-final policy and rules set out in this near-final PS have been developed by the PRA in accordance with its statutory objectives and informed by the regulatory principles and the matters to which it must have regard when making policy and rules as set out in FSMA 2000. In CP16/22, the PRA set out details of the applicable accountability framework in Appendix 6 – PRA statutory obligations and provided its assessment of relevant considerations separately in each chapter. The PRA has provided a summary of its updated overall assessment below, having considered the consultation responses. More detailed, updated explanations are included in the relevant chapters of this near-final PS, where the PRA has made changes to the draft policy.

1.24 The PRA considers that the near-final policy and rules in this near-final PS will advance its primary objective to promote the safety and soundness of the firms it regulates. They address shortcomings in the RWA framework revealed by the global financial crisis, including inadequate Pillar 1 capital requirements in some areas and a ‘worrying degree of variability’ in the calculation of risk weights noted by the Basel Committee on Banking Supervision (BCBS) in December 2017.footnote [13] Addressing these shortcomings is important to underpin confidence that firms are adequately capitalised, given the risks to which they are exposed, which in turn supports financial stability. In particular, the near-final policy and rules will advance the PRA’s primary objective by:

  • simplifying and reducing the range of approaches available for RWA calculations, thereby promoting the consistent application of approaches across firms through simpler and clearer requirements;
  • improving the risk-sensitivity of the credit risk SA, resulting in RWAs that are more reflective of risk;
  • constraining the use of the credit risk IRB approach in areas where RWAs cannot be modelled in a robust and prudent manner (such as for central government and central bank exposures). This will reduce unwarranted RWA variability across firms; and
  • introducing the output floor, limiting how low internally modelled RWAs can fall below those produced under the revised SAs and thereby reducing excessive variability and cyclicality in RWAs.

1.25 The near-final policy and rules within this near-final PS will also advance the PRA’s secondary objective to facilitate effective competition by:

  • increasing risk-sensitivity under the credit risk SA, which aims to ensure firms using the SA calculate RWAs that are better aligned to the risk of their exposures and to those calculated by firms with permission to apply the IRB approach;
  • requiring more robust and prudent modelling practices within the IRB approach, and limiting the reduction in RWAs that can be achieved through its use, thereby reducing the potential competitive advantage that firms using the IRB approach currently have over firms using the SA;
  • promoting a more consistent and level playing field by introducing an output floor, which would ensure RWAs for firms with IM permissions do not fall below a defined percentage of the RWAs calculated under the SAs; and
  • reducing barriers to entry for SA firms to use the IRB approach, for example by permitting the roll-out of the IRB approach by individual roll-out classes in certain circumstances, rather than requiring roll-out across all credit risk exposures, and by changing the approval threshold for new IRB permissions to ‘material compliance’ from ‘full compliance’.

1.26 Changes to the PRA’s accountability framework under the FSMA 2023 to add international competitiveness and growth (subject to alignment with international standards) as a secondary objective, do not formally apply to the near-final policy and rules set out in this near-final PS. These provisions are disapplied by regulation 4 of the FMSA 2023 (Commencement No. 2 and Transitional Provisions) Regulations 2023 as amended by regulation 18 of the FSMA 2023 (Commencement No. 4 and Transitional and Saving Provisions) (Amendment) Regulations 2023.

1.27 Nevertheless, as the PRA noted in CP16/22 and PS17/23, international competitiveness and growth considerations, and alignment with international standards, are factors the PRA must ‘have regard’ to. These factors, alongside the relative standing of the UK as a place for internationally active firms to operate, were strongly considered by the PRA in developing the near-final policy and rules on the Basel 3.1 standards.

1.28 In considering these factors and the characteristics of the UK’s financial system, the proposals in CP16/22 included some adjustments relative to international standards to support the international competitiveness of the UK, while advancing the PRA’s primary objective. For the elements of CP16/22 relevant to this PS, these included, for example, targeted measures to address concerns with respect to the risk-weighting of exposures to unrated corporates under the credit risk SA.

1.29 The cumulative changes to the draft policy in this near-final PS, informed by evidence submitted during the consultation process, represents a significant further step in facilitating growth and competitiveness. This includes, for example, changes to CFs to align more closely with those of other jurisdictions. And with the introduction of the SME lending adjustment and the infrastructure lending adjustment, such that the removal of the infrastructure and SME support factors does not result in an increase in overall capital requirements for infrastructure and SME lending, the near-final package represents a material change from the draft policy which will provide stability and certainty for firms to continue to lend to those important sectors of the UK economy.

1.30 In PS17/23, the PRA estimated, based on data submitted by firms, that the package of policies to implement Basel 3.1, as set out in CP16/22, would result in a 3.2% increase in Tier 1 capital requirements for major UK firms at the end of the transitional period on 1 January 2030. This estimate took into account the policies within CP16/22 and the PRA’s general principle that it would not require firms to calculate capital requirements for the same risks under both the Pillar 1 and Pillar 2 frameworks. Both of these considerations decreased the estimated impact.footnote [14] It did not take into account the changes in either PS17/23 or this near-final PS relative to CP16/22. Those changes will further reduce the expected impact on capital requirements of Basel 3.1 implementation. The PRA now estimates that the Tier 1 capital requirements for major UK banks in aggregate are likely to increase by less than 1% by 1 January 2030. By comparison, the European Banking Authority (EBA) has estimated the latest EU proposals would increase Tier 1 minimum capital requirements for EU firms by 9.9% when fully phased-in, and by 5.6% while EU-specific transitional arrangements are in place.footnote [15] US regulatory agencies are in the process of adjusting their original proposals and intend to publish the revised proposals for re-consultation alongside proposed changes to the globally systemic important bank (G-SIBs) surcharge. Taken together, regulatory agencies estimate these proposals would lead to a 9% increase in CET1 capital requirements for G-SIBs. For other large banks that are not G-SIBs, the new proposals would result in an estimated increase of 3.5-4.5%.footnote [16] 

1.31 The PRA considers that the near-final policy and rules still align with international standards, including by removing pre-existing deviations. As set out in PS17/23, alignment with international standards supports the UK’s international competitiveness, including the relative standing of the UK as a global financial centre, by:

  • strengthening key stakeholders’ confidence in the UK banking system, by addressing widespread and long-standing concerns around the credibility of the calculation of RWAs;
  • assuring regulators in other jurisdictions of UK authorities’ commitment to robust standards, which allows for deep interconnections between the financial systems of their jurisdictions and the UK;
  • minimising operational complexities for firms that operate across multiple jurisdictions; and
  • providing stability and predictability to firms and other stakeholders through the implementation of internationally aligned standards. This promotes a high-trust environment in the UK, where firms and stakeholders can confidently conduct their business over the medium term.

1.32 By delivering a package that includes evidence-based adjustments to reflect the UK’s financial system, but remains aligned with international standards, the PRA considers that the near-final rules will both support competitiveness and facilitate the flow of capital within the UK banking system. A vibrant banking system in turn underpins core economic activities in the UK and supports a healthy and growing domestic economy. Consumers and businesses can borrow, invest, and manage risk with confidence that individual institutions within the sector are sufficiently robust to withstand economic shocks and can therefore maintain lending to the real economy during downturns.

1.33 The PRA has included in this near-final PS a summary of the purpose of the proposed rules (set out in Appendix 3), a statement identifying CRR restatement provisions (set out in Appendix 5), and a statement identifying corresponding CRR rules (set out in Appendix 4).

Interactions with other frameworks

Leverage ratio

1.34 Effective as of 1 January 2022, following the 2021 Financial Policy Committee (FPC) and PRA review, the PRA implemented most of the changes to the calculation of the leverage exposure measurefootnote [17] that were made to international standards in the process of finalising Basel 3.1. However, some of the changes to those international standards were not implemented at that time as they related to changes to the credit risk SA in the risk-weighted capital framework set out in this near-final PS. While the PRA is not including policy with respect to the leverage ratio requirement in this near-final PS, the changes to the treatment of off-balance sheet items (ie CFs) in Chapter 2 – Credit risk – standardised approach will flow through to the leverage framework.

Large exposures

1.35 The large exposure requirements in the CRR have already been transferred to PRA rules and amended to implement the Basel III standards. However, the Basel 3.1-related large exposure standards were not implemented at that time because they depend on changes to the credit risk SA in the risk-weighted capital framework set out in this near-final PS. The PRA is not including changes to the large exposure requirement in this near-final PS, however changes to prudential standards will have a consequential impact on the large exposure requirements. For example, CRM techniques set out in Chapter 4 – Credit risk mitigation for calculating capital requirements will also be used for calculating net large exposures, subject to the broader conditions in the large exposure requirements.

Liquidity risk

1.36 Effective as of 1 January 2022, the PRA implemented the Basel III net stable funding ratio (NSFR) standard in PRA rules, and it transferred the BCBS liquidity coverage ratio (LCR) standard into PRA rules. The Basel 3.1 standards did not make amendments to either liquidity standard. Where relevant, however, the changes to prudential standards set out in this near-final PS will automatically flow through to the LCR and NSFR, including in relation to risk weights for mortgage loans under the credit risk SA.

Stress testing

1.37 The PRA buffer (also referred to as Pillar 2B) is an amount of capital firms should maintain in addition to their total capital requirement and the combined buffer. The PRA buffer absorbs losses that may arise under a severe stress scenario, while avoiding duplication with the combined buffers. Its size is informed by a range of factors including the results of relevant stress-testing, an assessment of whether a firm has significant risk management and governance weaknesses, and supervisory judgement. A number of measures in the PRA’s near-final rules, including transitional arrangements, the more risk-sensitive credit risk SA, and the output floor, may increase complexities in assessing how RWAs may evolve under a severe stress scenario for some UK firms. This, in turn, could impact the size of the PRA buffer.

1.38 The PRA acknowledges the interaction between Basel 3.1 and dynamics in stress. The PRA will provide firms with more guidance in due course on what to assume for Basel 3.1 in forming their stress test projections as part of the next concurrent exercise involving firm submissions of stressed projections.

Implementation and next steps

1.39 The policy material in this near-final PS is published as near-final. The PRA does not intend to change the policy or make substantive alterations to the instruments before the making of the final policy material.

1.40 The PRA intends to publish the final rule instruments, technical standards instrument and policy in a single, final PS covering the entire Basel 3.1 package, once HMT has made commencement regulations to revoke the relevant parts of the CRR that the final PRA rules will replace. The PRA has also published supporting materials relating to new or revised Authorisations processes.

1.41 In CP16/22, the PRA consulted on an implementation date of 1 January 2025, with a five-year transitional period. In PS17/23, the PRA took into account feedback from firms on the need for sufficient time to robustly implement the Basel 3.1 standards and the latest information at that time on the implementation timelines of other major jurisdictions, and moved the proposed implementation date for the Basel 3.1 standards back by six months to 1 July 2025 with a 4.5 year transitional period.

1.42 Since the publication of PS17/23, the PRA has continued to monitor the implementation timelines of other jurisdictions and to assess the adequacy of the period between publication of PRA rules and their implementation. Having considered these issues, the PRA has decided to move the implementation date for the Basel 3.1 standards by a further six months to 1 January 2026 with a transitional period of 4 years to ensure full implementation by 1 January 2030 in line with the proposals originally set out in CP16/22.footnote [18]

1.43 Any references related to the UK’s membership of the EU in the SSs and Statement of policy (SoP) covered by the near-final policy in this near-final PS will be updated as part of the final PS to reflect the UK’s withdrawal from the EU. Unless otherwise stated, any remaining references to EU or EU-derived legislation are to the version of that legislation which forms part of assimilated law in the UK.footnote [19]

2: Credit risk – standardised approach

Introduction

2.1 This chapter provides feedback to responses to Chapter 3 of consultation paper (CP)16/22 – Implementation of the Basel 3.1 standards, which set out proposals to change the standardised approach (SA), for calculating risk-weighted assets (RWAs) for credit risk in the Capital Requirements Regulation (CRR) and the Prudential Regulation Authority’s (PRA) expectations, in response to the Basel 3.1 standards. This chapter also sets out the PRA’s near-final policy on the SA for credit risk following the consultation.

2.2 In CP16/22, the PRA proposed to introduce changes to enhance the risk sensitivity and robustness of the SA, including:

  • the introduction of a more structured and granular exposure allocation and increased risk sensitivity of the treatment for real estate lending;
  • a more risk-sensitive treatment for exposures to unrated corporates;
  • changes to deliver a simpler, more risk-sensitive and prudent approach for risk-weighting infrastructure lending, including the introduction of specific treatments for ‘specialised lending’ and the removal of the infrastructure support factor;
  • changes to deliver a simpler, more transparent and prudent mechanism for determining risk weights for lending to small and medium-sized enterprises (SMEs), including the removal of the SME support factor;
  • the introduction of a more risk-sensitive approach to risk-weighting equity exposures, including a prudent treatment for higher-risk ‘speculative unlisted equity’;
  • changes to off-balance sheet conversion factors (CFs) aligned to local UK market conditions; and
  • changes to reduce mechanistic reliance on external ratings and the introduction of new due diligence requirements.

2.3 The PRA received 87 responses to its proposals on the SA for credit risk. Responses covered a wide range of the PRA’s proposals. Comments focused particularly on the following points:

  • the proposed treatment of unrated corporate exposures;
  • the proposals to remove the SME support factor and infrastructure support factor and other aspects of SME lending;
  • the proposed changes to the treatment of real estate exposures (including the proposed risk weight treatment and the proposal to require that the value of the property be the value at origination); and
  • the proposals to change the CFs for certain types of off-balance sheet items.

2.4 Having considered the responses to the consultation, the PRA has decided to make significant amendments to the draft rules and draft supervisory statements (SSs) in response to the feedback and evidence provided by respondents. This chapter describes the comments and amendments that the PRA considers most material. As described in Chapter 1 – Overview, the PRA has also made a number of less substantive amendments and clarifications to the draft rules, which are not described in this chapter. These amendments are reflected in the near-final PRA Rulebook: CRR Firms: (CRR) Instrument [2024] in Appendix 2. Please refer to the document titled Comparison of Draft PRA Rulebook (CRR) Instrument [2023] against Near-final PRA Rulebook: CRR Firms (CRR) Instrument [2024], which contains a comparison of the near-final rules with the draft rules as set out in CP16/22 for ease of identifying all of the changes made.

2.5 The appendices to this near-final policy statement (PS) contain the PRA’s near-final policy, which will:

  • introduce a new Credit Risk: Standardised Approach (CRR) Part of the PRA Rulebook (Appendix 2) relating to the SA approach to replace CRR articles that HM Treasury (HMT) plans to revoke;
  • introduce a new Credit Risk: General Provisions (CRR) Part of the PRA Rulebook (Appendix 2) relating to general provisions for credit risk to replace CRR articles that HMT plans to revoke;
  • amend the existing Credit Risk Part of the PRA Rulebook (Appendix 2);
  • remove the existing Standardised Approach and Internal Ratings Based Approach to Credit Risk (CRR) Part of the PRA Rulebook;
  • amend SS10/13 – Credit risk standardised approach (Appendix 6), SS13/16 – Underwriting standards for buy-to-let mortgage contracts (Appendix 7) and SS17/13 – Credit risk mitigation (Appendix 10); and
  • introduce a new SS3/24 – Credit risk definition of default (Appendix 8).

2.6 The sections below have been structured broadly along the same lines as chapter 3 of CP16/22, covering the main areas where the PRA received comments from respondents as follows:

  • exposure class allocation;
  • external credit ratings and due diligence;
  • exposure values for off-balance sheet items;
  • exposures to central governments, central banks, regional governments, local authorities, public sector entities (PSEs), and multilateral development banks (MDBs);
  • exposures to institutions and covered bonds;
  • exposures to corporates and specialised lending;
  • exposures to individuals and SMEs;
  • real estate exposures; and
  • capital instruments, defaulted exposures, and high-risk items.

2.7 Unless specified otherwise, the near-final rules and near-final amendments to SSs are consistent with those in CP16/22 and therefore the PRA considers its analysis of its objectives and ‘have regards’ in CP16/22 with respect to the areas mentioned in the paragraph above remains appropriate.

Exposure class allocation

2.8 The PRA proposed an exposure class hierarchy for reporting purposes in chapter 12 of CP16/22, but it did not propose to introduce an explicit provision in its draft rules on how firms should allocate exposures to the appropriate exposure class.

2.9 The PRA received one response to this proposal, requesting more clarity on the correct allocation of exposures among exposure classes.

2.10 Having considered the response, the PRA has amended its draft rules to include an exposure class hierarchy to promote a consistent allocation across firms. The near-final rules set out the criteria to determine how exposures will be assigned to an exposure class. Where an exposure meets the criteria for more than one exposure class, it will be assigned to an exposure class, in the following order of precedence:

  • securitisation positions;
  • exposures in the form of units or shares in collective investment undertaking (‘CIUs’);
  • subordinated debt, equity and other own funds instruments;
  • exposures associated with particularly high risk;
  • exposures in default;
  • exposures in the form of eligible covered bonds;
  • real estate exposures;
  • exposures to international organisations;
  • exposures to MDBs;
  • exposures to institutions;
  • exposures to central governments or central banks;
  • exposures to regional governments or local authorities;
  • exposures to public sector entities;
  • retail exposures;
  • exposures to corporates; and
  • other items.

2.11 In the near-final rules, the PRA has decided to make the following amendments to the relative precedence of each exposure class that was proposed for reporting purposes:

  • move the ‘units or shares in CIUs’ exposure class up the hierarchy to reflect the intent of the near-final rules in relation to this exposure class;
  • clarify that firms must allocate exposures to the ‘subordinated debt, equity and other own funds instruments’ exposure class ahead of the ‘exposures associated with particularly high risk’ exposure class. The PRA considers that this appropriately reflects the relative risk weight treatments for each exposure class in the near-final rules;
  • remove the ‘institutions and corporates with a short-term credit assessment’ exposure class as these exposures will be treated under the institutions exposure class or corporate exposure class, as appropriate;
  • add exposure classes that were omitted from the hierarchy proposed for reporting purposes to clarify their position in the exposure class hierarchy; and
  • move the ‘other items’ exposure class to the bottom of the hierarchy.

PRA objectives analysis

2.12 The PRA considers that the introduction of an exposure class hierarchy in its near-final rules will advance its primary objective to promote the safety and soundness of regulated firms by improving the consistency of firms’ allocation of exposures, resulting in capital requirements that appropriately reflect the risk of exposures.

2.13 The PRA also considers that its decision to introduce an exposure class hierarchy in its near-final rules will advance its secondary objective to facilitate effective competition. Providing clarity and certainty on the approach to exposure class allocation will result in a more consistent calculation of capital requirements across firms.

‘Have regards’ analysis

2.14 In developing these near-final rules, the PRA has had regard to its framework of regulatory principles and the matters to which it is required to have regard when proposing amendments to CRR rules. The PRA considers that the hierarchy for exposure class allocation in the near-final rules is aligned with the Basel 3.1 standards.

External credit ratings and due diligence

2.15 The PRA proposed a number of changes to enhance the robustness of the use of external credit ratings and introduce due diligence requirements. The PRA received 21 responses to this section of CP16/22. The substantive issues raised by respondents are set out below.

Nomination of external credit assessment institutions

2.16 The PRA proposed that firms should use credit ratings of their nominated external credit assessment institutions (ECAIs) consistently for all types of exposures, for risk management and risk-weighting purposes. The PRA received 11 responses to its proposals:

  • four respondents supported the proposals;
  • one respondent requested that the PRA clarify the interaction between its proposed requirements and the high-level requirements for using the internal ratings based (IRB) approach, which state that internal ratings should play an essential role in the risk management process;
  • one respondent did not support the proposal to require nominated ECAIs to be used consistently across exposure classes, arguing the requirements would reduce competition between ECAIs by favouring those with larger ratings coverage; and
  • two respondents requested that the PRA clarify the procedure and timeline for firms to change their nominated ECAIs, with one arguing that firms needed assurance that they could change nominated ECAIs easily.

2.17 Having considered the responses, the PRA has decided not to amend its draft rules in relation to the process by which firms nominate ECAIs. However, it has amended its draft rules to clarify that the nomination of ECAIs for risk weighting purposes should be consistent with the firm’s use of ECAIs for general risk management purposes. The near-final rules do not restrict or place additional requirements on the use of ECAIs for risk management purposes and broader business decisions (including the high-level requirements for using the IRB approach). In particular, if a firm does not use an ECAI for its risk management purposes or broader business decisions, it will not be required to do so under the near-final rules and the firm will be free to use multiple ECAIs (beyond those nominated for risk-weighting purposes) for risk management purposes. However, a firm will not be able to ‘cherry-pick’ a different ECAI for risk-weighting purposes from one of those it already relies on for general risk management purposes.

2.18 The PRA does not consider that its near-final rules disproportionately favour the use of ECAIs with wider ratings coverage, given firms can nominate multiple ECAIs. Where two credit ratings are available for a particular exposure from nominated ECAIs, the one resulting in the higher risk weight will be assigned. Where more than two credit ratings are available from nominated ECAIs, the two that would produce the lowest risk weights will be referred to and the higher risk weight of the two shall be assigned. This reduces the potential for firms to ‘cherry-pick’ their use of ECAI ratings to lower RWAs without inappropriately penalising firms for nominating multiple ECAIs (firms are not required to use the most conservative rating where more than two ratings from nominated ECAIs are available).

2.19 The PRA also considers that its near-final rules provide sufficient clarity on the process by which firms can nominate ECAIs. The PRA’s near-final rules do not prohibit firms revoking such a nomination, provided firms are not seeking to change ECAIs for the purpose of reducing capital requirements.

Credit ratings for institutions

2.20 The PRA proposed to align with the Basel 3.1 standards and not permit the use of credit ratings for institutions that incorporate assumptions of implicit government support. The PRA received the following responses in relation to these proposals:

  • one respondent argued that the PRA should apply a five-year transitional period where firms could continue to use credit ratings that consider implicit government support, arguing this would align with a national discretion in the Basel 3.1 standards and the approach taken by other jurisdictions;
  • one respondent requested that the PRA clarify if firms could use a credit rating that considers implicit government support in its methodology, if the ECAI has assigned no benefit from such implicit support in determining the credit rating;
  • one respondent requested that the PRA clarify if a rating for an institution that assumes no implicit government support should always be used instead of an issue-specific rating that assumes implicit government support; and
  • one respondent requested that the PRA clarify the definition of implicit support and how it defined a publicly-owned bank.

2.21 Having considered the responses, the PRA has decided to maintain its overall proposal that firms are not permitted to use credit ratings that incorporate assumptions of implicit government support, and to not introduce any transitional arrangements. The PRA considers this to be a proportionate approach that advances its primary objective and notes that a number of ECAIs have already launched, or announced their intention to launch, credit ratings that do not incorporate assumptions of implicit government support.

2.22 The PRA has clarified in its near-final amendments to SS10/13 (Appendix 6), that firms may use credit ratings which consider implicit government support in their methodologies, provided the ECAI’s credit assessment is clear that, while the existence of implicit government support has been considered, no benefit for implicit government support has been included in the credit assessment. The PRA expects firms to continue to monitor whether such ratings assign any benefit for implicit government support as part of their due diligence.

2.23 The PRA acknowledges that there could be cases where prohibiting the use of a directly applicable issue-specific credit rating that considers implicit government support results in an inappropriate reduction in the risk weight assigned to an exposure. It has therefore decided to amend its draft rules to clarify that if all available issue-specific credit ratings incorporate assumptions of implicit government support then firms must use the issue-specific credit rating if it would lead to a higher risk weight than if it were disregarded. This will ensure that risk weights are not lowered due to use of ratings that assume implicit government support, while still appropriately reflecting issue-specific risks.

2.24 Regarding the definition of implicit government support, the PRA considers that its draft rules are sufficiently clear and so has not made any amendments other than those set out above. Aligning the level of detail prescribed in the near-final rules with that in the Basel 3.1 standards and in the implementation of these standards by other jurisdictions will support UK firms having access to a wide range of credit ratings. However, the PRA will consider providing further guidance if evidence emerges of inconsistent application between firms.

External credit assessment institution mapping

2.25 The PRA proposed not to amend the mapping of external credit ratings to the credit quality steps (CQS) set out in Commission Implementing Regulation (EU) 2016/1799 of 7 October 2016.

2.26 Two respondents requested that the PRA update the approach to mapping external credit ratings to CQS, arguing that the existing UK approach is difficult to navigate as it is no longer aligned with the EU. Having considered the responses, the PRA has decided not to amend the mapping of external credit ratings as part of this near-final PS. However, the PRA intends to consult on its policy on the allocation of credit assessments to an objective scale of CQS in H2 2024 within a consultation on proposed changes to the CRR for banks, building societies and investment firms.

Other issues relating to external credit ratings

2.27 Alongside considering the responses to its proposals on the use of external credit ratings, the PRA has decided to make minor amendments to its draft rules to clarify when credit ratings that are not directly applicable or issue-specific are to be used, and when short-term credit ratings are to be used. The PRA considers that this will support consistent application of its near-final rules across firms.

Due diligence

2.28 The PRA proposed to introduce due diligence requirements comprising two elements: (i) a requirement for monitoring counterparties; and (ii) for externally-rated exposures, a requirement to increase risk weights in cases where a firm’s due diligence indicates that an exposure has higher credit risk than the external credit rating would imply.

2.29 The PRA received 18 responses to its proposals. Six respondents requested that the PRA clarify aspects of the proposed due diligence requirements, with five respondents arguing this was necessary to ensure they were proportionate for smaller firms. In particular, two respondents requested that the PRA clarify the factors that should determine the sophistication of due diligence assessments and one respondent requested further guidance on how and when firms would be expected to increase risk weights in cases where a firm’s due diligence indicates that an exposure has higher credit risk than the external credit rating would imply.

2.30 Having considered the responses, the PRA has decided to maintain its proposed approach. Firms are currently required to understand and manage their risks under the existing Fundamental Rules and Internal Capital Adequacy Assessment Process (ICAAP) Parts of the PRA Rulebook, so the PRA expects firms to already have a good understanding of how to monitor counterparties. The PRA has also decided not to provide further guidance at this time on the proposed requirement to increase risk weights in cases where a firm’s due diligence indicates that an exposure has higher credit risk than the external credit rating would imply, but may consider providing such guidance in the future if evidence emerges of inconsistent application between firms.

2.31 Six respondents also argued that Small Domestic Deposit Takers (SDDTs) should be excluded from the due diligence requirements and four respondents argued smaller firms were not able to better assess credit risk than their nominated ECAI. The SA for SDDTs is outside the scope of this near-final PS. However, as set out in CP7/24 – The Strong and Simple Framework: The simplified capital regime for Small Domestic Deposit Takers (SDDTs), the PRA is proposing to exempt SDDTs from due diligence requirements under the regime that applies to SDDTs.

Other issues relating to due diligence

2.32 Other substantive responses to this section of CP16/22 focused on the following issues:

  • Due diligence requirements for MDBs: one respondent requested that the PRA consider excluding MDBs that receive a 0% risk weight from due diligence requirements, consistent with the exemption for exposures to PSEs, central banks and central governments. Having considered this response, the PRA has decided to amend its draft rules to exempt exposures to these MDBs from the proposed due diligence requirements. The PRA has also decided to amend its draft rules to exempt exposures to international organisations named in Article 118 of the near-final Credit Risk: Standardised Approach (CRR) Part from the proposed due diligence requirements. These amendments will provide consistency between counterparties that share similar risk profiles and result in more proportionate requirements. The PRA still expects firms to monitor their counterparties in order to understand and manage their risks.
  • Due diligence requirements for exposures subject to SA for firms with IRB permissions: three respondents requested that the PRA clarify whether existing IRB risk management processes could be used to conduct due diligence on exposures risk weighted under the SA through permanent partial use or roll-out. The PRA clarifies that, under its near-final rules, for all exposures to which the SA is applied, due diligence processes should be used that meet the requirements set out in Article 110A of the near-final Credit Risk: Standardised Approach (CRR) Part. Where a firm applies a common risk management approach to a portfolio consisting of exposures, some of which are subject to the SA and some of which are subject to the IRB approach, the PRA considers it likely that this approach would meet the due diligence requirement to monitor counterparties for exposures on the SA. However, firms with IRB permissions will still be required to perform additional due diligence for exposures on the SA to assess whether an externally-rated exposure has a higher credit risk than the external credit rating would imply. The PRA considers that it would be inconsistent with its secondary objective to set different requirements for firms with IRB permissions compared with firms only using the SA in respect of exposures to which the SA is applied.
  • Due diligence requirements when calculating the output floor: one respondent requested that the PRA clarify that existing IRB processes can be used to meet due diligence requirements for the purpose of calculating the output floor. For the purpose of calculating the output floor, the PRA considers that firms with IRB permissions will generally meet the monitoring counterparties due diligence requirements for exposures risk-weighted under the IRB approach as a result of their application of IRB rules. They will also, for exposures subject to the IRB approach, be able to use approved IRB models to carry out the assessment of whether an ECAI’s credit rating results in an appropriate SA risk weight for the purpose of calculating the output floor.

PRA objectives and ‘have regards’ analysis

2.33 The near-final policy on the use of external credit ratings and due diligence requirements under the SA approach is materially aligned with the proposals in CP16/22 and the PRA therefore considers its analysis of its objectives and ‘have regards’ in CP16/22 remains appropriate.

Exposure values for off-balance sheet items

2.34 The PRA proposed a number of changes relating to off-balance sheet items, including:

  • the introduction of a new definition of ‘commitment’; and
  • a change to the applicable CF for some issued off-balance sheet items and commitments.

2.35 The applicable CF will depend on the approach used to calculate RWAs for credit risk. The PRA’s decisions in this section are relevant to firms using the SA, the foundation IRB (FIRB) approach and the slotting approach, as well as the advanced IRB (AIRB) approach where exposure at default (EAD) modelling will be prohibited (see Chapter 3 – Credit risk – internal ratings based approach, EAD estimation). Where EAD can be modelled, the CFs set out in this section will also affect the EAD input floors.

2.36 The PRA received 31 responses to this section of CP16/22. The substantive issues raised by respondents are set out below.

Definition of ‘commitment’ for off-balance sheet items

2.37 The PRA proposed to define a commitment as ‘any off-balance sheet contractual arrangement that has been offered by the institution and accepted by the obligor, including to extend credit, purchase assets or issue off-balance sheet items (but which is not itself an issued off-balance sheet item)’.footnote [20]

2.38 The PRA received 14 responses on the proposed definition and related exemptions to it for certain contractual arrangements:

  • one respondents supported the proposed definition of commitment;
  • four respondents requested that the PRA clarify the proposed treatment of unadvised limits within the definition of commitment. Of these, three argued that the proposed definition of ‘conversion factor’ set out in HMT’s consultationfootnote [21] should align with the proposed definition of commitment;
  • six respondents argued that the PRA should exempt certain arrangements to corporates and SMEs from the definition of commitment, noting that this would align with a national discretion in the Basel 3.1 standards;footnote [22]
  • one respondent requested that, if the PRA did not exercise the national discretion, the PRA should consider only applying the definition of commitment and applicable conversion factors to exposures that were originated post the Basel 3.1 implementation date;
  • one respondent argued that firms should be able to delay the point at which they are subject to capital requirements for facilities that remain within the control of the firm for each drawing, and three respondents argued that the PRA should permit firms to delay the recognition of capitalisation for facilities that are subject to mandatory regulatory approvals, for example, in merger and acquisition (M&A) transactions, due to the need for approval from competition regulators;
  • two respondents argued for Bankers Association for Finance and Trade Master Risk Participation Agreements (MRPAs) to be exempted from the definition of commitment; and
  • six respondents suggested that there should be a consistent definition of commitment for the SA and the IRB approach.

2.39 Having considered the responses, the PRA has decided to maintain its proposed definition of commitment. However, it has clarified in its near-final amendments to SS10/13 that firms are only expected to apply CFs to advised limits that have been contractually agreed between the firm and the obligor when calculating capital requirements for commitments under the SA. The PRA notes that HMT has stated it will update the definition of CF, if it is to remain in legislation at the point the PRA’s final rules take effect (such that the amended definition also takes effect at this point).footnote [23] This would align with the PRA’s near-final amendments to SS10/13, and would be consistent across the SA and the IRB approach.

2.40 As part of its decision to maintain the proposed definition of commitment, the PRA has decided to maintain its proposal not to exercise the national discretion in the Basel 3.1 standards to exempt certain arrangements in relation to corporates and SMEs. The PRA acknowledges that such arrangements may differ from other types of unconditionally cancellable commitments (UCCs) due to the need for a further credit assessment before the obligor can execute a drawdown. However, the PRA did not receive any quantitative evidence that such arrangements have a different risk profile from other unconditionally cancellable commitments and considers that applying zero capital requirements for such arrangements would be inconsistent with their risk. Firms are unable to perfectly predict the likelihood of obligor default when further credit checks are undertaken so there is a non-zero risk that an obligor passes such checks before defaulting after the commitment has been drawn down. Firms may also have an incentive to extend credit, regardless of the obligor’s creditworthiness, for reputational or business reasons.

2.41 The PRA has decided not to exempt any exposures from, or introduce any transitional arrangements in relation to, the proposed definition of commitment. As set out above, the PRA considers that applying zero capital requirements for such arrangements would be inconsistent with their risk, and therefore, considers that the definition of commitment should be implemented on 1 January 2026.

2.42 The PRA has also decided not to delay the point of recognition of capitalisation for facilities that are subject to mandatory regulatory approvals or exempt MRPAs from the definition of commitment because there is a non-zero risk of such commitments moving on-balance sheet.

Conversion factors for commitments

Conversion factors for UCCs

2.43 The PRA proposed to align with the Basel 3.1 standards and increase the CF for UCCs from 0% under the current requirements to 10%, reflecting the non-zero risk that these commitments will move onto the balance sheet.

2.44 One respondent did not support the PRA’s proposal, arguing that the legal ability to cancel such commitments at any time and without notice warrants the application of a 0% CF.

2.45 Ten respondents argued that the PRA should introduce transitional arrangements to phase in the increase in the CF, with one respondent suggesting a five-year transitional period and eight suggesting a longer-term transitional period aligned with that which is expected to be introduced in the EU. Most respondents argued that not introducing transitional arrangements would harm the UK’s relative standing.

2.46 Having considered the responses, the PRA has decided to maintain its proposal to introduce a 10% CF for UCCs because the available evidence suggests that retaining the existing 0% CF would be inconsistent with the risk of such arrangements and could result in imprudent outcomes. Even though a commitment is unconditionally cancellable, firms sometimes fail to cancel a commitment and allow obligors to continue to draw down on existing facilities for a variety of reasons, including reputational or business reasons, which in turn prevents UCCs from being riskless. This is in line with internal analysis conducted by the BCBS during its development of the Basel 3.1 standards, and with the PRA’s assessment that there is little empirical support for application of 0% CFs.

2.47 The PRA has also decided to maintain its proposal not to introduce a transitional period for the introduction of the 10% CF for UCCs. The PRA acknowledges that this may have a small negative impact on international competitiveness for a limited period relative to the EU’s expected approach. However, the PRA considers that, in order to advance its primary objective, it is necessary to end the practice of zero capital requirements for UCCs without undue delay.

Conversion factors for ‘other commitments’

2.48 The PRA proposed to set a 50% CF for ‘other commitments’. It considered that the 40% CF in the Basel 3.1 standards was likely to be an under-estimation of the risk of these commitments moving on balance sheet based on evidence set out in the BCBS’s 2014 Consultation Paper and information available to it relating to realised conversion rates for UK mortgage offers, which would typically be allocated to the ‘other commitments’ category before the loan is drawn down.footnote [24]

2.49 The PRA received a significant number of responses:

  • three respondents opposed the proposed CF and suggested that the PRA should align with other jurisdictions and the Basel 3.1 standards. Another four respondents highlighted competitiveness concerns with the proposal. One respondent argued that the increase in capital requirements, relative to the CRR, for ‘other commitments’ that have a maturity up to and including one year would be disproportionate to the risk;
  • ten respondents did not oppose assigning a CF of 50% for UK residential mortgage commitments; however, these respondents opposed assigning a CF of 50% for non-UK residential mortgage commitments and wholesale commitments. Two respondents argued that assigning a CF of 50% for UK residential mortgage commitments and a CF of 40% for non-UK residential mortgage commitments and wholesale commitments would be a more risk-sensitive approach that would not be difficult to operationalise; and
  • one respondent opposed the proposal to not consider maturity in the determination of the CF.

2.50 Having considered the responses, including empirical evidence shared by respondents during the consultation period, the PRA has decided to amend its proposal. The near-final rules apply a 50% CF only to UK residential mortgage commitments. A 40% CF will apply to the remaining types of ‘other commitments’. The PRA considers that a 50% CF for UK residential mortgage commitments is justified based on data from UK firms. However, the evidence for other types of ‘other commitments’ is more nuanced, making it appropriate to apply the lower 40% CF prescribed in the Basel 3.1 standards.

Other issues relating to commitments

2.51 Other substantive responses to this section of CP16/22 raised a range of issues:

  • Commitments with certain drawdown: four respondents requested that the PRA clarify the concept and scope of ‘other commitments with certain drawdown’. The PRA has amended its draft rules to require that a 100% CF is applied to other commitments with similar economic substance to the types of commitments otherwise listed as receiving a 100% CF in its near-final rules. However, the PRA does not consider, in general, that real estate exposures would be classified as having similar economic substance to the types of commitments listed as receiving a 100% CF under its near-final rules.
  • Commitment to issue off-balance sheet items: the PRA proposed that, where a commitment is made to issue an off-balance sheet item, the lower of the CF applicable to the commitment and the CF applicable to the issued off-balance sheet item would apply. Three respondents argued that this approach would not correctly reflect the interaction between the probability of the off-balance sheet item being issued and the probability of it then moving on-balance sheet. The PRA has decided to maintain its proposal. Attempting to capture the probability of the off-balance sheet item being issued and then the probability of it moving on-balance sheet in a more risk-sensitive way would be challenging in practice and inconsistent with the degree of risk sensitivity envisaged by the Basel 3.1 standards under the SA.
  • Merchant acquiring activity: one respondent argued that the appropriate CF for contingent liabilities arising from merchant acquiring activity should be 20%. As noted above, the PRA has amended its draft rules such that ‘other commitments’ receive a 40% CF, and the PRA has decided not to make an exception for merchant acquiring activity. The PRA considers that a 40% CF more appropriately reflects the risk of such commitments than the 50% CF proposed in CP16/22 and that respondents did not provide persuasive evidence that a 20% CF would adequately reflect economic downturn conditions. This is aligned with the Basel 3.1 standards and the approach expected to be taken by other major jurisdictions.
  • Scope of a commitment: one respondent requested that the PRA clarify whether entire facility amounts will be subject to a CF, rather than just the available or pledged amount. The PRA considers that, under its near-final rules, the CFs will apply to the full advised limit contractually offered by the institution and accepted by the obligor.
  • Syndicated commitments: one respondent argued that the PRA should provide additional clarity on the treatment of syndicated commitments. The PRA considers its draft rules are clear and has decided not to make any amendments in this respect. The amount of such commitments will be the full amount that is contractually agreed between counterparties. The SA balances risk sensitivity with simplicity and proportionality and the PRA did not receive persuasive evidence to justify a bespoke treatment for syndicated commitments.
  • Transitional arrangements: one respondent argued that the PRA should introduce transitional arrangements to smooth the increase in capital requirements resulting from the introduction of the 50% CF for UK mortgage commitments. The PRA does not expect that the stock of existing exposures that will be impacted by this change is significant, and firms will have time to adjust to the impact of this change on the flow of new lending. The PRA considers that requiring the exposure value to be calculated in a prudent manner from the date of implementation is proportionate given the expected impact on capital requirements, and that it would not be appropriate to introduce a transitional period in this circumstance.

Issued off-balance sheet items

Other transaction-related contingent items

2.52 The PRA proposed to align with the Basel 3.1 standards and assign a 50% CF to other ‘transaction-related contingent items’, which includes guarantees, warranties, and standby letters of credit that do not have the characteristics of credit substitutes.

2.53 In addition, the PRA proposed to maintain the existing 20% CF for short-term self-liquidating trade letters of credit (with a maturity of less than one year) and clarify that a 50% CF would therefore only apply to self-liquidating trade letters of credit with a maturity of greater than one year.

2.54 The PRA received ten responses that opposed the proposal to assign a 50% CF to transaction-related contingent items. They argued that a 50% CF was not an accurate reflection of the risk for such items and that empirical evidence provided by the International Chamber of Commerce and Global Credit Data demonstrated that a 20% CF was more appropriate.

2.55 One respondent supported the proposal to maintain a 20% CF for short-term self-liquidating trade letters of credit with a maturity of less than one year. However, three respondents did not support the application of different CFs to self-liquidating trade letters of credits based on the maturity of the exposure and suggested that a 20% CF should be assigned irrespective of maturity. These respondents argued that if the PRA decided to maintain its proposal for exposures subject to the SA, it should introduce a 20% CF for all exposures subject to the FIRB approach because maturity is already explicitly captured in the IRB risk weight formula.

2.56 Having considered the responses, the PRA has amended its draft rules such that a 20% CF will be assigned to all ‘other transaction-related contingent items’ that do not have the character of credit substitutes (including trade letters of credit). The PRA agrees with respondents that the empirical evidence provided by the International Chamber of Commerce and Global Credit Data sufficiently addresses the PRA’s previous concerns that a 20% CF may not fully reflect the probability of a ‘trigger event’ occurring or the behaviour in downturn conditions.footnote [25] A CF of 20% therefore advances the PRA’s primary objective in a more proportionate manner.

2.57 The PRA notes that the effect of this change will be to reduce the existing 50% SA CF for certain ‘other transaction-related contingent items’ (such as for shipping guarantees, custom and tax bonds) under the CRR, while retaining the existing 20% SA CF in other cases (such as for warranties and performance bonds). The PRA considers that this approach is prudentially justified based on the evidence available to it, and notes that it may facilitate international competitiveness depending on the approach taken in other jurisdictions. A CF of 20% is lower than the 50% CF in the Basel 3.1 standards, but the PRA considers this to be warranted based on the data provided by firms during the consultation period.

Other issues relating to issued off-balance sheet items

2.58 Other substantive responses to this section of CP16/22 raised two technical issues:

  • Conversion factor tables: one respondent suggested that CFs for the SA and the FIRB approach should be set out in separate tables. The PRA has not amended its draft rules in this respect as the applicable CFs are the same for the SA and the FIRB approach.
  • Definition of trade finance: two respondents argued that the PRA should complement the list of specified ‘transaction-related contingent items’ with a broader definition of ‘trade finance’ related items which makes clear that items relating to the trade of both goods and services are included. The PRA considers that the list of ‘transaction-related contingent items’ is sufficiently exhaustive and notes that the near-final rules will not require the specified items to be connected to the exchange of goods (as opposed to services). The PRA has therefore decided not to refer to ‘trade finance’ in its near-final rules.

PRA objectives analysis

2.59 With the exception of the changes to the CF calibration for 'other commitments' and 'transaction-related contingent items', incorporating the amendments and clarifications described above, the PRA considers that its near-final rules and near-final amendments to SS10/13 remain materially aligned with the proposals in CP16/22 and therefore the PRA considers its analysis of its objectives and have regards for those areas in CP16/22 remains appropriate.

2.60 The PRA's decision to assign a CF of 20% for 'transaction-related contingent items' will result in a CF that is either the same or lower than the current CF applied under the CRR depending on the item. Having analysed the empirical evidence provided by respondents, the PRA considers this decision will result in risk sensitive CFs and so will advance its primary objective. Maintaining the proposed CFs would have resulted in capital requirements that are overly conservative, given the evidence available to the PRA.

2.61 The PRA also considers that its decision to align with the Basel 3.1 standards for the CF applied to ‘other commitments’ (excluding UK residential mortgage commitments) will result in the appropriate capitalisation for these off-balance sheet risks, consistent with its primary objective.

‘Have regards’ analysis

2.62 In developing these near-final rules, the PRA has had regard to its framework of regulatory principles and the matters to which it is required to have regard when proposing changes to CRR rules. The PRA considers its analysis of its 'have regards', as presented in CP16/22, remains appropriate, subject to the following updates:

1. Sustainable growth:

  • The PRA's decision to assign a 20% CF for 'transaction-related contingent items' will either maintain or reduce RWAs for these items compared to under the CRR, while still adequately capturing risk and therefore supporting sustainable growth.
  • The impact of the proposed 40% CF for ‘other commitments’ (excluding UK residential mortgage commitments) will vary depending on the exact nature of the commitment, increasing the CF for some commitments and decreasing it for others. However, the PRA still considers that the proposed treatment should result in adequate risk capture and therefore support sustainable growth.

2. Relative standing of the UK as a place for internationally active firms to operate and competitiveness:

  • The PRA's decision to apply a 20% CF for 'transaction-related contingent items' may have some positive implications from a competitiveness perspective, as 'transaction-related contingent items' will be assigned a lower CF than in some other jurisdictions that have implemented, or have proposed to implement, a 50% CF for transaction-related contingent items.
  • The PRA’s decision to align with the Basel 3.1 standards and apply a 40% CF for ‘other commitments’ with the exception of non-UK residential mortgage commitments should support the relative standing of the UK given that these include categories of wholesale commitments that are typically international in nature. The PRA recognises that the proposal to apply a 50% CF for UK residential mortgage commitments is somewhat higher than that set out in the Basel 3.1 standards and the CF applied by other jurisdictions. However, the PRA considers the potential impact on relative standing is likely to be limited given UK mortgage lending is concentrated within domestic lenders.

3. Relevant international standards:

  • The PRA considers that its amendments to the proposed CFs for ‘other commitments’ align with the Basel 3.1 standards. While the CF for UK residential mortgage commitments will be somewhat higher than that set out in the Basel 3.1 standards, the Basel 3.1 standards envisage that national supervisors may choose to set stricter requirements. The PRA considers that this is warranted based on data specific to the UK mortgage market.
  • The PRA recognises that the CF for ‘transaction-related contingent items’ is lower than that set out in the Basel 3.1 standards but considers that there is sufficient evidence that a 20% CF is appropriate and risk sensitive to justify taking a different approach for this category.

4. Finance for the real economy:

  • The PRA considers that the impact of the changes in CFs for 'transaction-related contingent items' and ‘other commitments’ will vary across products but will likely decrease RWAs for certain items compared to their treatment under the CRR. The PRA considers that the likely decrease in RWAs following the decisions made on the CFs for ‘transaction-related contingent items’ and some ‘other commitments’ will result in capital requirements that are more reflective of the risk of the exposures. This should facilitate firms’ willingness to provide these off-balance sheet items and commitments which may be positive for economic activity.

Exposures to central governments, central banks, regional governments, local authorities, PSEs, and MDBs

2.63 The PRA proposed to broadly retain the existing risk weight treatment for exposures to central governments, central banks, regional governments, local authorities, PSEs and MDBs, which it considered to be broadly aligned with the Basel 3.1 standards.

2.64 The PRA received eight responses to this section in CP16/22. The substantive issues raised by respondents are set out below.

Exposures to public sector entities

2.65 The PRA proposed that firms would be required to apply risk weights to exposures to PSEs that correspond to their external credit rating, or to apply a 100% risk weight where an external credit rating is not available. The PRA proposed to not treat any PSEs as exposures to the central government, regional governments or local authorities of the UK. The CRR contains a provision that permits firms to apply central government, regional government or local authority risk weights to exposures to UK PSEs where the PRA has identified that there is no difference in risk. However, given the PRA had not previously identified any UK PSEs meeting the criteria, the PRA did not consider that its proposals reflected a substantive policy change.

2.66 Seven respondents argued that exposures to certain PSEs should receive the same risk weight treatment as exposures to their central government and that this would align with a national discretion in the Basel 3.1 standards. Six respondents argued that the PRA’s proposal would put the UK at a competitive disadvantage relative to jurisdictions who exercise the Basel national discretion. One respondent argued that the PRA should designate a global list of PSEs that are eligible for the same risk weight treatment as the relevant central government. Two respondents requested that the PRA treat export credit agencies (ECAs) the same as the relevant central government where there is no difference in risk between the central government and the ECA. Two respondents acknowledged that firms could apply the risk weight of the central government to PSEs under the credit risk mitigation (CRM) framework but argued this would be a complex and operationally costly approach.

2.67 Having considered the responses, the PRA has decided to maintain its proposal that firms would not be permitted to apply central government risk weights to any PSE exposures. The PRA considers that PSEs (both in the UK and internationally) are typically riskier obligors than central governments, meaning applying the central government risk weight would be imprudent. The PRA did not receive persuasive evidence to support respondents’ claims in relation to any category of PSEs or any PSE exposures.

2.68 As noted by respondents, where an exposure to a PSE is subject to a guarantee that meets the CRM eligibility requirements, firms are permitted to recognise the guarantee and reduce capital requirements through the CRM framework.

Exposures to unrated central banks

2.69 The PRA proposed to retain the CRR treatment of central bank exposures, where risk weights depend on the external credit rating for rated central bank exposures and a 100% risk weight applies to unrated central bank exposures (except for third-country exposures denominated in the obligor’s local currency, which can be assigned a lower risk weight in certain circumstances and based partly on an equivalence assessment of each third country’s prudential regulation conducted by HMT).

2.70 Five respondents argued that the PRA should permit firms to use the relevant central government’s external credit rating to risk weight unrated central bank exposures. One respondent argued that a 100% risk weight for unrated central bank exposures is punitive and proposed that a 0% risk weight should apply where the central government qualifies for CQS 1. This respondent argued that central banks can generate seigniorage revenues and have low default rates.

2.71 Having considered the responses, the PRA acknowledges the arguments made by respondents and has decided to amend its draft rules to permit firms to apply a risk weight to an unrated central bank exposure that corresponds to the risk weight of the relevant central government. The PRA considers its decision to be consistent with its primary objective because it is appropriate to treat central banks and central governments equally due to their interdependency and similar risk profiles. The PRA considers that this approach broadly aligns with the Basel 3.1 standards.

Exposures to multilateral development banks

2.72 Other substantive responses to this section in CP16/22 focused on the treatment of exposures to MDBs. Four respondents argued that the PRA should set out the criteria by which it would determine MDBs that would qualify for the 0% risk weight. These respondents argued that firms should be permitted to apply to the PRA to amend the list where necessary. The PRA has decided not to publish criteria by which it would determine additional MDBs who would qualify for the 0% risk weight, but notes that it could consult to amend its rules if it considers that it would be appropriate to change the list in future.

Exposures to regional governments or local authorities

2.73 Alongside considering the responses to the proposals covered in this section, the PRA has decided to amend its draft rules to clarify that, where a credit rating is not available for an exposure to a regional government or local authority, and there is not a credit rating available for the central government, the exposure will be assigned a risk weight of 100%. The PRA considers that this will support the consistent application of its near-final rules across firms.

PRA objectives analysis

2.74 With the exception of the change to the treatment of unrated central bank exposures, the PRA considers that the near-final rules regarding exposures to central governments, central banks, regional governments, local authorities, PSEs and MDBs remain materially aligned with the proposals in CP16/22. Therefore, the PRA considers its analysis of its objectives and have regards for those areas in CP16/22 remains appropriate.

2.75 The PRA considers that its near-final policy on unrated central bank exposures advances its primary objective. The approach introduces greater risk sensitivity reflecting the fact that the credit risk of a central bank and the central government are linked.

‘Have regards’ analysis

2.76 In developing these near-final rules, the PRA has had regard to its regulatory principles and the matters to which it is required to have regard when proposing changes to CRR rules. The PRA considers its analysis of its ‘have regards’, as presented in CP16/22, remains appropriate, subject to the following updates:

1. Relative standing of the UK as a place for internationally active firms to operate and competitiveness:

  • The PRA considers that its decision on the risk weighting of unrated central banks will support international competitiveness in cases where other jurisdictions apply a 100% risk weight to unrated central banks.

2. Relevant international standards:

  • The PRA considers its decision on the risk weighting of unrated central banks broadly aligns with the Basel 3.1 standards.

Exposures to institutions and covered bonds

2.77 The PRA proposed to make a number of changes to the treatment of exposures to institutions and covered bonds, including to:

  • adjust the approach for rated institutions, the ‘external credit rating approach’ (ECRA);
  • introduce a new approach for unrated institutions, the ‘standardised credit risk assessment approach’ (SCRA); and
  • change the risk weight treatment for eligible unrated covered bonds.

2.78 The PRA received 15 responses to its proposals for treatment of exposures to institutions and 7 responses relating to its proposals for the treatment of exposures to covered bonds.

Exposures to institutions

Short-term exposures

2.79 The PRA proposed to continue to permit firms to apply lower risk weights to short-term exposures relative to long-term exposures for rated and unrated institutions. Short-term exposures are currently defined under the CRR as having a residual maturity of less than three months. The PRA proposed that an original maturity of less than three months should instead be used for identifying short-term exposures. The PRA also proposed that exposures arising from the movement of goods across national borders with an original maturity of less than six months, would be risk weighted as short-term exposures.

2.80 The PRA proposed to remove the specific CRR risk weight treatment for exposures to institutions of a residual maturity of three months or less that are denominated and funded in the national currency of the obligor.

2.81 Four respondents argued that the proposed eligibility criteria for identifying short-term exposures were too restrictive. Seven respondents did not support the proposal to use original maturity rather than residual maturity, arguing that the proposed approach did not reflect the level of risk of exposures with residual maturity of less than three months. Four respondents argued that the treatment of exposures arising from the trade of services and the movement of goods within the UK should be aligned with the treatment of exposures arising from the movement of goods across national borders.

2.82 Having considered the responses, the PRA has decided to maintain its proposal for the definition of short-term exposures as set out in CP16/22. The PRA’s proposal aligns with the Basel 3.1 standards and the approach taken by other jurisdictions. The PRA considers that the original intention of the lower risk weight for shorter-maturity exposures in the Basel framework was to capture specific types of exposure, rather than generally assuming a lower credit risk associated with exposures with limited remaining maturity. This is consistent with the general principle in the Basel 3.1 standards that risk weights under the SA are not adjusted depending on an exposure’s residual maturity.

2.83 The PRA has decided however to amend its draft rules such that exposures arising from the movement of goods within the UK are treated in the same way as exposures arising from the movement of goods across national borders. The PRA recognises the issues raised by respondents and considers that a different treatment for the movement of goods within the UK is not justified on a risk basis.

2.84 Regarding the proposed treatment of exposures arising from the trade of services, the PRA has decided not to amend its draft rules and will not permit them to be risk weighted in the same way as exposures arising from the movement of goods. The delivery of services is different in nature to the movement of goods and a very broad range of exposures could be deemed to arise from the delivery of services. The PRA did not receive persuasive evidence that the risk weight treatment for short-term exposures should be extended and notes its approach is aligned with international standards and the approach of other jurisdictions. Exposures to institutions (including those arising from the trade of services) will continue to be risk-weighted as short-term exposures if they have an original maturity of less than three months.

Unrated institutions

2.85 The PRA proposed to introduce a new approach (SCRA) for exposures to institutions where no external credit rating is available. Under this approach, exposures to institutions would be categorised into one of three grades (A, B, or C) and assigned an associated risk weight, depending on the institution’s ability to meet or exceed published minimum regulatory requirements, and the firm’s internal assessment of the institution’s credit risk. The proposed treatment for exposures to unrated institutions would remove the link between the risk weighting of institutions and the credit rating of their sovereigns.

2.86 Several respondents argued that the PRA should provide further clarity on the application of the SCRA approach:

  • three respondents argued that the PRA should introduce more prescriptive criteria for determining the grade of an exposure;
  • one respondent argued that the PRA should encourage non-listed SDDTs to publish their CET1 ratio and their Tier 1 Leverage ratio data in their annual reports and accounts to help other firms apply the SCRA approach for exposures to them;
  • one respondent argued that the PRA should designate the jurisdictions which are considered to have implemented equivalent regulatory requirements and buffers;
  • one respondent argued that the PRA should permit firms to use third-party sources to carry out the SCRA approach;
  • one respondent requested that the PRA clarify whether firms would be permitted to reassign institutions between categories and how often they would be able to do so. Respondents argued that it would make it more challenging for small institutions to raise wholesale funding if they could not be reassigned to reflect a reduction in risk; and
  • one respondent requested that the PRA clarify whether firms would need to seek permission from the PRA to apply their assessment approach.

2.87 Having considered these responses, the PRA has decided not to amend its draft rules in relation to the SCRA for unrated institutions for the reasons set out below.

2.88 Regarding the level of prescription in the draft rules, the PRA considers that the approach is sufficiently detailed to support consistent application amongst firms and that firms should already have the capability to carry out a credit risk assessment as part of their broader risk management processes. However, the PRA will consider providing further clarification in future if evidence emerges of inconsistent application across firms.

2.89 The PRA is responsible for setting Pillar 3 disclosure requirements (see the Disclosure (CRR) Part of the PRA Rulebook). The form and content of financial statements is mandated under the Companies Act 2006, which does not fall under the remit of the PRA. In relation to the disclosure of non-listed SDDTs’ capital ratios, as set out in PS15/23 – The Strong and Simple Framework: Scope Criteria, Liquidity and Disclosure Requirements, the PRA will exclude non-listed SDDTs from the requirement to disclose a Pillar 3 report. However, some SDDTs without listed instruments may still wish to disclose (either publicly or directly to their lender) their prudential requirements.

2.90 While third-party sources of information may be used as an input for the application of the SCRA, the PRA considers it important that firms carry out this assessment for their own counterparties, alongside broader due diligence processes, rather than relying solely on third party sources.

2.91 Finally, the PRA clarifies that firms will not be required to apply for PRA permission to apply their assessment approach. The PRA also clarifies that its near-final rules do not restrict firms from re-categorising institutions into different grades where appropriate.

Other issues relating to exposures to institutions

2.92 One respondent argued that the proposed retention of a specific risk weight treatment for exposures to institutions in the form of minimum reserves required by the Bank of England should be amended to remove outdated references to a European Central Bank (ECB) regulation. Having considered this response, the PRA has decided to amend its draft rules to remove this treatment. The PRA does not consider there to be any applicable minimum reserve requirements currently set by the Bank of England that would qualify for the treatment and notes that direct exposures to the Bank of England that are denominated and funded in sterling will receive a 0% risk weight in any event.

2.93 The PRA also received responses on the use of external credit ratings for institutions. The PRA has provided feedback to these responses in the ‘External credit ratings and due diligence’ section of this chapter.

Exposures to eligible covered bonds

2.94 The PRA proposed to introduce changes to the risk weight treatment that applies to certain covered bonds. To be eligible for this treatment, covered bonds would be required to meet the definition of a ‘CRR covered bond’ (CRR covered bonds are, among other things, issued by a credit institution with its registered office in the UK) and be secured by exposures which meet the collateral eligibility criteria (including additional requirements for immovable property collateral).

2.95 As part of the above changes, the PRA proposed to introduce a more risk-sensitive mapping from the issuing institutions’ risk weights to unrated covered bonds risk weights. Consistent with the proposed treatment of institutions, the PRA proposed that the risk weight for unrated eligible covered bonds issued by rated institutions which have a credit rating corresponding to CQS 2 would be reduced from 20% to 15%, and that the risk weight for unrated eligible covered bonds issued by rated institutions which have a credit rating corresponding to CQS 3 would be increased from 20% to 25%. The PRA also proposed that the new due diligence requirements would apply to all covered bond exposures. In respect of eligible collateral, the PRA proposed to retain the requirement that exposures to residential and commercial mortgage loans would be required to meet the valuation requirements in Article 208 of the Credit Risk Mitigation (CRR) Part of its draft rules (including a requirement to revalue at default which was not required under the CRR). The PRA did not propose changes to the maturity assessment of short-term collateral.

2.96 Seven respondents provided comments on the proposed treatment of eligible covered bonds:

  • seven respondents argued that the eligibility criteria for covered bonds collateralised by exposures to institutions should depend upon residual maturity or argued all exposures to institutions that are CQS 2 (of any maturity) should be eligible collateral;
  • four respondents did not support the PRA’s proposal that the eligibility criteria for the lower risk weight treatment should be limited to those covered bonds which are issued by a credit institution with its registered office in the UK; and
  • two respondents argued that there should not be a requirement for real estate collateral used in eligible covered bond pools to be revalued at default, where the issuing institution applies the SA.

2.97 Having considered the responses, the PRA has decided to make a number of amendments to its draft rules, as set out below.

2.98 For covered bonds collateralised by exposures to institutions, the PRA has decided that exposures to institutions which have a credit rating corresponding to CQS 2 will be eligible covered bond collateral irrespective of maturity (notwithstanding any other eligibility requirements). This maintains an appropriate level of prudence and aligns with international standards.

2.99 The PRA has decided to maintain its proposal that only covered bonds which are issued by a credit institution with its registered office in the UK should be eligible covered bonds. It is important that only high-quality covered bonds qualify for lower risk weights, and the PRA considers that regulation in the UK underpins the quality of covered bonds issued by UK institutions.

2.100 The PRA has decided to remove its proposed requirement that real estate collateral in eligible covered bond pools be revalued at default. This was included in its draft rules as a result of a cross-reference to the foundation collateral method’s real estate valuation requirements. The PRA agrees with respondents that it is not necessary to maintain this requirement and that the amended approach aligns with the Basel 3.1 standards. The PRA has similarly excluded real estate covered bond collateral from the requirement for valuations to be adjusted to reflect prior charges, under Article 229 of the near-final Credit Risk Mitigation (CRR) Part, as the existing requirements on the use of real estate collateral in eligible covered bond pools already reflect the risks associated with prior charges.

PRA objectives and ‘have regards’ analysis

2.101 The PRA considers that the amendments set out above do not materially change the proposals in CP16/22 and therefore the PRA considers its analysis of its objectives and have regards in CP16/22 remains appropriate.

Exposures to corporates and specialised lending

2.102 The PRA proposed to enhance the risk sensitivity of the SA for calculating RWAs for rated and unrated corporate exposures. The PRA also proposed to align with the Basel 3.1 standards by not retaining the infrastructure support factor under the SA and introducing a specific treatment for specialised lending to align the SA more closely with the IRB approach.

2.103 The PRA received 27 responses to this section in the CP. The substantive issues raised are set out below.

Corporate exposures

Externally rated corporate exposures

2.104 The PRA proposed to continue to permit the use of external credit ratings for determining the RWAs applied to corporate exposures. Three respondents argued that the PRA should not allow the use of external credit ratings for the purpose of risk-weighting corporate exposures if the approach to unrated corporate exposures is not amended (see below). Having considered the responses, the PRA has decided to maintain its proposal to permit the use of external credit ratings for risk-weighting corporate exposures. The PRA considers this approach results in risk weights that better reflect the risk for rated exposures.

2.105 To align with the Basel 3.1 standards, the PRA also proposed to increase risk sensitivity by reducing the risk weight applicable to corporates that have a credit rating which corresponds to CQS 3 from 100% to 75%. One respondent explicitly supported the approach, noting that it would contribute to the increased risk sensitivity of the prudential framework and to a better allocation of firms’ capital requirements. The PRA has maintained the proposed approach in its near-final rules.

Unrated corporate exposures

2.106 The PRA proposed to introduce a more risk-sensitive approach relative to the CRR and the Basel 3.1 standards for unrated (non-SME) corporate exposures, to better reflect the underlying risk of different unrated corporate exposures. The proposal aimed to maintain an aggregate level of RWAs which was broadly consistent with the calibration under the Basel 3.1 standards – firms would apply one of two possible approaches to all of their unrated corporate exposures: (i) a risk-sensitive approach, available through a permission granted by the PRA, that would permit firms to assign a 65% risk weight to counterparties deemed ‘investment grade’ (IG) and a 135% risk weight to counterparties deemed ‘non-investment grade’ (non-IG); or (ii) a risk-neutral approach of a 100% risk weight applied to all unrated corporate exposures. The PRA asked respondents to provide quantitative and qualitative evidence to support responses, particularly in respect of how appropriate the proposed 135% risk weight for non-IG exposures would be under the risk-sensitive approach.

2.107 The PRA received 21 responses:

  • four respondents supported the proposal to give firms the option to choose between the two approaches. However, five respondents argued that firms with more IG exposures would select the risk-sensitive approach which would make them uncompetitive in non-IG lending, with firms using the risk-neutral approach being uncompetitive in lending to IG corporates and therefore being concentrated in non-IG lending;
  • the PRA received thirteen responses relating to the calibration of the risk weights assigned under the risk-sensitive approach. Seven respondents argued the proposed calibration of the risk weight for non-IG unrated corporates was too high. Two respondents supported the proposal to reduce the applicable risk weight for unrated corporates deemed IG to 65%, but two respondents argued that the proposed risk weight was too punitive for the lowest risk exposures. Two respondents argued that the risk weight for IG exposures should be increased and the risk weight for non-IG exposures should be decreased to narrow the differential between IG and non-IG exposures;
  • three respondents argued that the proposed calibration of the risk weight for non-IG unrated corporates would have a negative impact on the competitiveness of the UK relative to jurisdictions that assign a risk weight of 100%, and could cause the UK branches of internationally domiciled banks to become concentrated in non-IG corporate lending; and
  • four respondents argued that the PRA should provide guidance on the criteria for classifying entities as IG or non-IG. Two respondents supported the PRA’s proposal to not fully align with the definition of IG in the Basel 3.1 standards for jurisdictions that do not permit the use of external ratings for regulatory purposes. Respondents were particularly supportive of the PRA’s proposal not to implement the requirement that entities or parents be listed on a recognised exchange.

2.108 Having considered the responses, the PRA has decided to maintain its proposed approach. Firms will be permitted to use the risk-neutral approach or to use the risk-sensitive approach (subject to permission from the PRA). Taking into account the range of responses received, the PRA considers that its near-final rules strike an appropriate balance between increasing risk sensitivity while maintaining the simplicity and proportionality of the SA.

2.109 The PRA acknowledges that the treatment for unrated corporate exposures may result in a reallocation of exposures between lenders, depending on the risk-weighting approach they use. However, firms will have discretion to choose which approach to use based on their business model, broader risk-management approach and operational capabilities (subject to PRA permission to use the risk-sensitive approach).

2.110 The implications of the two approaches for lending to different types of corporates is nuanced, including for firms with IRB permissions who will have to apply the output floor. The provision of finance to unrated corporates is also dependent on factors other than capital requirements, such as business relationships, locality, and sector expertise.

2.111 The PRA recognises the importance of ensuring that the risk-sensitive approach is calibrated to apply prudent but proportionate risk weights to IG and non-IG exposures, given the wide range of unrated corporate exposures held by firms. The risk weights have been set so that they broadly align with the overall calibration of the Basel 3.1 standards, while introducing risk sensitivity to better reflect the range of credit risk of unrated corporate exposures. In finalising its proposals, the PRA has been conscious that SA risk weights are calibrated with the aim of ensuring that, on aggregate, risks are adequately capitalised across a range of exposures and so it should not be expected that the SA is complex enough to result in risk weights tailored to every individual exposure.

2.112 The PRA had significant engagement with industry on this topic, both before and after the publication of CP16/22. Although several respondents argued that the proposed risk weights were calibrated too conservatively, the PRA did not receive persuasive evidence, either from domestic or international experience, that supported these assertions. The data provided by respondents suggest that any lowering of the IG or non-IG risk weights would result in risk weights that no longer appropriately align with the with the overall calibration of the Basel 3.1 standards. As noted above, the risk weights assigned under the SA must cover a broad range of risk in a relatively simple and proportionate manner. This means that the risk weights assigned to exposures to some unrated IG corporates may be higher than indicated by comparators such as modelled risk weights or the risk weights assigned under the SA for exposures to rated corporates, but for others the risk weights may be lower than indicated by these comparators. The situation is similar for exposures to some non-IG corporates, and the calibration of both IG and non-IG risk weights seeks to ensure adequate capitalisation across a range of exposures. Overall, the PRA considers the 65% and 135% risk weights to be broadly appropriate for unrated IG and non-IG corporate exposures respectively.

2.113 Regarding the definition of IG, the PRA has decided to not publish further expectations on how firms should assess if an exposure to a corporate entity shall be deemed IG. The PRA considers that the definition of IG stated in the near-final rules is sufficiently clear. However, the PRA notes that it could choose to set further expectations or guidance if evidence of inconsistent application amongst firms emerges.

Exposures to unrated funds and special purpose vehicles (SPVs)

2.114 A number of respondents argued that the PRA should create a tailored approach for specific types of unrated corporate exposures to better reflect their risk. Three respondents suggested that the PRA create a specific treatment for SPV structures, where firms use SPVs to repackage securities to provide a return on the underlying collateral to investors. Five respondents suggested that the PRA create a bespoke treatment for exposures to unrated funds given the difference in the nature of these exposures and their risk compared to other corporates.

2.115 Having considered the responses, the PRA has decided to not introduce any bespoke treatments for funds or SPVs that repackage securities. The PRA acknowledges that exposures to funds and SPVs that repackage securities may have different risk characteristics compared to other corporate exposures. However, the level of risk will vary materially - funds and SPVs are not homogenous and different types of exposures will have different levels of credit risk. The PRA considers that it is important to maintain the simplicity and proportionality of the SA and notes that the Basel 3.1 standards similarly do not attempt to provide a differentiated approach. Given these are cross-border markets, if there were to be a different, more granular approach to risk weighting such exposures, the PRA considers it should be an agreed, internationally consistent approach.

Asset secured lending

2.116 Other responses to this section of CP16/22 focused on one technical point relating to asset secured lending: two respondents argued that the PRA should reconsider its proposals on the SA risk weighting of exposures secured by non-real estate collateral (‘asset secured lending’). They argued such exposures are lower risk than unsecured lending given the existence of collateral.

2.117 Having considered the responses, the PRA has decided to not apply a specific risk weight approach for non-real estate asset secured lending under the SA. Given the potential diversity of non-financial collateral, permitting its recognition in a risk-sensitive way would introduce a level of complexity that the PRA considers would not be proportionate or appropriate for the SA. The PRA also notes that the Basel 3.1 standards do not permit recognition of non-financial collateral for firms using the SA, with the exception of real estate collateral, which is recognised directly in risk weights.

Specialised lending exposures and the infrastructure support factor

2.118 Aligned with the Basel 3.1 standards, the PRA proposed to introduce a new specialised lending exposure subclass, which included specific risk weight treatments for commodities finance, object finance and project finance, and more closely aligned the SA with the IRB approach. As part of introducing the Basel 3.1 standards’ approaches, the PRA proposed to remove the infrastructure support factor.

2.119 The PRA received 14 responses to this section in CP16/22. The substantive issues raised are set out below.

Project finance and the infrastructure support factor

2.120 The PRA proposed to remove the infrastructure support factor. The infrastructure support factor applies to project finance exposures and other corporate exposures. The PRA has presented its feedback to the responses to its proposal to remove the infrastructure support factor alongside feedback to the responses to its proposed treatment of project finance exposures given significant overlap between project finance and infrastructure lending.

2.121 For unrated project finance exposures, the PRA proposed to differentiate between exposures to projects in the pre-operational phase versus the operational phase, in line with the Basel 3.1 standards. The PRA proposed that an unrated exposure in the operational phase would be risk-weighted at 100% and that an unrated exposure in the pre-operational phase would be risk-weighted at 130%.

2.122 The PRA also proposed to introduce greater risk sensitivity through a lower risk weight for unrated ‘high-quality’ project finance exposures that are in the operational phase. The PRA proposed that exposures that meet the ‘high-quality’ criteria would be risk-weighted at 80%. The PRA considered the proposed high-quality project finance approach would result in a broadly similar risk weight adjustment as provided by the infrastructure support factor for eligible exposures. It therefore considered that maintaining the infrastructure support factor in addition to the 80% risk weight for high-quality project finance would result in an imprudent duplication of risk weight adjustments for project finance exposures in Pillar 1 capital requirements.

2.123 The PRA received ten responses to its proposed treatment for project finance exposures and the proposed removal of the infrastructure support factor. Respondents were generally supportive of the introduction of the new high-quality project finance treatment. Six respondents opposed the removal of the infrastructure support factor, three neither opposed nor supported the removal, and one supported the removal. Respondents made the following arguments for retaining the infrastructure support factor:

  • six respondents argued that removing the infrastructure support factor would negatively impact the international competitiveness of the UK;
  • five respondents argued that the infrastructure support factor had increased the amount of lending to infrastructure projects and that removing it would increase the price of lending and decrease lending volumes;
  • three respondents argued that the introduction of a new high-quality project finance treatment would not fully offset the impact of removing the infrastructure support factor as lending by firms is predominantly to projects in the pre-operational phase and the high-quality risk weight only applies to projects in the operational phase;
  • four respondents argued that if the PRA removed the infrastructure support factor, transitional arrangements should be put in place in order to limit the impact on existing exposures;
  • six respondents argued that the PRA’s proposal would impact the financing of green infrastructure projects such as the development and operation of renewable energy sources. One respondent requested that alternative measures be put in place to support green and sustainable financing if the support factor is removed; and
  • one respondent requested that the proposed draft rules align with the ‘Solvency II Corporate Infrastructure Exemption’.

2.124 Having considered these responses, the PRA has decided to make material amendments to its draft policy. It has decided to:

  • amend its draft rules to maintain the proposed high-quality project finance 80% risk weight, but widen the scope of eligible entities that the counterparty is depending on for the revenue to include central banks, and certain international organisations and MDBs;
  • retain its proposal to risk-weight unrated project finance exposures at 100% where the project is in the operational phase and at 130% where the project is in the pre-operational phase; and
  • maintain its proposed removal of the infrastructure support factor in Pillar 1, but apply a firm-specific structural adjustment to reduce Pillar 2A capital requirements (the ‘infrastructure lending adjustment’) to ensure that the removal of the support factor does not result in an increase in overall capital requirements for infrastructure exposures.

2.125 Given the infrastructure support factor applies to exposures risk-weighted under both the SA and IRB approach, the PRA has considered the treatment of infrastructure lending under both approaches collectively. Although not part of the SA, the PRA is making a further change to lower risk weights for firms using the IRB slotting approach by introducing a new substantially stronger category in the slotting approach for project finance exposures. This is covered in Chapter 3 – Credit risk – internal ratings based approach, but is cross referenced here as it is an important change for the overall treatment of infrastructure lending. Taken together with the changes to the SA outlined above, the PRA considers that its near-final rules result in an overall Pillar 1 treatment for project finance exposures that is risk based and results in risk weights that are neither too low nor too high.

2.126 The new risk weights assigned under the SA introduce risk sensitivity in a manner that promotes consistency of application between firms. Some firms have not applied the infrastructure support factor to date because of the difficulty in assessing its complex criteria. The high-quality project finance treatment is simpler to apply than the infrastructure support factor and the PRA therefore expects this treatment to be applied more widely by firms.

2.127 Recognising the limited evidence base for the calibration of risk weights for lending to infrastructure, the PRA requested that respondents submit quantitative or qualitative evidence on this topic. The PRA did not receive persuasive evidence, based purely on the risk of the lending, to support retaining the infrastructure support factor in Pillar 1 in addition to the proposed changes in the underlying risk weights.

2.128 The PRA examined whether the proposed risk weights for exposures to projects in the pre-operational phase should be lower than those in the Basel 3.1 standards, given these exposures may have been eligible for the infrastructure support factor but will be ineligible for the ‘high-quality’ project finance risk weight. The PRA conducted a two-part analysis considering:

(a) the relative riskiness of project finance exposures in the pre-operational phase with corporate exposures more generally; and

(b) the relative riskiness of project finance exposures in the pre-operational and operational phases.

2.129 The PRA used Moody’s data on default rates and recovery rates for corporate bonds and project finance loans to inform both parts of the analysis and considered probability of default (PD) and loss-given default (LGD) effects separately in each case.

2.130 In the first part of the analysis, the PRA used stressed default rate data for corporate bonds rated Ba and B (which are subject to RWs of 100% and 150% respectively under the SA) as a comparator for project finance exposures in the pre-operational phase in order to analyse PD effects.footnote [26],footnote [27] Due to the limited data available, the PRA used projects in years 1 and 2 as a proxy for pre-operational phase projects.

Table 1: Default rates

Project Finance (Years 1-2)

Corporate bonds rated Ba

Corporate bonds rated B

Peak default rate in early 1990s

c. 7.0%

3.84%

13.12%

Peak default rate in early 2000s

c. 7.2%

1.77%

4.57%

2.131 This analysis is sensitive to significant variations in peak corporate default rates in different economic scenarios, but, given that risk weights should reflect plausible downturn scenarios, suggests the relative riskiness of project finance exposures that are in the pre-operational phase should sit between the risk associated with corporate bonds rated Ba and B, which are assigned risk weights of 100% and 150% respectively under the SA. The stressed default rates for project finance exposures are significantly higher in both periods than for Ba rated bonds which are assigned a risk weight of 100% under the SA.

2.132 In respect of LGD effects, the PRA compared recovery rates for project finance loans in the construction phase (which the PRA took to be a proxy for the pre-operational phase) with recovery rates for corporate loans and bonds.

Table 2: Implied LGD from recovery ratesfootnote [28]

Project Finance –Construction phase

Corporate Loansfootnote [29]

Senior Unsecured Bondsfootnote [30]

Implied LGD

27.5%

16.9%

52.6%

2.133 The data show that project finance loans in the construction phase have higher implied LGDs than corporate loans but lower implied LGDs than unsecured corporate bonds. However, additional data for project finance distressed sales recovery rates (which was not disaggregated by construction phase) show higher implied LGDs. Given that the risk weights assigned to unrated corporate exposures should reflect a mix of loans and bonds, and project finance risk weights should reflect a mix of recoveries from retaining the assets and distressed asset sales, the PRA considers there is insufficient evidence to support a downward adjustment of risk weights for project finance exposures based on LGD effects.

2.134 In the second part of the analysis the PRA compared PDs for projects in years 1-2 (which the PRA took as a proxy for the pre-operational phase) with the PDs for projects in year 5 (which the PRA took as a proxy for the operational phase). Moody’s data suggests that projects in years 1-2 have a default rate of around 1% which is around twice as high as the default rate for projects in year 5 (around 0.5%). Using the IRB risk weight formula to infer PDs under stressed conditions, the PRA estimated that there should be an approximate 40% uplift in the risk weight for project finance exposures in the pre-operational phase relative to the operational phase due to the differences in PD (the uplift would be higher still if data beyond year 5 had been considered).

2.135 In relation to LGD effects, the best available evidence shows that the average implied LGD for construction phase projects (which the PRA took as a proxy for the pre-operational phase) is approximately 43% higher than for operational phase projects (implying a 43% uplift in risk weights).footnote [31] Combining the PD and LGD effects, the available evidence suggests that pre-operational phase risk weights should be approximately 100% higher than operational phase risk weights. In comparison, the 130% risk weight the PRA has decided to introduce for project finance exposures in the pre-operational phase is only 30% higher than the 100% risk weight assigned to project finance exposures in the operational phase and only 63% higher than the 80% high-quality project finance risk weight for operational phase exposures.

2.136 There is a significant degree of uncertainty in both pieces of analysis. The uncertainty is consistent with the lack of data available publicly and lack of data provided by firms during the consultation period. However, taken together, the PRA considers that they indicate that the 130% risk weight in the Basel 3.1 standards for project finance exposures in the pre-operational phase is broadly appropriate and that introducing a lower Pillar 1 risk weight for these exposures (including for ‘high-quality’ projects without increasing the risk weight for projects that did not meet this criteria) would not be consistent with the underlying risk.

2.137 The PRA has however amended its draft rules to improve risk sensitivity where this can be done in a manner that is consistent with its primary objective. In particular, for the purpose of identifying high-quality project finance exposures, the PRA will widen the list of eligible main entities that the counterparty’s revenue is dependent on to include international organisations that are prescribed a risk weight of 0%, MDBs that are prescribed a risk weight of 0%, and central banks with a risk weight of 80% or lower. The PRA considers that it is risk sensitive to treat these counterparties in a similar manner to the eligible counterparties included in the PRA’s original proposals, given they lead to the exposure having a similar level of risk. This change further aligns the high-quality project finance counterparty eligibility criteria with the criteria for the infrastructure support factor under the existing CRR.

2.138 The PRA continues to consider that the infrastructure support factor should be removed from Pillar 1 capital requirements based purely on its assessment of the risk of eligible exposures. The PRA also considers the materiality of the infrastructure support factor to be limited as only ten firms are currently applying the infrastructure support factor (across the SA and IRB approaches). Given the limited data available, and consistent with the lack of data provided by firms in response to the consultation, it is also difficult to assess how the existing stock of loans currently benefitting from the infrastructure support factor will be classified within the corporate exposure class. Taking into account the above points, and the amendments to the draft rules described above (including the important changes set out in Chapter 3 – Credit risk – internal ratings based approach which will reduce risk weights for certain project finance exposures under the slotting approach), the PRA considers that its near-final rules are unlikely to have a material impact on Pillar 1 capital requirements for infrastructure lending.

2.139 However, the PRA recognises the concerns raised by respondents on the potential impact on UK competitiveness and growth of even limited increases in capital requirements on infrastructure lending. To minimise any potential disruption to infrastructure lending and therefore growth resulting from the removal of the infrastructure support factor, the PRA will apply the infrastructure lending adjustment. This will ensure that the removal of the infrastructure support factor under Pillar 1 does not result in an increase in overall capital requirements for infrastructure exposures.

2.140 The PRA considers that the infrastructure lending adjustment, which will be applied where firms choose to submit the necessary data to the PRA, strikes an appropriate balance in delivering a prudent but proportionate approach, supporting UK competitiveness and growth. It will ensure that the Pillar 1 risk weights assigned to infrastructure exposures appropriately reflect their risk, supporting the PRA’s primary objective. The removal of the infrastructure support factor from Pillar 1 capital requirements also aligns with international standards. The infrastructure lending adjustment will minimise any potential disruption to infrastructure lending as there will be no increase in overall capital requirements for infrastructure exposures. This will support growth without conflicting with the PRA’s primary objective or undermining economic growth in the medium to long-term.

2.141 The PRA considers that the infrastructure lending adjustment, when taken alongside the broader package of changes being introduced, will result in an overall level of firm capitalisation that is consistent with its primary objective. The PRA considers that the package will therefore support firms’ ability to lend through the economic cycle and support UK competitiveness and growth.

2.142 The PRA will invite firms in due course to provide the necessary data to inform the calibration of the infrastructure lending adjustment as part of an off-cycle review of firm-specific Pillar 2 capital requirements (as set out in PS17/23 and see Chapter 6 – Pillar 2 for further information).footnote [32] Given the infrastructure lending adjustment will minimise any disruption to infrastructure lending from the removal of the infrastructure support factor, the PRA has decided there is no need to amend its draft rules to introduce any transitional arrangements for the removal of the infrastructure support factor from Pillar 1 capital requirements.

Project finance exposures with government and MDB support

2.143 Other substantive responses to this section of CP16/22 focused on the recognition of government and MDBs’ support for projects.

2.144 One respondent argued that it should be easier for projects with support from governments and MDBs to qualify for lower regulatory capital requirements. The PRA considers that the scope of the proposed criteria for high-quality project finance exposures is appropriate. Additionally, the CRM framework provides a way for firms to recognise CRM-eligible guarantees to lower capital requirements for these exposures where this is justified by the level of credit risk.

Object finance

2.145 The PRA proposed that object finance exposures would be assigned a risk weight of 100%, reflecting the broad universe of underlying assets with different risk profiles that fall in this asset class.

2.146 Five respondents argued that firms should be permitted to apply lower risk weights to ‘high-quality’ object finance exposures. Four of these respondents argued that the PRA should provide guidance on the criteria for an object finance exposure to be considered high-quality. Three respondents argued that available data on shipping and aviation finance supported a lower risk weight for associated object finance exposures.

2.147 Having considered the responses, the PRA has decided to maintain its proposed treatment for object finance exposures. The PRA did not receive persuasive evidence that supported firms’ assertions that a lower risk weight for certain object finance exposures was appropriate. As recognised by respondents, object finance collateral is idiosyncratic in nature and generally exhibits poor resilience to stress. The proposed risk weight applicable to object finance exposures reflects the full range of potential exposures and is calibrated to downturn scenarios.

Definition of specialised lending exposure

2.148 The PRA proposed to introduce a definition of specialised lending exposure that would capture exposures that met one of several stated criteria. Following the PRA’s review of responses and the PRA’s decisions relating to the IRB approach (see Chapter 3 – Credit risk – internal ratings based approach, IRB exposure classes and sub-classes), the PRA has decided to align the definition of specialised lending in the SA with the definition used in the IRB approach for project finance, object finance and commodities finance, such that all (rather than one) of the stated criteria must be met for an exposure to be classed as a specialised lending exposure. The PRA considers that there is strong merit in applying a consistent definition across the SA and IRB approach.

PRA objectives analysis

2.149 With the exception of the eligibility criteria for the high-quality project finance risk weight, the PRA’s near-final rules remain materially aligned with the proposals in CP16/22 and therefore the PRA considers its analysis of its objectives and have regards for those areas in CP16/22 remains appropriate.

2.150 The PRA considers that widening the eligibility criteria for exposures eligible for the lower high-quality project finance risk weight advances its primary objective in a proportionate and risk sensitive manner.

2.151 Furthermore, the PRA considers that its decision to proceed with the removal of the infrastructure support factor and to apply the infrastructure lending adjustment will advance its primary objective in a proportionate manner. The PRA’s decision will ensure that Pillar 1 capital requirements are prudent and risk-sensitive, minimise any disruption to infrastructure lending and thereby support UK growth, and maintain an appropriate overall level of capitalisation.

2.152 The PRA considers that its decision to apply the infrastructure lending adjustment is consistent with its secondary competition objective given it will be available to all firms with eligible exposures, regardless of the credit risk approach used.

‘Have regards’ analysis

2.153 In developing these near-final rules, the PRA has had regard to its framework regulatory principles and the matters to which it is required to have regard when proposing changes to CRR rules. The PRA considers its analysis of its ‘have regards’, as presented in CP16/22, remains appropriate, subject to the following updates:

1. Relevant international standards:

  • The PRA considers that although the wider eligibility criteria for high-quality project finance exposures are not fully aligned with the explicit criteria in the Basel 3.1 standards, they are consistent with the Basel 3.1 standards list of other eligible main entities that the counterparty’s revenue is dependent on, as the entities in question would be viewed as low risk elsewhere in the SA framework.

2. Relative standing of the UK as a place for internationally active firms to operate and competitiveness:

  • The PRA considers that the wider eligibility for lower risk weights for certain infrastructure exposures will support the competitiveness of UK firms relative to the CP proposals and further mitigate any perceived negative impact that the removal of the infrastructure support factor might have on the relative standing of the UK. The impact should also be seen in combination with the changes set out in Chapter 3 – Credit risk – internal ratings based approach which will reduce risk weights for certain project finance exposures under the slotting approach.
  • The PRA’s decision to apply the infrastructure lending adjustment will also support the competitiveness of UK firms by striking an appropriate balance between delivering prudent Pillar 1 capital requirements that align with international standards and maintaining an appropriate overall level of capitalisation.

3. Finance for the real economy:

  • The PRA considers that the near-final package of Pillar 1 capital requirements is unlikely to have a material impact on the availability and cost of infrastructure lending. However, the PRA considers that its decision to implement the infrastructure lending adjustment will minimise any potential disruption to infrastructure lending, and therefore the impact on growth, from the removal of the infrastructure support factor.

Exposures to individuals and small and medium-sized enterprises

2.154 Aligned with the Basel 3.1 standards, when capitalising lending to individuals and small businesses, the PRA proposed to:

  • remove the SME support factor but change risk weights for exposures to SMEs;
  • change the definition and treatment of ‘regulatory retail’ exposures (excluding real estate);
  • introduce a new treatment for ‘retail’ exposures with currency mismatch; and
  • maintain the existing requirement that exposures to SMEs that are secured by commercial real estate receive a risk weight that is no lower than 100%.

2.155 The PRA received 40 responses to its proposals for exposures to SMEs and 11 responses to its proposals for retail exposures, including lending to individuals. The main issues raised by respondents are set out below.

Exposures to small and medium-sized enterprises

2.156 The PRA proposed changes to the treatment of exposures to SMEs, reflecting the Basel 3.1 standards, including the introduction of a new ‘corporate SME’ exposure sub-class. The PRA proposed to retain the existing lower CRR risk weight of 75% for retail SME exposures, introduce a new lower risk weight of 85% for unrated corporate SME exposures, and introduce a lower risk weight treatment for ‘transactor’ regulatory retail exposures to SMEs of 45%. This was alongside changes proposed to the risk weight treatment for certain exposures secured by real estate, which will also apply to exposures to SMEs secured by real estate (but the PRA proposed to maintain the existing requirement that risk weights for exposures secured by commercial real estate shall be no lower than 100%).

2.157 Given the reductions in SME risk weights in the Basel 3.1 standards, relative to the risk weights for larger corporates, the PRA also proposed to remove the existing SME support factor as retaining it would result in an unwarranted ‘doubling-up’ of lower risk weight treatments in Pillar 1. The PRA proposed to base the eligibility for SME risk weights on the CRR definition of SME.

2.158 Respondents were generally supportive of lower risk weights for SMEs, relative to larger corporates, though ten respondents argued that the lower risk weights did not go far enough in offsetting the proposed removal of the SME support factor. Eighteen respondents argued that risk weights for SME exposures secured by real estate should be lower than for unsecured SME exposures, which was not the case under the PRA’s proposed risk weight treatment for commercial real estate (see 'Real estate exposures’ section of this chapter for further detail).

2.159 The PRA received 40 responses to its proposal to remove the SME support factor. The most common arguments against the proposal were that its removal:

  • would reduce lending to SMEs and increase the lending rates which SMEs pay to borrow, because changes in the cost of capital would be passed on to obligors;
  • was not justified on prudential grounds, due to a lack of evidence that the SME support factor leads to SME exposures being undercapitalised in a stress; and
  • would disproportionately affect smaller banks and building societies (including those who are seeking to grow and compete with larger banks) who are an important source of finance for SMEs.

2.160 These respondents argued that the PRA should retain the SME support factor. However, if the PRA proceeded with the proposal to remove it, 13 requested that it is phased out gradually, or that it is retained for exposures existing at implementation date (until final repayment of the exposures). Two respondents suggested removing the SME support factor and making compensatory changes to Pillar 2A as necessary (to address specific risks on a firm-by-firm basis) or alternatively retaining the SME support factor but offsetting some of its effect in Pillar 2A calculations.

2.161 Seven respondents noted the current definition of SME was difficult to apply consistently, given the complexity of assessing linked and parent firms. They requested the PRA aligns with the definition in the Basel 3.1 standards.

2.162 Having considered these responses, the PRA has decided to make significant changes across a range of areas to its draft policy. It will:

  • maintain its proposed removal of the SME support factor in Pillar 1, but apply a firm-specific structural adjustment under Pillar 2A (the ‘SME lending adjustment’) to ensure that the removal of the support factor does not result in an increase in overall capital requirements for SME exposures;
  • maintain its proposed unrated corporate SME risk weight of 85%, which is a reduction from the 100% risk weight which currently applies to unrated corporate SME exposures under the CRR, aligning with the Basel 3.1 standards;
  • maintain its proposed retail SME risk weight of 75% applicable to exposures to SMEs which meet the requirements to be treated as regulatory retail exposures, and its proposed 45% risk weight for ‘transactor’ exposures to retail SMEs;
  • amend its draft rules to remove the 100% risk weight floor for exposures to SMEs secured by commercial real estate, where repayment is not ‘materially dependent on cashflows from the property’ and where the exposure meets the ‘regulatory real estate’ definition. This will result in significantly lower risk weights for certain SME exposures secured by commercial real estate than under the existing CRR treatment and the CP16/22 proposals for lending secured on an SME’s own commercial premises (see the 'Real estate exposures’ section of this chapter for further detail); and
  • amend the draft rules to include a new simplified and expanded definition of SME to align with the Basel 3.1 standards and permit more exposures to benefit from the lower SME risk weights.

2.163 The PRA has decided to make the above changes to its proposals where respondents provided new evidence or highlighted implications that had not previously been considered in the PRA’s analysis. In light of the significant feedback related to the proposed removal of the SME support factor, the PRA has reviewed the appropriateness of the risk weighting of SME exposures in the round, including the proposed underlying risk weights, the treatment of secured versus unsecured SME exposures, and the definition of SME. The package of measures in Pillar 1 relevant to SME exposures has been assessed as a whole.

2.164 The PRA aims to adopt risk weights that are not too low to adequately capture the riskiness of a given class or sub-class of exposures but are also not higher than required to be commensurate with that risk. On the basis of the available evidence, the PRA considers that its near-final rules will capitalise SME exposures under Pillar 1 at broadly the right level and advance its primary objective. To inform its analysis, the PRA considered the riskiness of corporate and retail SMEs relative to larger (non-SME) corporate exposures in terms of idiosyncratic and systemic risk. Unrated corporate exposures are assigned a 100% risk weight under the risk-neutral approach and provide a natural benchmark against which to compare against.

2.165 There are two main factors to weigh up when comparing the appropriate risk weighting of lending to smaller corporates with lending to larger corporates. First, if lending to smaller corporates would typically be riskier than lending to larger corporates due to the risks specific to the individual corporate such as the corporate’s individual business model (higher idiosyncratic risk). Second, if lending to small corporates would be less risky because the performance of such exposures is less correlated with the economic cycle (lower systemic risk). Then the combined effect of these two factors (idiosyncratic and systemic risk) can be assessed.

2.166 For idiosyncratic risk, the PRA considered the following portfolio level data from the annual cyclical scenario (ACS) on corporate and retail SME exposures for firms with IRB permissions:footnote [33]

Table 3: Average probability of default (PD) and loss given default (LGD) for different obligor types

Average PD

Average LGD

% secured

Mid-to-large corporate

0.94%

39.45%

40.45%

Corporate SME

1.73%

31.87%

85.77%

Retail SME

2.65%

43.47%

34.64%

2.167 The data show that corporate SMEs have higher idiosyncratic default risk (PD) than non-SME corporates but have a lower average LGD. The lower LGDs are likely driven by firms seeking collateral more often when lending to smaller businesses compared to larger ones. Although the average PD is higher for corporate SMEs, analysis of these data using the relevant IRB formulae set out in the Basel 3.1 standards suggests that the effect of the higher PD and lower LGD for corporate SMEs (relative to non-SME corporates) broadly balance out when considering potential losses. For retail SMEs, the data indicate that these have higher idiosyncratic risk than non-SME corporates in respect of both PD and LGD.

2.168 The PRA has examined several sources of evidence to understand the systemic risk of lending to SMEs – ie whether the risk of SME lending is materially correlated with the economic cycle (as opposed to the risks inherent to a particular firm such as its own particular strengths and weakness). The PRA considered the asset value correlation (AVC) parameters used in the IRB formulae set out in the Basel 3.1 standards (see the ‘Calculation of RWAs and EL’ section of Chapter 3 – Credit risk – internal ratings based approach). These parameters indicate that both corporate and retail SME exposures have lower systemic risk than non-SME corporate exposures. For corporate SMEs, the AVC parameter is reduced by between 0 and 4 percentage points depending on size, and for retail SMEs, the AVC parameter is reduced by approximately 7.4 percentage points on average.

2.169 There is limited empirical data to validate these IRB parameters for lending to UK SMEs (although long-run data on business failure rates in the UK indicate that SME lending risk has increased in past downturns). Some academic studies into the systemic risk of SME lending in other jurisdictions find that the risk of SME lending is materially correlated with the economic cycle (relative to wider corporate lending), but several do not, and the results are likely to reflect factors related to the relevant jurisdictions. Therefore, while some of these studies may suggest the AVC parameters in the IRB formulae are insufficiently conservative, the PRA considers them to be the most appropriate benchmark available given they reflect the international consensus.

2.170 The PRA then analysed the reduction in risk weights that would be consistent with these differences in both idiosyncratic and systemic risk. The PRA’s analysis shows the following, when comparing to the 100% SA risk weight for unrated corporates:

  • for corporate SMEs, a risk weight of approximately 88% would be consistent with the differences in idiosyncratic and systemic risk. This is broadly in line with the 85% calibration in the Basel 3.1 standards; and
  • for retail SMEs, a risk weight of approximately 89% would be consistent with the differences in idiosyncratic and systemic risk. This is higher than the 75% calibration in the Basel 3.1 standards.

2.171 This analysis indicates that the effective discount (relative to the 100% risk weight for unrated corporates) incorporated in the risk weight for corporate SMEs under the Basel 3.1 standards is broadly appropriate. Meanwhile, the analysis indicates the effective discount incorporated in the risk weight for retail SMEs under the Basel 3.1 standards may be too high. However, this is a single piece of analysis, and the PRA does not consider that it, in itself, justifies deviating from the international consensus set out in the Basel 3.1 standards. But it supports the view that reducing risk weights in Pillar 1 for corporate or retail SMEs further than the Basel 3.1 standards already have would be inconsistent with an assessment purely based on the risk.

2.172 The PRA continues to consider that the SME support factor should be removed from Pillar 1 capital requirements based on its assessment of the risk of eligible exposures. This will also be consistent with Basel 3.1 standards. The PRA considers that the net effect of the near-final Pillar 1 package for SME exposures is unlikely to have a material impact on the availability and cost of SME lending. While the internationally published evidence is mixed, overall it indicates that the SME support factor has not had a significant impact on SME lending. The PRA considers, based on this evidence, that the SME support factor has not significantly increased the provision of financing to SMEs and so removing it will not significantly reduce lending to SMEs.

2.173 The PRA’s estimate of the overall impact of its near-final package for SME exposures (across both the SA and the IRB approach and including the changes to commercial real estate set out in the ‘Real estate exposures’ section of this chapter) is an eight percentage point increase in average Pillar 1 risk weights – from 56% to 64%.

2.174 The PRA has estimated the impact of the overall package on SME lending pricing using two approaches. The first uses data from a Bank of International Settlements repository of 83 empirical studies assessing the impact of previous changes in minimum regulatory capital requirements on lending pricing. The second uses the PRA’s own calculations, based on a model of how changes in capital requirements translate into changes in lending pricing. Taking an average across these two approaches would suggest that the change in risk weights for SME lending could lead to an increase in pricing of 0.14 percentage points (without reflecting the introduction of the new transactor exposure sub-class, which will permit a subset of exposures to qualify for a new, lower risk weight which in turn should reduce the overall pricing impact).

2.175 These estimates are uncertain, reflecting the many factors which influence the pricing and availability of lending of which the capital framework is only one. It is challenging to accurately estimate the impact of changes in capital requirements on pricing, including because there is a significant discretionary element to firms’ lending decisions which differs across firms and products, and which may change over time. The extent to which changes in capital requirements for specific types of exposure are passed on by firms to obligors is dependent on several factors, including firm strategy, market dynamics and the level of competition between firms. There is also uncertainty around these estimates due to the limited data available to assess how the existing stock of loans will be classified within the new classes and sub-classes of exposures.

2.176 Although the near-final package is expected to result in a limited increase to Pillar 1 capital requirements for finance to SMEs, the PRA considers this necessary for both the safety and soundness of firms and the provision of finance to SMEs through the economic cycle. Capital requirements should be calibrated appropriately, including for downturn conditions. This means they should not be too high, but they should not be too low either. The failure of Silicon Valley Bank in the US demonstrated how depositors and investors can focus on lender resilience during stress, and how bank failures in such circumstances can leave SME customers without access to finance. This is particularly important given the reliance of SMEs on bank lending relative to other forms of finance.

2.177 However, the PRA recognises concerns raised by respondents on the potential impact on UK competitiveness and growth of even limited increases in capital requirements on SME lending. To minimise any potential disruption to SME lending and therefore growth resulting from the removal of the SME support factor, the PRA will apply the SME lending adjustment to ensure that the removal of the SME support factor under Pillar 1 does not result in an increase in overall capital requirements for SME lending. The PRA considers that its near-final policy strikes an appropriate balance in delivering a prudent but proportionate approach.

2.178 The PRA considers that the SME lending adjustment, which will be applied where firms choose to submit the necessary data to the PRA, will minimise any potential disruption to SME lending and growth resulting from the removal of the SME support factor, without conflicting with the PRA’s primary objective or undermining economic growth in the medium to long-term. When taken alongside the broader package of changes being introduced, the approach will result in an overall level of firm capitalisation that is consistent with its primary objective and supports firms’ ability to lend through the economic cycle.

2.179 The PRA will invite firms in due course to provide the necessary data to inform the calibration of the SME lending adjustment as part of an off-cycle review of firm-specific Pillar 2 capital requirements (as set out in PS17/23 and see Chapter 6 – Pillar 2 for further information).footnote [34] Given the SME lending adjustment will minimise any disruption to SME lending from the removal of the SME support factor, the PRA has decided there is no need to amend its draft rules to introduce any transitional arrangements for the removal of the SME support factor from Pillar 1 capital requirements.

2.180 In addition to the above changes, the PRA is introducing a revised SME definition. The new definition will require firms to assess the turnover of entities on the basis of the approach to accounting consolidation taken in the obligor’s jurisdiction. This will enable firms to assess SME status more easily and eliminate the requirement for firms to undertake a bespoke assessment of linked firms beyond those linkages that are captured through accounting consolidation. This will lower the operational burden, improve consistency and comparability across firms, and increase the number of exposures which qualify as SMEs.

Treatment of regulatory retail exposures

Criteria for identifying regulatory retail exposures

2.181 The PRA proposed to update the criteria for identifying regulatory retail exposures and introduce three qualifying elements relating to the product type, exposure value and granularity of the exposure. In particular, the PRA proposed that the value of the exposure (either individually or when aggregated with all other retail exposures and including undrawn commitments and off-balance sheet items) to a single obligor or group of connected clients should not exceed £0.88 million, to align with the Basel 3.1 standards.

2.182 One respondent requested that the PRA clarify how firms would be required to calculate the aggregate exposure for the purpose of applying the exposure value criterion. The PRA also identified an unintended circular reference where the calculation of the exposure value criterion would have depended on whether an SME exposure met that criterion (as an exposure to an SME would only have been treated as a retail exposure if it had satisfied the criteria for regulatory retail exposures).

2.183 The PRA has decided to amend its draft rules to materially align with the current treatment under the CRR, such that the value criterion is based on the total amount owed to the firm, excluding all residential real estate exposures. The PRA notes that this has the following effects:

  • undrawn commitments will be excluded from the value criterion calculation; and
  • only residential real estate exposures will be excluded from the value criterion calculation (as opposed to all exposures which do not meet the definition of a retail exposure, including all real estate exposures, as proposed in CP16/22).

2.184 While this does not fully align with the Basel 3.1 standards, the PRA considers the approach to be appropriately prudent when considered alongside the other criteria for classifying retail exposures. The PRA further considers that its decision advances its secondary competition objective by more closely aligning the treatment between exposures subject to the SA and the IRB approach. It is also a more appropriate and prudent criterion for determining if commercial real estate exposures to SMEs should receive a 75% counterparty risk weight, than a criterion based solely on a firm’s aggregate retail exposures to the obligor (see the ‘Real estate exposures’ section of this chapter).

2.185 Other substantive responses to this section of CP16/22 focused on two issues:

  • Off-balance sheet items: one respondent argued that the proposal to include commitments to issue off-balance sheet items to SMEs in the product criterion, but to exclude off-balance sheet items that have been issued to SMEs, is inconsistent. The respondent argued that issued off-balance sheet items should also be included in the product criterion. The PRA agrees the proposed treatment was inconsistent. However, it considers most off-balance sheet items and commitments to purchase such items are not retail in nature, so has decided to amend its draft rules to also exclude commitments to issue off-balance sheet items from the definition of regulatory retail exposures.
  • Asset-backed lending: one respondent requested that the PRA clarify how firms would be required to classify their asset-backed lending portfolio, identifying that if a product was offered to a natural person, it would appear to not meet the regulatory retail definition as it would not be one of a range of similar exposures. The PRA confirms that, if an asset-backed exposure to a natural person is not one of a range of similar exposures, the exposure would not meet the criteria to be categorised as a regulatory retail exposure.

Transactors, non-transactors and other retail exposures

2.186 The PRA proposed to add granularity within the ‘retail’ exposure class, by breaking it down into three sub-exposure classes for the purpose of determining the appropriate risk weight: regulatory retail exposures that are ‘transactor exposures’footnote [35] subject to a 45% risk weight; regulatory retail exposures that are ‘non-transactor exposures’ subject to a 75% risk weight; and ‘other retail’ exposures subject to a 100% risk weight.

2.187 Respondents broadly supported of the introduction of the new, lower risk weight for transactor exposures, although one argued that the definition of retail exposures had been complicated by the introduction of the transactor category. Six respondents requested the PRA introduce a materiality threshold to assess whether exposures meet the conditions to qualify as a ‘transactor exposure’ to introduce proportionality.

2.188 Having considered the responses, the PRA has decided to maintain its proposal to introduce a specific risk weight treatment for ‘transactor exposures’. The PRA considers its decision to introduce the ‘transactor exposures’ category improves the risk sensitivity of retail risk weights and aligns with the Basel 3.1 standards. However, the PRA has clarified in its near-final amendments to SS10/13, that, where an exposure has breached the transactor conditions due to certain technical events, the exposure should still be classified as a ‘transactor exposure’ as such technical events do not affect the risk of such exposures. A broader materiality threshold (for example based on interest or charges applied to the obligor) would be inappropriate given late payments or drawdown of overdrafts affect the credit risk of such exposures, regardless of the length of the delay or period of drawdown.

Exposures with a currency mismatch

2.189 The PRA proposed to introduce a risk weight multiplier of 1.5x for unhedged foreign exchange retail and residential real estate exposures to individuals, subject to a maximum risk weight of 150%. The multiplier would be applied to exposures where a mismatch is identified (at origination and throughout the loan term) between the currency of the loan and that of the obligor’s main source of income.

2.190 The PRA also proposed an ‘alternative approach’ to identifying currency mismatches where information about an obligor’s main source of income cannot be verified for lending prior to Basel 3.1 implementation. The PRA proposed that firms applying the ‘alternative approach’ would instead use, as a proxy, any currency mismatch between the currency of the exposure and the domestic currency of the country of residence of the obligor.

2.191 Three respondents argued that the PRA should limit the scope of the mismatch multiplier to regulatory residential real estate exposures, and retail exposures to individuals in the form of instalment loans or non-revolving loans with pre-specified schedules of repayment, in line with some other international jurisdictions. Five respondents requested that asset-based lending to high-net-worth individuals should be excluded from the currency mismatch treatment, arguing that for this type of lending, the collateral reduces the risks associated with currency mismatch, and that the source of income of an obligor was not a part of the underwriting criteria. Therefore, they argued that applying the proposed requirements to identify if a currency mismatch exists for such exposures would represent a disproportionate burden.

2.192 Having considered the responses, the PRA has decided to maintain its proposal that the currency mismatch multiplier should apply to all residential real estate exposures and retail exposures to individuals. The PRA did not receive persuasive evidence that foreign exchange risks were reduced for specific types of retail lending. This also aligns with the Basel 3.1 standards. Consistent with this decision, the PRA has decided to amend its draft rules to include ‘other real estate exposures’ within the scope of the currency mismatch multiplier treatment to correct a drafting error. However, the PRA notes that its proposed definition of loan instalment amount may not be clear when assessing if revolving facilities are hedged. Therefore, the PRA has decided to amend its draft rules to clarify that firms should use the minimum instalment required under the revolving exposure contract, and that for partially drawn revolving facilities, firms should consider the facility as fully drawn when calculating loan instalments.

2.193 Having considered the responses, the PRA has decided not to change the scope of the currency mismatch multiplier by exempting asset-based lending to high-net-worth individuals. However, recognising the risk mitigation that assets held as collateral provide, the PRA has decided to amend its draft rules to permit certain assets to be considered as a hedge when determining if loans are subject to the currency mismatch multiplier. This treatment will be limited to pledged assets that are eligible collateral under the CRM financial collateral comprehensive method and will be subject to the application of relevant haircuts.

Other issues relating to the currency mismatch multiplier

2.194 Other substantive responses to this section of CP16/22 focused on the following issues:

  • Financial hedges: three respondents requested that the PRA clarify what constitutes a ‘financial hedge’ for the purposes of identifying when to apply the currency mismatch multiplier. Having considered the responses, the PRA has decided to not provide further clarification on the definition of a financial hedge. Firms are already required to consider what constitutes a financial hedge in other contexts, so providing additional detail for this aspect could lead to increased burdens on firms if it is inconsistent with their existing approaches.
  • Assessing an obligor’s source of income: four respondents noted that the PRA’s draft rules specify that firms should consider a currency mismatch between an obligors ‘source of income’ and the loan, while CP16/22 referred to an obligor’s ‘main source of income’. These respondents argued that the latter would be more appropriate. The PRA has decided to maintain its draft rules but has clarified in its near-final amendments to SS10/13 how it expects firms to apply them. Firms should evaluate foreign exchange risk between an obligors’ main source of income and a loan where this source of income constitutes a substantial proportion of the loan instalments. Where this is not the case, firms should consider other income sources to ensure that they have correctly identified material foreign exchange risks. This addresses the inconsistency highlighted by respondents and ensures firms are identifying foreign exchange risk in a proportionate manner.
  • Non-natural persons: one respondent argued that the treatment of currency mismatch for real estate exposures was inconsistent between natural persons and non-natural persons with the characteristics of an individual obligor. Having considered the response, the PRA has decided to amend its draft rules to specify that real estate exposures to SPVs with an individual guarantor must be treated as an exposure to a natural person for the purpose of applying the currency mismatch multiplier to residential real estate exposures. The PRA agrees with the respondent that this provides consistency between obligors who share similar characteristics and prevents arbitrage opportunities.
  • Assessment of currency mismatch post-origination: three respondents requested clarification on the assessment of a currency mismatch post-origination, arguing it should be the obligor’s responsibility to make firms aware of any material change in circumstances. Having considered the responses, the PRA has clarified in its near-final amendments to SS10/13 the circumstances where firms should re-assess currency mismatch.
  • Alternative approach (ongoing application): three respondents argued that the PRA should expand the alternative approach to apply to all exposures on an ongoing basis. The PRA does not consider it appropriate to permit the alternative approach to be used in respect of exposures originated on or after 1 January 2026. Firms should have adequate processes in place by this time to obtain the necessary information from obligors at origination.
  • Alternative approach (country of employer): three respondents argued that the PRA should permit the country of employer to be used as a proxy for assessing an obligor’s income under the alternative approach. Having considered the responses, the PRA has decided to widen the scope of the alternative approach to permit firms to use country of employment as a proxy. The PRA considers this to be a reasonable alternative to country of residence for identifying a currency mismatch under the alternative approach.

PRA objectives

2.195 With the exception of the introduction of a new SME definition, the PRA considers that its analysis of its objectives presented in CP16/22 remains appropriate.

2.196 The PRA considers that the new definition of SME in its near-final rules will advance its primary objective by improving the consistency of application by firms and therefore the comparability of risk weights across firms. In respect of the PRA’s secondary objective, the PRA considers that the revised definition of SME should be easier for all firms to apply, and therefore support smaller firms being able to compete with larger firms when seeking SME business.

2.197 Furthermore, the PRA considers that its decision to proceed with the removal of the SME support factor and to apply the SME lending adjustment to ensure this does not result in an increase in overall capital requirements will advance its primary objective in a more proportionate manner. The PRA’s decision will ensure that Pillar 1 capital requirements are prudent and risk-sensitive, minimise any disruption to SME lending and therefore growth, and maintain an appropriate overall level of capitalisation.

2.198 The PRA considers that its decision to apply the SME lending adjustment is consistent with its secondary competition objective given it will be available to all firms with eligible exposures, regardless of the credit risk approach used.

‘Have regards’ analysis

2.199 In developing these near-final rules, the PRA has had regard to its framework of regulatory principles and the matters to which it is required to have regard when proposing amendments to CRR rules. The PRA considers its analysis of its ‘have regards’, as presented in chapter 3 of CP16/22, remains appropriate, subject to the following updates:

1. Relevant international standards:

  • The PRA’s new definition of SME is consistent with international standards.

2. Relative standing of the UK as a place for internationally active firms to operate and competitiveness:

  • The PRA considers that the package of near-final rules for exposures to SMEs will support the relative standing of the UK as a place to operate by lowering capital requirements for exposures to SMEs relative to non-SME corporates where it is prudentially justified to do so.
  • The PRA’s decision to apply the SME lending adjustment will also support the competitiveness of UK firms by striking an appropriate balance between delivering prudent Pillar 1 capital requirements that align with international standards and maintaining an appropriate overall level of capitalisation.
  • The PRA considers that its decisions to revise the treatment of SME lending secured by commercial real estate (covered in the ‘Real estate exposures section of this chapter), and to introduce a new expanded definition of SME, will support the competitiveness of UK firms relative to the CP proposals and further mitigate any perceived negative impact that the removal of the SME support factor might have on the relative standing of the UK.

3. Finance for the real economy:

  • The PRA considers that the near-final package for Pillar 1 capital requirements is likely to support growth, with little effect on the availability and cost of SME lending. However, the PRA considers its decision to apply the SME lending adjustment should address concerns raised by respondents on the potential impact of even limited increases in capital requirements on SME lending. The SME lending adjustment will support growth and minimise potential disruption and any impact on the provision of SME lending, and therefore the impact on growth, from the removal of the SME support factor.

4. Proportionality:

  • The PRA considers that the definition of SME is more proportionate for firms to apply and will require fewer resources to operationalise relative to the current definition.
  • The PRA considers that its decision to permit transactor exposures to retain their status if the threshold is breached due to a technical event, to be proportionate. It ensures risk weights do not increase due to technical events that do not affect the risk of the exposure. Similarly, the PRA considers its decision to permit firms to use the least burdensome loan instalment amount when assessing if a revolving exposure is hedged, and the introduction of new proxy measures under the alternative approach, lowers the operational costs of implementing the currency mismatch multiplier.

5. Different business models:

  • The PRA considers its decision to permit assets to be considered as a hedge when assessing currency mismatch should result in an application of a currency mismatch multiplier that better reflects the risks of exposures for firms specialising in asset-based lending and lending to high-net-worth individuals.

Real estate exposures

2.200 The PRA proposed changes to the treatment of exposures secured on immovable property, including:

  • a more structured and granular exposure allocation and greater risk sensitivity of the treatment for real estate, including introducing a new treatment for exposures that are materially dependent on cash flows from property;
  • new requirements for determining the value of the property, including a requirement that the value of the property should be the value at origination; and
  • a specific treatment for ‘land acquisition, development and construction exposures’ (ADC) exposures.

2.201 The PRA received 58 responses to this section in CP16/22. The substantive issues raised by respondents are set out below.

Definition of residential real estate and commercial real estate

2.202 The PRA proposed to introduce a definition for residential real estate that the property predominantly has, or will have, the nature of a dwelling. The PRA proposed to include a list of property types that were excluded from the definition for residential real estate as the PRA considered the use of such property was not consistent with residential real estate. The list included care homes, purpose-built student accommodation, and property that was predominantly used for holiday lets.

2.203 Respondents generally argued that the proposed exclusion of holiday lets from the residential real estate definition was not reflective of the risk, as such exposures may be more akin to buy-to-let residential property, with nine respondents arguing that holiday lets should be considered residential property when the nature of the dwelling is like standard residential housing.

2.204 Two respondents requested that the PRA clarify the treatment of purpose-built student accommodation, and one respondent did not support the proposed exclusion of purpose-built student accommodation from the definition of residential real estate.

2.205 Having considered the responses, the PRA has decided to amend its draft rules to remove the provisions excluding specific types of property from the definition of residential real estate. While the feedback mainly pertained to holiday lets and purpose-built student accommodation, the PRA considers that the feedback is also applicable, in principle, to care homes. In the PRA’s near-final amendments to SS10/13, the PRA has introduced expectations that firms should treat property as residential real estate only where it could be used as a standard residential dwelling. In particular, the PRA has clarified that it considers it unlikely that holiday lets, care homes and purpose-built student accommodation would meet the definition of residential real estate unless the property is capable of being resold as a standard residential dwelling in the event of the obligor’s default.

2.206 The PRA has also decided to make further amendments to the definition of real estate exposures to align with the approach to ADC exposures set out below.

Other issues relating to the definition of residential real estate and commercial real estate

2.207 Other substantive responses to this section of CP16/22 focused on mixed-use properties. Two respondents requested that the PRA clarify its proposed treatment of mixed-use properties (ie where a single property is used for residential and commercial purposes).

2.208 Having considered the responses, the PRA has decided to retain the proposed predominant nature test for defining whether a property is residential real estate for operational simplicity. The PRA has decided to set a further expectation in its near-final amendments to SS10/13 that firms should have a clear documented policy for determining the predominant nature of a property for the purpose of applying the definition of residential real estate.

2.209 Where an exposure is secured by multiple properties, the PRA has clarified in its near-final amendments to SS10/13 that firms must assess the predominant nature of each property separately to determine if it is residential or commercial real estate. The approach to exposures that are secured by both residential and commercial real estate is set out below.

Requirements for ‘regulatory real estate’

2.210 The PRA proposed to introduce criteria for a loan to be classified as a ‘regulatory real estate exposure’ which are aligned with the Basel 3.1 standards, as it considered these to be proportionate and prudent. This included a proposed requirement that the exposure is secured on a finished property. In cases where the requirements are not met, the PRA proposed that the exposures would be classified as ‘other real estate’ and receive a higher risk weight, given the elevated risks associated with collateral that does not fully meet the regulatory real estate requirements. The PRA acknowledged that the proposed approach may affect self-build mortgage loans, as these would not meet the ‘finished property’ requirement during the construction phase, but considered, based on the evidence it had, that the greater risk of these exposures justified a higher risk weight.

2.211 The PRA received a significant number of responses to its proposal not to exempt self-build mortgage loans from the ‘finished property’ requirement. Seven respondents argued that the applicable risk weights for ‘other residential real estate’ would be too high for self-build property mortgages given what they considered to be their low-risk nature and the existing risk management practices of firms. Six respondents also noted that providing such an exemption would align with the Basel 3.1 standards, which allow discretion for national authorities to exempt self-build mortgage loans from the ‘finished property’ requirement. A number of respondents provided empirical data to support their arguments.

2.212 Having considered the responses and the data provided, the PRA agrees that classifying self-build mortgage loans as ‘other residential real estate’ could result in risk weights that are not proportionate to the level of risk associated with these exposures. The PRA has therefore amended its draft rules to exempt self-build mortgage loans from the ‘finished property’ requirement. If all other requirements are met, loans secured by unfinished self-build properties will be classed as ‘regulatory real estate exposures’, and use of the loan-splitting approach to risk weight these exposures will therefore be permitted. Specifically, the PRA has decided to exempt exposures from the ‘finished property’ requirement where the exposure is: (i) secured by property that has been acquired or held for development and construction purposes; (ii) the property does or will not have more than four residential housing units; and (iii) the property will be the obligor’s primary residence.

2.213 The valuation and market liquidity of unfinished self-build properties is subject to substantial uncertainty compared with finished properties, particularly in a downturn. To account for this, the PRA has decided to require that, for the purposes of determining the value of an unfinished self-build property for the purpose of risk-weighting the exposure, firms must use the higher of the most recent valuation with a 20% haircut applied, or the land value. The PRA considers that these changes in aggregate address respondents’ concerns and advance the PRA’s primary objective in a proportionate manner.

Other issues relating to the requirements for regulatory real estate

2.214 Other substantive responses to this section of CP16/22 focused on the issues set out below.

Regulatory real estate criteria – value of the property must not materially depend on the performance of the obligor

2.215 The PRA proposed that the value of the property must not materially depend on the performance of the obligor. The PRA has decided to amend its draft rules to clarify that a real estate exposure must meet this requirement in order to be classified as a regulatory real estate exposure, as opposed to this being a requirement on how a valuation is obtained. This more closely aligns with the Basel 3.1 standards.

Regulatory real estate criteria – second charges

2.216 One respondent requested that the PRA clarify whether a loan secured by a second charge mortgage can be deemed a ‘regulatory real estate exposure’. Three respondents argued that the PRA should include the additional clarification provided in the Basel 3.1 standards regarding the situation where there is an intermediate charge held by another lender.

2.217 Having considered these responses, the PRA clarifies that an exposure secured by a second charge on a property may be treated as a ‘regulatory real estate exposure’ if certain requirements are met (including where there is an intermediate charge held by another lender). The PRA has decided to amend its draft rules to more clearly reflect this approach.

Insurance against damage

2.218 One respondent argued that the PRA should retain the CRR requirement that firms must monitor that immovable property is adequately insured against the risk of damage. Having considered this response, the PRA agrees that this would promote safety and soundness so it has amended its draft rules to introduce a requirement that firms must have procedures in place to monitor that the immovable property is adequately insured against the risk of damage in order for the exposure to be classified as a ‘regulatory real estate exposure’.

Monoline insurance

2.219 Three respondents considered that the PRA should recognise monoline-insured loans as meeting the regulatory real estate requirement for charges over the property, where the specific requirements in the Basel 3.1 standards are met.footnote [36]

2.220 Having considered the responses, the PRA has decided not to recognise monoline-insured loans as meeting the regulatory real estate requirement for being secured by appropriate charges over the property. The PRA considers that there was insufficient evidence provided in responses to justify implementing a bespoke treatment for such exposures and notes a firm may be able to use eligible credit risk mitigation techniques to reflect the credit insurance.

Underwriting standards

2.221 Two respondents requested that the PRA provide further guidance on good underwriting practice. The PRA notes that buy-to-let exposures are already subject to specific expectations set out in SS13/16 and does not consider that further guidance or requirements for underwriting standards are necessary at this stage.

Valuation of real estate collateral

2.222 In CP16/22, the PRA proposed that the value for residential and commercial property should be the value at origination, which the PRA considered to be the valuation obtained by a firm when it issues a new mortgage loan for the purchase of the property, or when a firm issues a new mortgage loan to an existing or new obligor for the property securing the loan. The PRA proposed to permit some exceptions for firms to use updated valuations, in particular to require a new valuation of the property when: (i) an event occurs that results in a likely permanent reduction in the property’s value; (ii) there is a significant decrease in the market value of the property as a result of a broader decrease in market prices; or (iii) modifications are made to property that unequivocally increase its value.

2.223 The PRA also proposed that the valuation of a property would need to be appraised using prudently conservative valuation criteria, and that this valuation would need to be undertaken by a suitably qualified valuer who was independent from the institution’s mortgage acquisition, loan processing and loan decision process.

2.224 The PRA considered that the proposed approach achieved a balance between ensuring accurate valuations while mitigating the risk of excessive cyclicality in values that could flow through to risk weights.

2.225 The PRA received 47 responses. The material issues raised were:

  • Use of value at origination: some respondents explicitly supported parts of the proposals on valuing real estate collateral. In particular, three respondents supported permitting revaluations to be performed at the point a loan is refinanced and two respondents agreed with the proposal that the value should reflect the value at origination. However, 29 respondents did not support using value at origination. These respondents argued that it was not risk sensitive as it would not reflect the current market value, that it would be unsuitable for specific products with longer term fixed periods, and it would create inconsistency between firms using the SA and those using the IRB approach. Respondents also raised concerns over the operational complexity of maintaining two values for a property: one for risk management purposes reflecting current market value and a separate origination value for capital requirement purposes.
  • Revaluation events: ten respondents requested that the PRA clarify what constitutes a revaluation event and to publish a non-exhaustive list of examples of valid revaluation events.
  • Revaluation due to a broader decrease in market prices: ten respondents requested that the PRA clarify the proposed requirement to update the value of the property where there has been a significant decrease in its market value as a result of a broader decrease in market prices. Respondents suggested the PRA issue guidance on how to comply with the requirement, including the triggers that should be used to indicate a significant decrease in the market value of the property and when there had been a broader decrease in market prices. Eight respondents argued that if the value of the property was revised downwards due to a broader decrease in market prices, firms should be permitted to reflect subsequent price increases up to the previously recognised property value.
  • Valuation of existing exposures: two respondents agreed with the PRA’s proposal to permit firms to use the valuation obtained at the most recent revaluation event for exposures incurred prior to implementation date, where it is not reasonably practicable to use the origination value of the property. However, two respondents argued that it could lead to inconsistent approaches across firms, with some arguing it is unclear what is meant by the term ‘reasonably practical’ and one arguing that the requirement to use the origination value should only be applied to new exposures.
  • Prudent valuation criteria: 15 responses related to the proposed requirement that the valuation must be adjusted to consider the potential for the current market price to be significantly above the value that would be sustainable over the life of the loan. Respondents requested that the PRA clarify the proposed requirement, and some argued that the PRA should not introduce the requirement due to the complexity of adjusting valuations at individual property level. Two respondents supported the proposed implementation of a prudent valuation approach.
  • Use of automated valuation models (AVMs): 19 respondents argued that the PRA should explicitly permit AVMs to be used for property valuations, noting the trend towards increased levels of automation across the industry.

Value at origination

2.226 Having considered the responses, the PRA has decided to maintain the proposed requirement that the value of the property should be the value at origination but has made several important amendments to its draft rules.

2.227 The PRA continues to consider that using the value at origination strikes a good balance between ensuring an accurate valuation is used while mitigating potential cyclicality in values that could lead to excessive cyclicality in risk weights. However, the PRA acknowledges that, without compromising the advancement of its primary objective, changes are appropriate in order to make the valuation approach more product agnostic and make aspects of the approach easier for firms to operationalise.

2.228 The PRA has amended its draft rules to introduce a revaluation ‘backstop’ that will require firms to obtain an updated valuation every five years after the most recent valuation has been obtained. This addresses respondents’ concerns that using value at origination could impact firm and obligor incentives and behaviours in relation to the length of fixed-rate products and therefore not be product agnostic. The PRA has amended its draft rules to introduce two exceptions to this requirement:

  • where the amount of the loan is more than £2.6 million or 5% of the own funds of the firm, the firm must obtain an updated value of the property every three years. This reflects the greater risk of loss in the event of a default for such exposures if they are inaccurately valued and is consistent with the existing requirement for the valuation of such immovable property collateral; and
  • where the most recent valuation has been obtained due to a broader decrease in market prices (see below), a firm may either: (i) obtain an updated valuation either three or five years (as applicable) after this revaluation; or (ii) obtain an updated valuation either three or five years (as applicable) after the most recent origination event that occurred prior to this most recent valuation.

Revaluation due to a broader decrease in market prices

2.229 The PRA acknowledges the concerns raised by respondents relating to a lack of clarity on the trigger for firms to obtain an updated valuation where there has been a significant decrease in the market value of a property due to a broader decrease in market prices. The PRA has decided to amend its draft rules to retain the existing requirement that firms shall continue to monitor the market value of the property on a frequent basis and carry out more frequent monitoring where the market is subject to significant changes in conditions. In addition, the PRA has decided to amend its draft rules to only require that firms obtain an updated valuation if the firm estimates that the market value of the property has fallen more than 10% from the recorded value of the property at the last valuation event. The PRA considers that this simpler approach should promote the consistent application of its rules across firms. 

Valuation of existing exposures

2.230 The PRA has maintained its proposed policy for assessing origination valuation for exposures incurred prior to 1 January 2026. The PRA clarifies that the most recent available origination value, whether at the point of purchase of the property or a valuation at another origination event (including when an obligor remortgages) must be used. The PRA notes that the revaluation ‘backstop’ will mean firms will be required to update existing valuations that are older than five years (or three years if the exception above applies) on 1 January 2026. The PRA considers that most regulatory real estate exposures will have relatively recent origination values, with limited scope for inconsistent approaches between firms.

Prudent valuation criteria

2.231 The PRA agrees with respondents that the proposed requirement to adjust a valuation to reflect the value of the property that would be sustainable over the life of the loan could be complex for firms to operationalise and may lead to inconsistent approaches. Having considered the responses, the PRA has amended its draft rules to remove this requirement.

Use of automated valuation models

2.232 In response to comments related to the use of automated valuation models (AVMs), the PRA clarifies that it did not intend to propose that only physical valuations by a qualified surveyor would be permitted and has amended the draft rules to state that firms can use robust statistical valuation methods. This may include the use of AVMs or indices, where it is prudent to do so.

Definition of real estate where repayment is materially dependent on the cash flows generated by the property

2.233 The PRA proposed to implement criteria for determining whether repayment of the loan was ‘materially dependent on the cash flows generated by the property’ that were aligned with the Basel 3.1 standards, and proposed guidance in its draft amendments to SS10/13 to support consistent implementation across firms. It proposed that a real estate exposure would be considered materially dependent on the cash flows generated by the property where payment of the mortgage loan over a representative period (or the prospects of recovery in the event of default) materially depends on the cash flows generated by the property securing that exposure, rather than on the capacity of the obligor to pay the mortgage loan from the other sources.

2.234 In addition to the general criteria, the PRA proposed that houses in multiple occupation (HMOs) would automatically be treated as materially dependent on the cash flows generated by the property. The PRA also proposed exceptions to the definition of materially dependent on the cash flows generated by the property. These included exposures secured by a property that was the obligor’s primary residence and social housing exposures.footnote [37]

2.235 The PRA proposed to permit an exposure to an individual who has no more than three mortgaged residential properties in total, regardless of the firm that provided the loan on the other properties, to be exempt from being classified as materially dependent on the cash flows generated by the property (the ‘three property limit’). For the purposes of the three-property limit, the PRA proposed that:

  • properties of an individual would include properties held through structures where the individual was the ultimate beneficial owner; and
  • the three-property limit would not include the individual’s primary residence unless the individual depends on cash flows generated by their property portfolio to meet the mortgage payments on that primary residence.

2.236 The PRA proposed that if the three-property limit was exceeded by an individual, all residential real estate exposures to that individual would be treated as being materially dependent on the cash flows generated by the property (except for the obligor’s primary residence unless that property itself counts towards the three-property limit).

General criteria for loans to be materially dependent on the cash flows generated by the property

2.237 Apart from responses on the application of the three-property limit, which are covered in the following subsection, the PRA received responses on the following points:

  • General criteria for loans to be materially dependent on the cash flows generated by the property: nine respondents raised concerns relating to the general criteria for determining whether the exposure was materially dependent on the cash flows generated by the property. Respondents argued it would be operationally challenging to obtain data to assess whether the repayment of the loan was materially dependent on the cash flows generated by the property over a representative period of sufficient length to include a mix of good and bad years, and argued that the proposed requirement was not sufficiently clear.
  • High-net-worth individuals: one respondent argued that the PRA should exempt exposures to high-net-worth individuals from being determined as materially dependent on the cash flows generated by the property and provide guidance on how to account for individuals who own more than three properties but generate no rental income.
  • Houses in multiple occupation: six respondents commented on the proposed treatment of HMOs. Respondents challenged the automatic inclusion of HMOs in the definition of materially dependent on the cash flows generated by the property and some requested that the PRA clarify the interaction of this proposed requirement with the three-property limit.
  • Housing associations or cooperatives of individuals: two respondents argued that the PRA should exempt exposures secured by residential real estate property to associations or cooperative of individuals from the definition of materially dependent on the cash flows generated by the property, which would align with the Basel 3.1 standards.

2.238 Having considered the responses, the PRA has amended its draft rules to simplify and clarify the definition of materially dependent on the cash flows generated by the property. Reflecting this approach, the PRA has also removed the related expectations from its draft amendments to SS10/13.

2.239 The PRA has amended its draft rules to remove the general requirement to assess whether the payment of the mortgage loan over a representative period (or the prospects of recovery in the event of default) materially depends on the cash flows generated by the property securing that exposure, rather than on the capacity of the obligor to pay the mortgage loan from other sources.

2.240 For a residential real estate exposure, the PRA has amended its draft rules to require that the exposure must be treated as materially dependent on the cash flows generated by the property unless one of the following criteria is met:

  • the exposure is secured by a single property that is the obligor’s primary residence;
  • the exposure is to an individual or, in certain cases, to a special purpose entity where an individual is a guarantor, and does not exceed the three-property limit, assessed at a loan origination event (including refinancing);
  • the exposure is to a public housing company or not-for-profit association regulated in the UK that exists to serve social purposes and to offer tenants long-term housing; or
  • the exposure is to an association or a cooperative of individuals that exist with the sole purpose of granting its members the use of a primary residence in the property securing the loans.

2.241 These amendments will make it operationally easier for firms to assess whether exposures are materially dependent on the cash flows generated by the property, and should result in more consistent application of the near-final rules across firms, while advancing the PRA’s primary objective. The PRA acknowledges that the change may result in some exposures meeting the definition of materially dependent on the cash flows generated by the property which would not have met the definition proposed in CP16/22. However, the PRA considers that a simpler framework, which addresses respondents’ concerns relating to operational complexity, is a proportionate response and will result in a prudent framework that will improve the consistency and comparability of real estate RWAs. The PRA did not receive persuasive evidence that a narrower or more risk-sensitive definition of whether exposures are materially dependent on the cash flows generated by the property (including an exemption for high-net-worth individuals) would achieve this.

2.242 The PRA considers that its decision to permit certain exposures to special purpose entities to be treated as not being materially dependent on cash flows from the property should result in exposures of similar economic substance but different legal form being treated in a similar manner that better reflects the risk of the exposures.

2.243 The PRA has also decided to remove the explicit inclusion of HMOs from the materially dependent on the cash flows generated by the property definition. As a result, firms will be required to assess HMOs in the same way as other residential real estate exposures.

2.244 The PRA considers that it is appropriate to exempt exposures to housing associations or cooperatives from being determined as materially dependent on the cash flows generated by the property, given the similar nature of these exposures to social housing exposures (as reflected in the Basel 3.1 standards). Consistent with this judgement, the PRA has decided to implement a minimum counterparty risk weight of 75% for these exposures in a similar manner to the risk weighting of social housing exposures (see below).

2.245 For commercial real estate exposures, the PRA has also decided to amend its draft rules to require that an exposure is treated as materially dependent on the cash flows generated by the property unless each commercial real estate property that the exposure is secured on is predominantly used by the obligor for its own business purposes (eg where it is used as an office or a production facility by the business, rather than being leased out to generate income). For commercial real estate exposures, the PRA considers it appropriate that firms should update their assessments of the use of the property at least annually.

Calibration and monitoring of the three-property limit

2.246 The PRA received eighteen responses which argued that the proposed requirement to monitor the three-property limit for an individual obligor across lenders on a continuous basis would be operationally complex and place a high burden on firms. Some requested that the PRA clarify its proposed requirements for ongoing monitoring. Sixteen respondents did not support the proposed calibration of the three-property limit, and some argued that the limit should be increased to reflect the relative risk of larger professional landlords compared to those with smaller portfolios. Some requested that the PRA clarify the proposed treatment of properties held through corporate structures (eg special purpose entities) and whether these exposures would be in scope of the three-property limit.

2.247 The PRA continues to consider that individual obligors who have no more than three mortgaged residential properties (excluding their primary residence) typically represent a lower risk than obligors with more mortgaged properties. The exception from treatment as materially dependent on cashflows for exposures to individual obligors with no more than three residential properties (the ‘three property limit’) is prudent and proportionate and advances the PRA’s primary objective, given the increased risk associated with larger property portfolios.

2.248 However, based on the responses received, the PRA has decided to amend its draft rules on how firms calculate the three-property limit for exposures to individuals and exposures to special purpose entities. For exposures to individuals, the number of properties will be the sum of:

  • properties used as security for a residential real estate exposure to the individual; and
  • properties used as security for a residential real estate exposure to a special purpose entity where the individual is a guarantor, and the individual receives the economic benefit from the residential real estate.

2.249 In counting the number of properties:

  • the obligor’s primary residence will not be considered in any circumstances;
  • properties that act as security for residential real estate exposures will be counted regardless of the lender; and
  • firms must consider a single housing unit that is a separate part within a multi-unit property as one property.

2.250 The PRA will only require firms to update their assessment of material dependence on cashflows from the property (including assessing the number of properties mortgaged by the obligor where relevant) when a new or replacement mortgage is issued. When the three-property limit is exceeded, all residential real estate exposures to that individual or entity must be treated as being materially dependent on the cash flows generated by the property (unless they are secured by the obligor’s primary residence).

2.251 These amendments to the calculation of the three-property limit will result in a simpler and more proportionate approach. The PRA considers that properties used as security for a residential real estate exposure to other lenders to be relevant in determining the risk, and only considering exposures to a single firm could therefore lead to imprudent outcomes.

Risk-weighting of real estate sub-classes

Treatment of ‘regulatory residential real estate’ exposures that are not materially dependent on the cash flows generated by the property

2.252 The PRA proposed to introduce more risk-sensitive risk weights for regulatory residential real estate exposures. For regulatory residential real estate exposures that are not materially dependent on the cash flows generated by the property, the PRA proposed to specify that the risk weight treatment would be the loan splitting approach as set out in the Basel 3.1 standards. The loan splitting approach assigns a 20% risk weight to the part of the exposure value up to 55% of the property value and the risk weight of the counterpartyfootnote [38] to any residual part of the exposure.

2.253 For second charge mortgage loans, the PRA proposed that, where there are charges on the property not held by the firm, the loan splitting approach should be adjusted, such that the amount of exposure that would be assigned a 20% risk weight would be reduced to reflect other charges on the property and appropriately reflect the risk of the firm’s exposures.

2.254 Two respondents supported the proposed risk weights, and four respondents supported the increased risk sensitivity in the proposed treatment for regulatory residential real estate more generally. However, four respondents did not support the proposed calibration of the risk weight treatment.

2.255 Three respondents argued that the proposed treatment for second charge mortgage loans under the loan splitting approach for regulatory residential real estate exposures that are not materially dependent on the cash flows generated by the property was overly punitive. The respondents argued that the loan splitting approach would result in increased risk weights for second charge mortgage loans that are not justified by the risk, and risk weights that would be broadly similar to unsecured exposures which they consider too conservative. The respondents argued that the whole loan approach available in the Basel 3.1 standards for regulatory residential real estate exposures that are not materially dependent on the cash flows generated by the property should be available for second charge mortgage loans.

2.256 Having considered the responses, the PRA has decided to maintain its proposed risk weight treatment for regulatory residential real estate exposures that are not materially dependent on the cash flows generated by the property. However, the PRA has adjusted the counterparty risk weights that apply to exposures to SMEs (see the section on the treatment of regulatory commercial real estate below). The PRA notes that its proposal will significantly increase risk sensitivity compared to the CRR treatment and considers that this is one of the most important developments in the SA. The PRA did not receive persuasive evidence to support respondents’ assertions that the calibration in the Basel 3.1 standards was overly conservative. The package is prudent and reflects the risk of exposures, while facilitating effective competition by significantly reducing the gap between SA risk weights and risk weights calculated under the IRB approach. It is also in line with international standards.

2.257 The PRA has also decided to maintain its proposed risk weight treatment for second charge mortgage loans. The PRA considers that the loan splitting approach captures the risks associated with second charge mortgage loans, whereas other approaches (including the ‘whole loan’ approach) would not. Second charge mortgage loans are generally riskier than first charge mortgage loans. Even if recoveries are maximised, the second charge holder does not receive any recoveries until the first charge holder is fully paid meaning the LGD on the second charge will typically be higher. In a downturn, falls in the value of the property may result in the second charge holder being reliant on unsecured recoveries. The loan splitting approach reflects these dynamics and ensures total capital requirements remain unchanged if a single mortgage loan is split into two sequential charges. It would not be prudent for capital requirements to reduce in these circumstances.

Treatment of ‘regulatory residential real estate’ exposures that are materially dependent on the cash flows generated by the property

2.258 For regulatory residential real estate exposures that are materially dependent on the cash flows generated by the property, the PRA proposed to require firms to risk-weight the whole loan amount of such exposures using the relevant risk weight determined by the loan to value (LTV) of the exposures, ranging between 30% and 105% (the ‘whole loan’ approach). This would result in higher risk weights for these exposures than for regulatory residential real estate exposures that are not materially dependent on the cash flows generated by the property. In cases where the firm has a junior charge and there are senior charges not held by the firm, the PRA proposed that firms should multiply the applicable risk weight by 1.25 (unless the LTV is less than or equal to 50%).

2.259 Five respondents did not support the proposed increased risk weights for regulatory residential real estate exposures that are materially dependent on the cash flows generated by the property relative to exposures that are not materially dependent on the cash flows generated by the property. Five respondents did not support the proposed increased risk-weights for higher-LTV exposures that are materially dependent on the cash flows generated by the property relative to the current treatment. Two respondents argued that there should be further risk sensitivity in the final treatment.

2.260 One respondent supported the proposed requirement to use the whole loan approach for regulatory residential real estate exposures that are materially dependent on the cash flows generated by the property. However, two respondents argued that the whole loan approach would introduce operational complexity as firms may be required to use two different risk weight calculation methods across regulatory real estate portfolios.

2.261 Having considered the responses, the PRA has decided to maintain the requirement for firms to assign higher risk weights to regulatory residential real exposures that are materially dependent on the cash flows generated by the property and also to maintain the proposed approach to assign a risk weight to the whole exposure based on the LTV. The PRA did not receive persuasive evidence to support respondents’ assertions that the level of risk was the same as for residential real estate exposures that are not materially dependent on the cash flows generated by the property, particularly in a downturn. The PRA continues to consider that exposures which are materially dependent on the cash flows generated by the property are generally higher risk than exposures that are not materially dependent on the cash flows generated by the property.

2.262 However, the PRA agrees that the risk weight treatment could be made more risk sensitive. The PRA has amended its draft rules to introduce greater risk sensitivity in the treatment for exposures that fall in the range 60% < LTV ≤ 80%, assigning a 40% risk weight for exposures that fall in the range 60% < LTV ≤ 70% and a 50% risk weight for exposures that fall in the range 70% < LTV ≤ 80% (in CP16/22 the PRA proposed to assign a 45% risk weight for all exposures that fall in the range 60% < LTV ≤ 80%).

Other issues relating to the treatment of regulatory residential real estate

2.263 Other substantive responses to this section of CP16/22 focused on the following issues:

  • Social housing exposures: two respondents did not support the minimum counterparty risk weight of 75% for social housing exposures and noted that it could result in secured exposures attracting a higher risk weight than unsecured exposures to the same counterparty. The PRA has decided to maintain the proposed minimum counterparty risk weight of 75% for social housing exposures. The PRA considers that this approach is risk-sensitive and proportionate. Social housing exposures will not be considered to be materially dependent on the cash flows generated by the property and will therefore not be subject to the higher risk weights under the whole loan approach, despite the obligor’s ability to repay the loan, at least in part, being reliant on the performance of the individual tenants. The PRA considers that this risk should instead be reflected in the application of the retail counterparty risk weights under the loan splitting approach. As set out above, the PRA has decided to amend its draft rules to introduce a similar treatment for exposures to housing associations and cooperatives.
  • Social housing exposures: one respondent argued that only drawn commitments should be considered for determining the applicable risk weight under the loan splitting approach. The PRA has decided to maintain the proposed requirement that undrawn balances must be considered for determining the risk weight. It considers this to be a more risk-sensitive approach which reflects the non-zero chance of such off-balance sheet items moving on to the balance sheet.
  • Loan splitting approach calculation: one respondent suggested that the PRA should include the examples of the loan splitting approach calculation set out in the Basel 3.1 standards in a supervisory statement. The PRA has provided further examples of the loan splitting approach calculation in its near-final amendments to SS10/13.

Treatment of ‘regulatory commercial real estate’

2.264 For commercial real estate exposures, the PRA proposed that risk weights should be no lower than 100% given the default experience of this asset class in stressed environments and broader factors such as market price stability.

2.265 For regulatory commercial real estate exposures that are not materially dependent on the cash flows generated by the property, the PRA proposed that firms would be required to risk weight the exposure by assigning the risk weight of the counterparty or 60% (whichever is lower) to the part of the exposure up to 55% of the property value, and apply the risk weight of the counterparty to the residual part of the exposure. If the resultant risk weight was lower than 100%, a risk weight of 100% would be applied to the exposure.

2.266 For regulatory commercial real estate exposures that are materially dependent on the cash flows generated by the property, the PRA proposed to require firms to apply a 100% risk weight, unless the LTV of the exposure was greater than 80%, in which case the risk weight would be 110%.

2.267 Eighteen respondents did not support the PRA’s proposal that risk weights should be no lower than 100% for regulatory commercial real estate exposures, particularly for exposures which are not materially dependent on the cash flows generated by the property. Respondents argued that a 100% risk weight was overly conservative for these exposures and that this would result in a lack of risk sensitivity which diverged from the stated objective of the Basel 3.1 standards, placing UK firms at a disadvantage relative to international competitors. Respondents also argued that the proposals would create perverse incentives for firms given secured commercial real estate exposures to SMEs would receive a higher risk weight than unsecured exposures. Some respondents requested that the PRA publish analysis to justify the calibration of the risk weights for commercial real estate exposures.

2.268 The majority of respondents argued that the PRA should align the risk weight treatment with the Basel 3.1 standards, while some respondents argued commercial real estate exposures that are not materially dependent on the cash flows generated by the property should be subject to a lower minimum risk weight.

2.269 The treatment of commercial real estate under the PRA’s near-final rules is summarised in Table 4. Having considered the responses, the PRA acknowledges the arguments in relation to the risk-sensitivity of the draft rules. It has amended the draft rules to remove the 100% minimum risk weight in the cases outlined below, distinguishing between commercial real estate exposures to SMEs and non-SMEs, and between exposures that are materially dependent on the cash flows generated by the property and exposures that are not materially dependent on the cash flows generated by the property:

  • for regulatory commercial real estate exposures to SMEs that are not materially dependent on the cash flows generated by the property, firms will be required to risk weight exposures based on the loan splitting approach. Firms will be required to assign a risk weight of 60% to the part of the exposure up to 55% of the property value, and apply the risk weight of the counterparty to the residual part of the exposure; and
  • for regulatory commercial real estate exposures to corporates that are non-SMEs that are not materially dependent on the cash flows generated by the property, firms will be required to apply the counterparty risk weight or 60% (whichever is greater), unless the risk weight would be greater than if the exposure were materially dependent on the cash flows generated by the property, in which case the risk weight applicable to the exposure will be the same as the risk weight for exposures materially dependent on the cash flows generated by the property (100% or 110%, according to LTV).

2.270 Having considered the responses, the PRA has decided to maintain its proposal in relation to regulatory commercial real estate exposures to both SMEs and non-SMEs that are materially dependent on cash flows from the property. The PRA has retained the minimum risk weight of 100% in its near-final rules and firms will be required to risk weight exposures with LTV ≤ 80% at 100% and LTV > 80% at 110%.

Table 4: Near-final treatment of exposures secured by commercial real estate

SME

Non-SME corporates

Regulatory and not materially dependent on cash flows from the property

Loan splitting approach
60% risk weight up to 55% of the property value
Apply the counterparty risk weight to the residual part of the exposure (if any)footnote [39]

Counterparty risk weight with a 60% risk weight floor and capped at the ‘materially dependent on cashflows’ treatment as set out in the row below

Regulatory and materially dependent on cash flows from the property

Apply a 100% risk weight floor.
LTV ≤ 80% = 100% risk weight
LTV > 80% = 110% risk weight

Apply a 100% risk weight floor.
LTV ≤ 80% = 100% risk weight
LTV > 80% = 110% risk weight

Non-regulatory and not materially dependent on cash flows from the property

Counterparty risk weightfootnote [40]

No collateral benefit, apply the counterparty risk weight with 60% risk weight floor

Non-regulatory and materially dependent on cash flows from the property


150% risk weight


150% risk weight

2.271 These changes will result in a more risk-sensitive approach, reflecting the PRA’s analysis of the differing risk mitigation provided by commercial property collateral depending on the status of the obligor. Analysis of stress testing data for firms with approved IRB models shows that, on average, exposures to SMEs that are secured on collateral have both a lower PD and LGD than unsecured exposures (see Table 5).

2.272 For exposures to non-SME corporates, the analysis shows that, on average, the overall riskiness of secured and unsecured exposures is similar. Secured exposures have a lower LGD but a higher PD than unsecured exposures, which results in similar implied overall losses for secured and unsecured exposures. The analysis does not justify giving credit for holding commercial real estate collateral in respect of exposures to non-SME corporates. Equally, though, it does not suggest that the riskiness of secured borrowing warrants a risk weight that is no lower than 100%. However, the PRA considers that a minimum risk weight of 60% is necessary to ensure a minimum level of prudence, and to reduce the cliff edge in risk weights between exposures that are materially dependent on cashflows from the property and those that are not.

Table 5: IRB modelled PDs and LGDs for corporate SMEs and non-SME corporates (annual cyclical scenario stress test, base case data)

Corporate SMEs

Non-SME corporates

Secured

Unsecured

Secured

Unsecured

PD

1.66%

2.01%

1.29%

0.69%

LGD

29.69%

42.65%

32.64%

43.44%

2.273 The PRA has amended its draft rules to clarify that, where a firm has a real estate exposure to an SME that meets the PRA’s revised definition of a retail exposure (except the requirement that it is not a real estate exposure), the counterparty risk weight shall be 75%. The counterparty risk weight for other SMEs will continue to be 85%. This is relevant for commercial real estate exposures but will apply in respect of any relevant real estate exposures where the counterparty risk weight is used to assign a risk weight to an SME exposure. This increases alignment between the risk weight treatments of secured and unsecured exposures.

2.274 The PRA continues to consider that its proposed treatment for commercial real estate exposures that are materially dependent on cash flows generated by the property appropriately reflects the greater risks associated with commercial real estate lending where the obligor is likely to be reliant on rental income for repayment of the loan.

Treatment of ‘other commercial real estate’

2.275 The PRA proposed that commercial real estate exposures which do not meet all the requirements of regulatory commercial real estate would be classed as other commercial real estate. Other commercial real estate exposures which are not materially dependent on the cash flows generated by the property would be risk-weighted according to the risk weight of the counterparty, but the risk weight would be floored at 100%. Other commercial real estate exposures which are materially dependent on the cash flows generated by the property would be risk-weighted at 150%.

2.276 As with regulatory commercial real estate exposures, respondents argued that secured exposures should not receive a higher risk weight than unsecured exposures.

2.277 For consistency with the treatment of regulatory commercial real estate exposures, the PRA has revised its draft rules regarding the treatment of other commercial real estate exposures which are not materially dependent on the cash flows generated by the property, distinguishing between exposures to SMEs and non-SMEs:

  • the risk weight for other commercial real estate exposures to SMEs which are not materially dependent on the cash flows generated by the property will be the relevant counterparty risk weight (75% for retail SMEs and 85% for corporate SMEs); and
  • the risk weight for other commercial real estate exposures to non-SME corporates which are not materially dependent on the cash flows generated by the property will be the risk weight of the counterparty or 60% (whichever is greater).

2.278 The PRA has decided to maintain its proposed 150% risk weight for other commercial real estate exposures which are materially dependent on cashflows from the property. This is consistent with the Basel 3.1 standards.

Exposures with multiple types of property collateral

2.279 Other substantive responses to this section of CP16/22 focused on the treatment of exposures with multiple types of property collateral.

2.280 Three respondents requested that the PRA clarify the proposed treatment of exposures where the loan is secured against a mixture of residential real estate and commercial real estate. The PRA has decided to amend its draft rules to state that a mixed real estate exposure will be split into a residential real estate exposure and a commercial real estate exposure. The PRA has also provided further clarification on the approach firms are expected to take to risk-weight exposures in its near-final amendments to SS10/13. Firms will be required to use a value-weighted allocation approach, where the proportion of the total value of the property collateral for each type of real estate collateral is used to weight the allocation of the loan amount (and the respective exposure value) across real estate exposure categories.

2.281 Both the residential real estate part and the commercial real estate part of a mixed real estate exposure must separately meet the requirements to be deemed a regulatory real estate exposure (see above) in order for both parts to be risk-weighted as regulatory real estate exposures. In all other cases (including if one part meets the requirement to be regulatory real estate and the other part does not), both parts must be risk-weighted as other real estate exposures.

Land acquisition, development and construction exposures

2.282 The PRA proposed to introduce a specific treatment for ‘land acquisition, development and construction’ (ADC) exposures that would apply to exposures to a corporate or special purpose entity financing land acquisition for development and construction purposes, or financing development and construction of any residential or commercial real estate. The PRA considered that higher risk weights for ADC exposures reflected the higher risk of these exposures, given that the source of repayment of the loan was a planned, but uncertain sale of the property, or substantially uncertain cash flow. The PRA proposed to assign a risk weight of 150% to ADC exposures, except that in cases where certain conditions were met, a loan granted for the purpose of developing residential real estate would be assigned a risk weight of 100%.

2.283 Four respondents argued that the PRA should consider exemptions to the definition of ADC exposures to support decarbonisation, regeneration, and the build-to-rent sector. One respondent raised concerns with the difference in risk weights for residential development finance in the SA compared to the IRB approach. One respondent noted that the ADC exposure class would capture exposures not currently classified under items associated with high risk as speculative immovable property financing. Four respondents argued that the PRA should provide guidance to support firms’ assessment of when a 100% risk weight can be assigned to ADC exposures, with some respondents arguing that a lack of clarity could lead to inconsistent application by firms.

2.284 Having considered the responses, the PRA has decided to maintain its proposed risk weight treatment for ADC exposures. The PRA considers the scope of exposures captured remains appropriate given the risk profile of these exposures and considers that it would not be appropriate to introduce exclusions to the treatment. ADC exposures are generally higher risk than other real estate exposures and unsecured corporate lending, as the repayment of ADC loans at origination of the exposure is either dependent on the future uncertain sale of the property, or the source of cash flows is substantially uncertain. There are various reasons for the increased risk associated with the uncertain repayment of the loan, such as the risk of delayed expected completion due to a deterioration in a counterparty’s financial condition or market conditions.

2.285 The PRA has decided to substantially maintain its proposed approach for assigning a 100% risk weight for ADC exposures to allocate exposures that meet both the definition of regulatory real estate and the definition of ADC. However, the PRA has decided to make amendments to its draft rules to clarify that, to be eligible, an exposure must be subject to prudent underwriting standards, including for the valuation of any real estate used as security for the exposure. The PRA will consider providing further clarification in future of which ADC exposures are eligible for a 100% risk weight if evidence emerges of inconsistent application across firms.

Definition of ADC exposure

2.286 Other substantive responses to this section of CP16/22 focused on the allocation of exposures as ADC. Two respondents requested that the PRA clarify how firms would be required to allocate an exposure when it meets the definition of regulatory real estate and ADC. One respondent argued that the proposed treatment for regulatory real estate exposures would be more appropriate for the risk profile of such exposures. The PRA has amended its draft rules to clarify that the definition of ADC exposures relates to the purpose of the lending, and that exposures which meet the criteria for regulatory real estate and ADC must be treated as ADC exposures. This definition aligns with the Basel 3.1 standards and is prudent given the heightened risk of ADC exposures.

PRA objectives analysis

2.287 The PRA considers that its analysis of its objectives in CP16/22 remains appropriate, subject to the following updates to reflect changes to the proposals for real estate.

2.288 The PRA considers that its amended proposals for the valuation of real estate promote safety and soundness by dampening the cyclicality of property valuations, relative to indexed approaches or other more frequent valuation approaches. Although the amended proposals are likely to be marginally more cyclical than those in CP16/22, the PRA considers that its amended proposals appropriately balance risk sensitivity and cyclicality.

2.289 The PRA considers that its decision to change the approach to assessing whether an exposure is materially dependent on cashflows from the property promotes safety and soundness by improving the clarity of the requirements which should result in a more consistent and comparable approach across firms. The PRA considers that its adjustments to the risk weight applicable to residential real estate exposures which are materially dependent on the cash flows generated by the property in the 60% < LTV ≤ 70% and 70% < LTV ≤ 80% categories promotes safety and soundness, because it adds further risk sensitivity to the framework.

2.290 The PRA considers that its decision to reduce risk weights for certain commercial real estate exposures results in an overall treatment that advances its primary objective. The proposed changes would improve risk sensitivity relative to the existing treatment and are based on evidence of the risk of particular types of commercial real estate exposures, including exposures which are materially dependent on cashflows from the property which are subject to a minimum 100% risk weight because the collateral represents the main source of repayment for the loan.

2.291 In respect of the PRA’s secondary objective to facilitate effective competition, the PRA’s decision to permit some self-build properties to be treated as regulatory real estate will result in a smaller difference in risk weights between the SA and the IRB approach for self-build loans relative to the proposals in CP16/22, thereby facilitating effective competition. The PRA also considers that its near-final approach to assessing whether an exposure is materially dependent on cashflows from the property facilitates effective competition by promoting consistent application across firms. The PRA’s decision to reduce risk weights for certain commercial real estate exposures brings the average level of capital requirements under the SA for these exposures closer to that of firms applying IRB approaches, facilitating effective competition.

‘Have regards’ analysis

2.292 In developing these near-final rules, the PRA has had regard to its framework of regulatory principles and the matters to which it is required to have regard when proposing amendments to CRR rules. The PRA considers its analysis of its ‘have regards’, as presented in chapter 3 of CP16/22, remains appropriate, subject to the following updates:

1. Relevant international standards:

  • The PRA’s decision to apply lower risk weights than those proposed under CP16/22 and the CRR to certain commercial real estate exposures results in a treatment that is closer to the baseline set out in the Basel 3.1 standards than the existing treatment.

2. Relative standing of the UK as a place for internationally active firms to operate and competitiveness:

  • The PRA generally expects that its approach to real estate valuation will produce outcomes that are comparable with other major jurisdictions, but notes that the revised approach expected to be taken by the EU to valuations is likely to produce real estate risk weights that are more cyclical than the PRA’s approach. Lower cyclicality of risk weights is of benefit to UK firms, by reducing volatility in capital requirements.
  • The PRA’s amended approach to risk weighting commercial real estate exposures will result in risk weights that are closer to the EU’s expected approach, relative to the existing treatment, while some jurisdictions have proposed more conservative risk weights than the PRA’s amended approach. The PRA considers that its approach supports the relative standing of the UK and the ability of UK firms to compete in a market with cross-border lending.

3. Different business models:

  • The PRA’s decision to permit some self-build real estate exposures to be classed as regulatory real estate exposures means that firms specialising in self-build mortgage loans will not be affected in the way described in CP16/22. The net effect of the PRA’s near-final rules affecting self-build mortgage loans – treating some self-build mortgage loans as regulatory real estate and introducing specific valuation requirements for self-build mortgage loans – will be lower capital requirements relative to CP16/22. Relative to the current treatment, capital requirements could increase or decrease dependent on the specific nature of firms’ exposures.
  • The PRA’s analysis in CP16/22 indicated that specialist firms with exposures to holiday lets, purpose-built student accommodation and care homes could be affected, because these exposures were to be explicitly excluded from treatment as residential real estate and would have received a higher risk weight than now, as commercial real estate. This is now less likely to be the case because holiday lets, purpose-built student accommodation and care homes may in some cases be able to qualify for treatment as residential real estate under the near-final rules and near-final amendments to SS10/13.
  • The PRA’s decision on the approach to assessing if residential real estate exposures are materially dependent on the cash flows generated by the property may have a greater impact on firms who target lending towards portfolio landlords and high-net-worth individuals, relative to the current treatment. The PRA’s decision – for both residential and commercial exposures – is intended to balance risk sensitivity with promoting consistent implementation across firms, including smaller firms who may otherwise face disproportionate complexity in the calculation of capital requirements.

4. Sustainable growth and growth:

  • The PRA’s decision to apply lower risk weights than those in the current treatment and proposed in CP16/22 to certain commercial real estate exposures helps to promote sustainable growth, by aligning risk weights for specific exposures (and specific counterparties like SMEs) with their relative risk, as reflected in data on modelled PD and LGDs.

5. Proportionality:

  • The PRA considers that its approach to assessing whether an exposure is materially dependent on cashflows from the property is proportionate, as a result of the assessment being simple, objective and conducted only at origination events. The PRA considers that this proportionate approach is likely to result in a similar level of capitalisation in aggregate, relative to a more complex approach such as the proposals in CP16/22. The removal of the proposal to require real estate valuations to be adjusted to reflect the sustainable value over the life of the loan should make the amended approach for the valuation of real estate more proportionate than the CP16/22 proposals.

Capital instruments, defaulted exposures, and high-risk items

2.293 Aligned with the Basel 3.1 standards, the PRA proposed a number of changes to enhance the risk sensitivity and robustness of the treatment of the SA risk weights for: (i) equities, subordinated debt and other capital instruments; (ii) defaulted exposures; and (iii) high-risk items.

2.294 The PRA proposed to remove the use of the IRB approach to calculate RWAs for equity exposures, and, as a result, the proposed SA treatment would apply to all equity exposures (subject to proposed transitional arrangements).

2.295 The PRA received 14 responses to this section of CP16/22. The substantive issues raised by respondents are set out below.

Equities, subordinated debt and other capital instruments

2.296 The PRA proposed to apply a 400% risk weight to venture capitalfootnote [41] exposures and a 250% risk weight to other equity exposures. The PRA proposed that the new SA equity risk weights would be phased-in over a five-year period.

2.297 Two respondents did not support the proposals, arguing that the proposed increase in the equity risk weights was not prudentially justified as equity portfolios showed limited downside risks of losses, while offering greater potential profit than other exposure classes. Five respondents argued the PRA should permit firms to reduce equity risk weights to reflect diversification benefits, as is the case under the CRR IRB simple risk weight approach. Respondents argued that diversification limits the risk of downside losses by offsetting losses with profits on other investments.

2.298 Four respondents requested that the PRA clarify the proposed definition of ‘venture capital’ and, in particular, the PRA’s intended definition of ‘newly established enterprises’. One respondent argued for an approach based on the age of the business being invested in for defining exposures subject to the 400% risk weight, arguing it would be the most practical method.

2.299 Having considered the responses, the PRA has decided to maintain its proposal that equity exposures should either be assigned a 400% or a 250% risk weight. Applying higher risk weights to equity exposures than to other types of exposures reflects that equity exposures are riskier, given their position in the creditor hierarchy and the expected low level of recoveries in default. The PRA also considers that, in line with the Basel 3.1 standards and approaches expected to be adopted by other jurisdictions, there is a subset of equity exposures that carry a materially higher credit risk, and therefore require a risk weight of 400%.

2.300 The proposed risk weight calibration aligns with the Basel 3.1 standards and with other major jurisdictions such as the EU and the US (based on their proposed implementation). The PRA did not receive persuasive evidence from firms to support their assertions that lower risk weights for equity exposures are justified, including regarding the impact of diversification on the aggregate risk of equity exposures. In particular, the diversification of equity portfolios and the level of credit risk will vary materially across portfolios. This is difficult to capture under the SA and it is important to maintain the simplicity and proportionality of the SA approach. The PRA also notes that the Basel 3.1 standards similarly do not provide a diversification benefit for equity exposures.

2.301 Having considered the responses, however, the PRA has made significant amendments to the criteria for assigning a 400% risk weight to equity exposures. The PRA has amended its draft rules to require firms to apply a 400% risk weight to exposures to businesses that are less than five years old, instead of to ‘venture capital exposures’ as proposed in CP16/22.footnote [42] The PRA has also decided to set out expectations on how firms should determine the age of a business in its near-final amendments to SS10/13. Once a business is over five years of age, a 250% risk weight will be applied to any exposures to that business.

2.302 The PRA is taking this approach because respondents indicated that it would be difficult to define ‘venture capital’ in a way that could be consistently implemented by firms and because the available data,footnote [43] displayed in the chart below, shows there is a clear correlation between the age of a business and its risk of failure. The PRA considers that the revised criteria for applying a 400% risk weight should maintain the right level of risk sensitivity while remaining broadly aligned with the intent of the Basel 3.1 standards, and should address respondents’ concern by providing a clearer definition that will reduce the likelihood of inconsistent application.

Chart 1: US and UK business failure rate (a)

Business failure rate in the UK and US. Data shows a decrease in the failure rate in the US and UK after 3 years following initial high rates of failure. The US data, which continues for 10 years as opposed to the UK which is only over a 5 year period, shows another more marked decrease at 5 years, with failure rate plateauing after, at a low level.

Footnotes

  • Business failure rate in the UK and US. Sample of data taken from a selection of years within specified date ranges.footnote [44]
  • (a) US data has been adjusted by one year to align methodologies with ONS data. BLS defines as business death as no trading activity for 4 quarters and therefore had 100% business survival in year one.

Other issues relating to equities, subordinated debt and other capital instruments

2.303 Other substantive responses to this section of CP16/22 focused on the following issues:

  • Exposures to companies where there is an ongoing business relationship: two respondents argued that the PRA should reduce risk weights for equity exposures to companies where the firm intends to have a long-term business relationship. The PRA has considered the response and has decided to maintain its proposal. The PRA did not receive persuasive evidence to support respondents’ assertions that an ongoing business relationship reduces credit risk.
  • Exposures arising from debt/equity swaps: two respondents argued that exposures arising from debt/equity swaps should be excluded from the 400% risk weight, aligning with a national discretion in the Basel 3.1 standards. The PRA has decided to not amend its draft rules. The PRA considers that exposures arising from debt/equity swaps generally carry the same level of risk as other equity exposures and the PRA did not receive persuasive evidence to suggest otherwise.
  • Nationally legislated programmes: four respondents argued that equity exposures to national legislated programmes should be risk weighted at 100%, aligning with a national discretion in the Basel 3.1 standards. The PRA has decided not to amend its draft rules. Co-investment with government bodies does not provide a guarantee of returns, and the PRA considers that the CRM framework enables firms to reflect guarantees where appropriate. The PRA is not aware of any UK programmes which fall within scope of the national discretion.
  • Exposures to subsidiaries: two respondents argued that the PRA should retain a 100% risk weight for certain subsidiary investments instead of the proposed 250% risk weight for equity exposures, arguing that firms investing in their subsidiaries will have detailed information to support the valuation of subsidiary share-based capital. The PRA has decided to not amend its draft rules because it does not consider that access to additional information in and of itself provides a meaningful mitigation against credit risk for equity investments.
  • Capital instruments with the economic substance of debt: two respondents argued that capital instruments with the economic substance of debt should be risk-weighted in a similar manner to corporate loans and not be treated as ‘equity exposures’, aligning with the national discretion in the Basel 3.1 standards. The PRA clarifies that that capital instruments (including equity instruments) that do not meet the criteria set out in Articles 133(1) to (2) of the near-final Credit Risk: Standardised Approach (CRR) Part will not be risk-weighted as equity exposures and will be assigned a 150% risk weight. The PRA did not receive persuasive evidence that these instruments have the economic substance to justify receiving the same risk weight treatment as loans that qualify to be risk-weighted as corporate exposures.
  • Other instruments with the economic substance of equities: the PRA has decided to amend its draft rules to remove the requirement that firms must request permission from the PRA to exclude certain instruments from being risk-weighted as equity exposures. However, firms will still be required to be able to demonstrate that such instruments should be treated as debt positions according to the criteria in the PRA’s near-final rules. The PRA considers that this will advance its objective of safety and soundness in a more proportionate way.
  • Listed equities and CIUs: one respondent requested that the PRA clarify how firms would be required to treat listed equities held for trading purposes and the treatment of shares held in the banking book as part of an unlisted CIU that is a hedge fund. The PRA clarifies that firms should apply the trading book boundary criteriafootnote [45] to determine whether exposures are held under the banking book or trading book. Firms should apply Article 133 of the near-final Credit Risk: Standardised Approach (CRR) Part to determine the treatment of exposures held in the banking book. For exposures to CIUs, firms should consult Articles 132 to 132C of the near-final Credit Risk: Standardised Approach (CRR) Part to determine the correct risk weighting approach. Where a firm is applying the look-through approach for risk-weighting exposures to CIUs, the firm should apply the definitions in Article 133 to the CIU’s underlying exposures s that fall within the scope of that article.

Defaulted exposures

2.304 The PRA proposed to introduce several changes to the SA risk weights for defaulted exposures, including restricting the flat 100% risk weight treatment to residential real estate exposures that are not materially dependant on the cash flows generated by the property, in line with the Basel 3.1 standards. The PRA also proposed to clarify the criteria that are used under the SA to determine whether exposures are treated as retail exposures for the purpose of applying the definition of default (as set out in set out in Article 178 of the near-final Credit Risk: Internal Ratings Based Approach (CRR) Part). In the case of such exposures, an institution may apply the definition of default at the level of an individual credit facility rather than in relation to the total obligations of an obligor.

2.305 Five respondents argued that the PRA should exercise a national discretion in the Basel 3.1 standards which enables a 50% risk weight to be applied to defaulted exposures where specific provisions are no less than 50% of the outstanding amount of the loan. Having considered the responses, the PRA has decided not to amend its draft rules. The PRA did not receive persuasive evidence that exercising the national discretion would result in risk weights that, on average, better reflect the risks associated with defaulted exposures.

2.306 Two respondents did not support the proposal to restrict the availability of the 100% risk weight to defaulted residential real estate exposures which are not materially dependent on cashflows from the property. Having considered this response, the PRA has decided not to amend its draft rules. The PRA considers that the risk weight treatment for defaulted exposures appropriately reflects the higher risks associated with commercial real estate and real estate that is materially dependent on cash flows from the property and is consistent with the risk-weighting of real estate exposures more broadly.

2.307 The PRA has however amended its draft rules to simplify the treatment of defaulted exposures under the SA. All exposures that meet the definition of a ‘retail exposure’ under the SA and all real estate exposures that would otherwise meet this definition shall be treated as retail exposures for the purpose of applying the definition of default.

2.308 This treatment aligns with the Basel 3.1 standards and appropriately reflects the broader changes to the classification of retail exposures outlined in the ’Exposures to individuals and small and medium-sized enterprises’ section of this chapter.

2.309 The PRA also proposed that the risk weights for defaulted exposures would apply to the part of each defaulted exposure that is not covered by collateral or a guarantee. The PRA has made minor changes to its draft rules to clarify that the size of the uncovered part should be calculated as set out in the near-final Credit Risk: Credit Risk Mitigation (CRR) Part.

High risk items

2.310 The PRA proposed to retain elements of CRR Article 128 that apply a 150% risk weight to exposures that are associated with particularly high risk. Three respondents argued that the updated treatment for ADC and equity exposures rendered the high-risk exposures treatment redundant. Having considered the responses, the PRA has decided to maintain its proposed approach. The PRA acknowledges that broader changes to the SA framework will mean fewer exposures will be classed as being of particularly high risk but considers that maintaining the exposure class will preserve existing risk sensitivity in the framework.

PRA objectives analysis

2.311 With the exception of the change to the criteria for assigning a 400% risk weight to equity exposures, the PRA considers that the objectives analysis set out in CP16/22 remains appropriate.

2.312 The PRA considers that the decision to introduce a new age-based criteria to determine whether an equity exposure is assigned a 400% risk weight will result in RWAs that better reflect the risks associated with equity and other capital instruments and should promote consistent application of the rules across firms. The PRA therefore considers that it is consistent with both its primary and secondary objectives.

‘Have regards’ analysis

2.313 In developing these near-final rules, the PRA has had regard to its framework of regulatory principles and the matters to which it is required to have regard when proposing amendments to CRR rules. The PRA considers its analysis of its ‘have regards’, as presented in chapter 3 of CP16/22, remains appropriate, subject to the following update:

1. Proportionality:

  • The PRA continues to acknowledge that allocating equity exposures to the different risk weight categories will result in some initial operational costs for firms but considers this to be proportionate to the benefits that the simplified criteria provide.
  • Permitting equity exposures to be assigned a 250% risk weight once the business is over 5 years of age should result in a proportionate approach as the 400% risk weight will only apply to younger equity issuers and then reduce as the equity gets older, which is justified by the level of credit risk.

3: Credit risk – internal ratings based approach

Introduction

3.1 This chapter provides feedback to responses to chapter 4 of consultation paper (CP) 16/22 – Implementation of the Basel 3.1 standards, which set out proposals to implement the Basel 3.1 standards for the internal ratings based (IRB) approach to credit risk. This chapter also sets out the Prudential Regulation Authority’s (PRA) near-final policy on the IRB approach for credit risk following the consultation.

3.2 In CP16/22, the PRA proposed to introduce restrictions on the use of the IRB approach for equities and low default portfolios such as exposures to banks, financial corporates, large corporates, and sovereign exposures. Other proposals included changes to the risk parameters used in IRB modelling, including new input floors for probability of default (PD), loss given default (LGD) and exposure at default (EAD), and greater specification of parameter estimation practices to reduce variability in risk-weighted assets (RWAs) for portfolios where the IRB approach remains available.

3.3 The PRA received 34 responses to its proposals on the IRB approach for credit risk. Responses covered a wide range of the PRA’s proposals. Comments focused on:

  • the appropriate approaches to credit risk for different types of exposures;
  • the calibration of IRB risk weight parameters;
  • the proposals to remove the small and medium-sized enterprises (SME) support factor and the infrastructure support factor;
  • the PRA’s proposed timelines for firms to make modelling changes; and
  • the proportionality and clarity of the PRA’s proposed requirements and expectations.

3.4 Having considered the responses to the consultation, the PRA has decided to amend the draft rules and draft supervisory statements (SSs) in certain areas. This chapter describes the comments and amendments that the PRA considers are material. As described in Chapter 1 – Overview, the PRA has also made a number of less material amendments and clarifications to the draft rules and draft SSs, which are not described in this chapter. These amendments are reflected in the near-final PRA Rulebook: CRR Firms: (CRR) Instrument [2024] in Appendix 2 Please refer to the document titled Comparison of Draft PRA Rulebook (CRR) Instrument [2023] against Near-final PRA Rulebook: CRR Firms (CRR) Instrument [2024], which contains a comparison of the near-final rules with the draft rules as set out in CP16/22 for ease of identifying all of the changes made.

3.5 The appendices to this near-final policy statement (PS) contain the PRA’s near-final policy, which will:

  • introduce a new Credit Risk: Internal Ratings Based Approach (CRR) Part of the PRA Rulebook (Appendix 2) to replace the CRR articles and associated technical standards that HM Treasury (HMT) plans to revoke and a technical standard that the PRA will revoke;
  • introduce a new Credit Risk: General Provisions (CRR) Part of the PRA Rulebook relating to general provisions for credit risk to replace CRR articles that HMT plans to revoke;
  • amend the Credit Risk Part of the PRA Rulebook and remove the existing Standardised Approach and Internal Ratings Based Approach to Credit Risk (CRR) Part of the PRA Rulebook;
  • introduce a new SS4/24 – Credit risk internal ratings based approach (Appendix 9);
  • introduce a new SS3/24 – Credit risk definition of default (Appendix 8);
  • withdraw SS11/13 – Internal Ratings Based (IRB) approaches; and
  • amend SS17/13 – Credit risk mitigation (Appendix 10).

3.6 The sections below have been structured broadly along the same lines as chapter 4 of CP16/22, covering the main areas where the PRA received comments from respondents as follows:

  • Implementation timelines
  • Permission to use the IRB approach
  • IRB exposure classes and sub-classes
  • Restrictions on IRB modelling
  • Roll-out, permanent partial use and reversion
  • Calculation of RWAs and expected loss (EL)
  • General requirements for use of the IRB approach
  • Definition of default
  • Input floors
  • PD estimation
  • LGD estimation
  • EAD estimation
  • Maturity
  • Specialised lending

3.7 Unless specified otherwise, the near-final rules and near-final SSs are consistent with those in CP16/22 and therefore the PRA considers its analysis of its objectives and ‘have regards’ in CP16/22 with respect to the areas mentioned in the paragraph above remains appropriate.

Implementation timelines

3.8 In CP16/22, the PRA set out its proposed timelines for firms to implement changes arising from the PRA’s proposals relating to the IRB approach. The PRA received ten responses to these proposed timelines.

3.9 Respondents requested that the PRA clarify and provides further information on several aspects of the process and timelines for updating existing permissions, new applications, material model changes and in-flight applications:

  • three respondents requested that the PRA use its resources efficiently and effectively when reviewing IRB model applications;
  • five respondents requested that the PRA clarify its approach and timelines for the repapering of existing permissions, new model applications and material model change applications;
  • five respondents requested that, where firms’ models are not fully compliant with requirements on the implementation date, firms should be able to update existing non-compliance notifications and post model adjustments (PMAs) that are in place for the IRB roadmap;
  • four respondents requested that the PRA clarify which compliance standard would be applied to applications between publication of the near-final rules and the implementation date;
  • two respondents requested that the PRA clarify timelines for existing, non-Basel 3.1-related model applications. One of the respondents argued that a transitional period should be considered given the delayed implementation of hybrid mortgage models;
  • one respondent requested that the PRA monitor the approaches of other jurisdictions regarding their implementation timelines for material model changes and provide guidance to firms on how these approaches could impact their implementation plans; and
  • two respondents requested that the PRA clarify timelines for overseas model approach (OMA) applications and whether these will be updated to align with other regulators’ approaches to Basel 3.1 timelines.

3.10 The PRA has considered the requests for further clarity relating to implementation timings and process and has set out further detail below.

3.11 In line with the process set out in CP16/22, the PRA expects that HMT will save existing IRB permissionsfootnote [46] which the PRA will then vary using its power under section 144G of the Financial Services and Markets Act (FSMA) 2000 where necessary such that the saved permissions will operate in line with the following changes that firms will be required to make with effect from 1 January 2026:

  • all restrictions on the scope of IRB models (eg restrictions on modelling EAD and mandatory use of the SA or the foundation IRB (FIRB) approach, subject to the transition arrangement for equity exposures);
  • all changes to LGD and EAD under the FIRB approach and all changes to the maturity calculation; and
  • all IRB input floors.

3.12 In addition, the PRA expects that it will use its section 144G FSMA 2000 powers to amend saved permissions for roll-out and permanent partial use (PPU) such that the scope of the saved permissions is restricted to make them consistent with the near-final rules. The PRA clarifies that firms will not need to apply to make the changes referred to in the previous paragraph.

3.13 The PRA recognises that firms may wish to apply for permissions relating to the IRB approach under the near-final rules before 1 January 2026. Examples could include:

  • applications to change roll-out plans and PPU permissions to take advantage of additional flexibility in the near-final rules; and
  • applications to revert portfolios to less sophisticated credit risk approaches (eg from the IRB approach to the SA or from the AIRB approach to the FIRB approach).

3.14 The PRA has communicated details of the application process, including when firms would need to apply in order for there to be sufficient time for applications to be determined before 1 January 2026. The PRA notes that firms should discuss their intentions regarding applications with their supervisors.

3.15 As proposed in CP16/22, and in contrast to the changes described above, the PRA will not require that IRB models themselves are fully compliant on 1 January 2026 provided that:

  • firms have an appropriate remediation plan in place that has been agreed with the PRA, including an appropriate date for the implementation of their new models or model changes, or that they demonstrate that the effect of the non-compliance is immaterial; and
  • during this period, firms assess whether a PMA is necessary in order to cover any shortfall in RWAs.

3.16 The PRA is aware that a number of applications relating to the IRB approach (eg new IRB permission applications and material model change applications), may be ‘in flight’ (not determined), when the final rules take effect on 1 January 2026. Any such applications made under the CRR would fall away when the relevant CRR provision is revoked. However, the PRA has set out a process whereby firms can inform the PRA that it wishes for an application originally made under the CRR to instead be treated as an application made under the near-final rules (once they are finalised). The PRA may request that firms submit additional information in support of such applications.

3.17 The PRA expects firms to engage with their supervisors regarding whether, once the PRA has made the final rules, future applications relating to the IRB approach should be treated as having been made under the CRR or these rules, and whether existing applications should be treated as having been made under these rules.

3.18 Where the PRA approves an application made, or treated as having been made, under these rules before 1 January 2026, the approval would not take effect until 1 January 2026. The PRA expects that it will be appropriate in the majority of cases for applications relating to the IRB approach to be made or to be treated as having been made under the near-final rules once these have been finalised.

3.19 The PRA clarifies that it plans to repaper firms’ IRB permission documentation to reflect all changes that will take effect from 1 January 2026.

3.20 The PRA has considered the requests regarding the amendment of existing non-compliance notifications and PMAs that are in place for the IRB roadmap and has made no changes to its proposals because it considers that these are consistent with the respondents’ requests.

3.21 The PRA has considered the request for further clarity on the approval standard that will be applied to IRB applications received before the 1 January 2026 implementation date. The PRA confirms that where an application has been made (and is treated as having been made) under the CRR, the approval standards set out in the CRR will apply. Where an application has been made (or is treated as having been made) under the near-final rules (once finalised), the approval standards set out in those rules will apply, including the PRA’s material compliance approval threshold for IRB applications.

3.22 The PRA has considered respondents’ comments regarding the approaches taken by other regulators and OMA applications. The PRA confirms that its existing OMA policy remains in place and firms can continue to apply to the PRA under this policy. Firms wishing to submit OMA applications should raise any considerations relating to their specific circumstances with their supervisor.

Permission to use the IRB approach

3.23 The PRA proposed a number of changes relating to firms’ permission to use the IRB approach, including:

  • to change the threshold for approval of IRB permission applications from ‘full compliance’ to ‘material compliance’; and
  • to change the standard for approval of IRB model change applications for firms that already have permission to use the IRB approach such that they may be approved where materially non-compliant if, among other things, they reduce the overall level of non-compliance in the firm’s IRB approach.

3.24 The PRA received 12 responses to this section of CP16/22. The substantive issues raised by respondents are set out below.

Standards for IRB application approval for new models and material model changes

3.25 The PRA proposed to change the threshold for approval of IRB permission applications and material model change applications from ‘full compliance’ to ‘material compliance’ with the PRA’s requirements. For this purpose:

  • ‘material compliance’ would mean where any non-compliance is immaterial, meaning it would result in a minimal impact on the quantitative and qualitative aspects of the firm’s IRB approach; and
  • the materiality of the non-compliance would be assessed at model level and in aggregate to help ensure that immaterial non-compliance across multiple models would not become material overall.

3.26 Where firms have already received permission to use the IRB approach, the PRA proposed that firms would continue to be required to notify the PRA where they no longer comply with any of the requirements applicable to the IRB approach. However, the PRA proposed to retain the rule that firms would not be required to present a remediation plan in respect of non-compliance if they can demonstrate that the effect of the non-compliance is immaterial.

3.27 The PRA considered these proposals overall to be more proportionate than the existing approach and would promote a more level playing field between firms with existing IRB permissions and IRB aspirants.

3.28 The PRA received 12 responses to its proposals. All were generally supportive, but some requested clarification on certain aspects.

3.29 Four respondents requested greater detail on the meaning of ‘immaterial non-compliance’. The PRA has decided not to provide any additional detail at this time on the meaning of ‘immaterial non-compliance’ given the idiosyncratic nature of individual firms’ IRB applications. This makes it difficult to provide guidance that would apply to firms generally. The PRA considers it appropriate to maintain flexibility to assess materiality on a case-by-case basis in the context of the firm’s overall compliance level.

3.30 One respondent requested clarification on whether materially non-compliant IRB model change applications would be approved for firms that already have permission to use the IRB approach where approval of the application would reduce the overall level of non-compliance in a firm’s IRB approach, and the firm has a remediation plan in place. The PRA clarifies that this interpretation of the proposal is correct.

3.31 One respondent requested that the PRA revise its proposal that it would approve material model changes that do not comply with the IRB rules provided that the non-compliance is immaterial at the model level and in aggregate. They requested that the PRA defines the aggregate level as covering only the models that are related to the change being sought, rather than all models that the firm has permission to use. Under that alternative definition, the PRA could approve immaterially non-compliant model changes where the firm is materially non-compliant with the IRB rules overall, provided that the non-compliance stems from models that are not related to the change being sought. Having considered the response, the PRA has decided to maintain its proposal. This is consistent with the PRA’s policy intention that overall non-compliance of firms with the PRA’s IRB requirements should not be material. The PRA considers that this is consistent with its primary objective to promote the safety and soundness of regulated firms by ensuring that firms with permission to use the IRB approach continue to comply, to a material extent, with the requirements that apply to that approach.

Other issues relating to permission to use the IRB approach

3.32 Other substantive responses focused on a range of technical issues:

  • Self-assessments of model changes: two respondents requested that the PRA clarify whether the self-assessments that would accompany applications for material model changes should be submitted by a Senior Management Function (SMF) holder. The PRA proposed that, when firms submit applications for permission relating to the IRB approach, or notify the PRA of an extension or change to an IRB model, they should provide the PRA with a self-assessment of whether they would comply with all relevant CRR articles, PRA rules and SS expectations. For the avoidance of doubt, the PRA does not expect this self-assessment to be made by an SMF holder. Rather, the PRA proposed that an SMF holder should submit an annual attestation as to whether the firm’s IRB approach complies with the CRR, PRA rules and PRA SS expectations, and cover the implementation of an appropriate remediation plan where relevant – this proposed expectation has been maintained in the near-final SS4/24 – Credit risk internal ratings based approach.
  • Annual attestations by SMF holders: one respondent argued that the PRA’s proposed expectation that an SMF holder attests to the compliance of the firm’s modelling approaches is onerous where models and external environments do not change year-on-year. The PRA has decided to maintain its proposal and considers that it is a proportionate expectation. The PRA does not expect that the burden on firms and SMF holders would be significant where neither the model nor the external environment has changed since the SMF holder made their last attestation.
  • Documentation requirements for model change notifications: two respondents requested that the PRA does not proceed with its proposal to require firms to submit technical process documents and the reports of firms’ independent review or validation in support of notifications for extensions and changes to IRB models. Those respondents argued that providing this information would be disproportionate to the scale of change being made. The PRA notes that firms are already asked to provide this information in the proforma application for making such notifications to the PRA. The PRA considers that the proposed requirement is proportionate because the information required is only that which is relevant to the change being made. As such, the PRA has decided to maintain its proposal.

PRA objectives and ‘have regards’ analysis

3.33 The near-final policy on permission to use the IRB approach is as proposed in CP16/22. Therefore, the PRA considers that the objectives and ‘have regards’ analysis presented in CP16/22 remains appropriate.

IRB exposure classes and sub-classes

3.34 The PRA proposed a number of definitional changes to existing IRB exposure classes, and to introduce new exposure sub-classes in some cases to improve the clarity of the framework. The PRA received eight responses to this section of CP16/22. The substantive issues raised by respondents are set out below.

Corporates exposure class

3.35 The PRA proposed to define a large corporate for the purpose of assigning exposures to the ‘financial corporates and large corporates’ exposure sub-class as a corporate with consolidated assets equal to or greater than £440 million, or having consolidated annual sales of more than £440 million, or belonging to a group where the annual turnover for the consolidated group is more than £440 million. For the purpose of the PRA’s proposal, consolidated annual sales would be calculated as the average annual amount over the last three years. The PRA considered this proposed definition to be broadly aligned with the Basel 3.1 standards while adding clarification on certain elements.

3.36 Two respondents requested that the PRA clarify the appropriate basis of consolidation to be applied in its proposal that firms should assess the consolidated annual sales and consolidated assets of the obligor. They requested that the PRA confirms that firms would be permitted to use internal measures of these values rather than published financial statements. The PRA has considered the responses and has decided that firms should be required to use audited financial statements. The PRA considers that this will lead to an appropriately robust assessment of obligor size and achieve an appropriate degree of consistency between firms. The PRA has also clarified in its near-final rules that firms will be required to assess the highest level of consolidation that includes the obligor and for which audited financial statements are available. Firms will therefore use the accounting standard that is applicable to those statements.

3.37 One respondent identified an error in the PRA’s proposal that the definition should depend, in part, on the obligor’s consolidated assets being more than £440 million. Having considered this response, the PRA clarifies that the appropriate test is only that the obligor’s annual sales are greater than £440 million. As such, the PRA has amended its near-final rules to remove the reference to consolidated assets being more than £440 million in the definition.

3.38 Two respondents requested that the PRA revise its proposal that consolidated annual sales are calculated as an average amount over the last three years. These respondents argued that the PRA should alternatively permit firms to assess consolidated annual sales as of the most recent year available and argued that this is in line with the Basel 3.1 standards which permit either approach. Having considered the responses, the PRA has decided to maintain its proposal. The PRA considers it desirable for the definition to be calibrated against a three-year moving average in order to reduce volatility in the classification of obligors over time. The PRA has decided not to permit firms to choose whether to use a three-year moving average or the most recent year available in order to increase the consistency of approaches across firms and so that firms do not ‘cherry-pick’ approaches in order to achieve lower risk weights.

3.39 Three respondents argued that the PRA’s proposal to redenominate the threshold for large corporates from Euro to Sterling, in line with the PRA’s approach to currency redenomination, would increase complexity for firms that seek model approvals in multiple jurisdictions. The PRA has decided to maintain its proposal. It considers that it is in line with its overall approach to currency redenomination and would reduce complexity for firms which do not seek model approvals in multiple jurisdictions. Further, the PRA’s near-final rules should reduce the extent to which fluctuations in exchange rate affect the assignment of exposures to exposure sub-classes for firms that operate predominantly in the UK.

3.40 In its near-final rules, the PRA has also amended the definition of large corporates by referring only to the obligor’s annual revenue, rather than to annual sales and annual turnover. The PRA does not consider these terms to have different meanings, but it considers that using a single term will avoid any ambiguity in the framework.

Definition of specialised lending

3.41 Five respondents commented on the PRA’s proposed definition of the ‘specialised lending’ exposure sub-class. The PRA proposed to align with the Basel 3.1 standards by including a condition that the borrowing entity has ‘little or no other material assets or activities’ in the definition of specialised lending. Two respondents supported this proposal in general, but all five respondents raised concerns regarding specific aspects of the definition. Three respondents argued that the proposed definition would lead to inconsistent classification of exposures between firms. One respondent requested that the PRA provide guidance on the aspects of the definition that they considered are open to different interpretations by firms. Two respondents thought that the PRA’s proposal would not significantly impact the set of exposures classed as specialised lending. One respondent was concerned about potential interactions with the definition of securitisation. One respondent argued that other characteristics are more important to the classification of exposures as specialised lending than the ones the PRA proposed.

3.42 Having considered the responses, the PRA has decided to maintain its proposal to align the definition of specialised lending with the Basel 3.1 standards. While the PRA recognises that the proposed change to the definition relative to the CRR may have limited impact, the PRA considers that it is desirable for firms to consider the additional condition when assessing whether an exposure is specialised lending. Indeed, the existence of other assets or activities by which the obligor may repay the obligation makes treating the exposure as a specialised lending exposure less appropriate. The PRA notes that HMT has stated it will update the definition of securitisation in the Securitisation Regulations 2024 such that it aligns with the PRA’s near-final rules. The PRA does not consider that it is necessary to provide further guidance on the definition of specialised lending at this time.

Definition of retail small and medium-sized enterprises

3.43 Five respondents commented on the PRA’s proposed amendments to the criteria for SMEs to be included in the ‘retail’ exposure class. The PRA proposed that the value of the exposure (either individually or when aggregated with all other retail exposures) to a single obligor, excluding exposures secured by residential immovable property, should not exceed £0.88 million to be classed as retail. The PRA also proposed to include undrawn commitments in the threshold and considered that this would align with the Basel 3.1 standards and reduce the likelihood that exposures are reclassified when facilities are drawn down and repaid.

3.44 Four respondents opposed the proposal to include undrawn commitments in the threshold arguing that it would lead to reclassification of obligors currently treated as retail to the ‘corporates’ exposure class. One respondent requested that the definition of retail SMEs should be based on financial turnover rather than the size of a firm’s exposure to the obligor.

3.45 Having considered the responses, the PRA has decided to amend its draft rules regarding the inclusion of undrawn commitments in the assessment of SMEs against the proposed £0.88 million size threshold to align with the CRR approach. Firms should only include the total amount owed to the institution when calculating the size of firms’ exposures to obligors against the threshold. This is aligned with the revised definition of retail SME under the SA as described in Chapter 2 – Credit risk – standardised approach. While this does not fully align with the Basel 3.1 standards, the PRA considers the approach to be appropriately prudent when considered alongside the other criteria for classifying exposures as retail SMEs.

3.46 In response to the proposed threshold level, respondents argued that the £0.88 million threshold is outdated and should be increased, as it is based on a threshold that has not been updated by the Basel Committee on Banking Supervision (BCBS) since 2004. Two respondents also argued that firms should be permitted to use either a £0.88 million or a €1 million threshold level when categorising exposures to avoid inconsistencies within an international group. One respondent requested that special purpose vehicles (SPVs) set up exclusively to hold buy-to-let residential properties should be exempt from the £0.88 million threshold, and instead be subject to a £3 million threshold.

3.47 Having considered the responses, the PRA has decided to maintain its proposed threshold level for retail SMEs. The PRA considers that the £0.88 million threshold is consistent with the Basel 3.1 standards. The PRA does not consider it appropriate for firms to be permitted to use different currencies when calculating the threshold level. This is consistent with the PRA’s general approach to currency redenomination and ensures consistent (in pound sterling terms) thresholds over time and across all PRA-regulated firms. Finally, the PRA does not consider it appropriate to permit an additional carve-out from the threshold level for SPVs set up to hold buy-to-let properties.

Other issues relating to IRB exposure classes and sub-classes

3.48 Other substantive responses focused on a range of technical issues:

  • ‘Central governments and central banks’ exposure class: one respondent requested that the PRA delete the proposed ‘central governments and central banks’ exposure class given that firms would not be permitted to apply the IRB approach to exposures to those obligors under the PRA’s proposals. The PRA has decided to proceed with its proposal. The PRA considers it desirable to define exposure classes for all exposures so that it is clear which approaches firms are permitted to use. The PRA has also revised this exposure class to include exposures to quasi-sovereigns in line with its decision to withdraw the IRB approach for those exposures (see ‘Restrictions on IRB modelling’).
  • Alignment of the ‘qualifying revolving retail exposures’ exposure sub-class with the SA: one respondent requested that the PRA align the proposed ‘qualifying revolving retail exposures’ exposure sub-class with the revolving facilities that might qualify as regulatory retail exposures in the SA. The PRA has decided not to align these definitions. The PRA considers that the definition of the ‘qualifying revolving retail exposures’ exposure sub-class is appropriate given the treatments that apply to those exposures under the IRB approach, which are aligned with the Basel 3.1 standards.
  • ‘Qualifying revolving retail exposures’ exposure sub-class definition: three respondents requested that the PRA clarify its proposal that the ‘qualifying revolving retail exposures’ exposure sub-class is defined as exposures where, among other things, the maximum amount to a single individual is £90,000. They requested clarification on whether this includes undrawn amounts and, if so, whether those amounts should be calculated following the application of conversion factors (CFs). They requested that the PRA clarify whether ‘maximum exposure’ means the largest aggregate exposure to a single individual or the largest single exposure to a single individual. The PRA confirms that ‘maximum exposure’ includes the total notional undrawn amount and that it means the largest aggregate exposure to a single individual. The PRA considers that this clarified definition reduces volatility in the classification of exposures as either qualifying revolving retail exposures or ‘other retail’ as customers draw on their credit limits. The PRA further considers that its decision appropriately classifies exposures as qualifying revolving retail exposures where customers have multiple facilities and is aligned with the Basel 3.1 standards. The PRA has amended its near-final rules to reflect these clarifications.
  • ‘Equity’ exposure class: one respondent requested that units or shares in CIUs should be a separate exposure class rather than included in the ‘equity’ exposure class as the PRA proposed. The PRA considers that this would improve the clarity of the draft rules as it more appropriately reflects differences between CIU exposures and equity exposures. As such, the PRA has adopted this approach in its near-final rules. One respondent requested that the PRA clarify why it proposed the ‘other equity’ exposure subclass given that it had proposed elsewhere to withdraw the IRB approach for exposures to equities. The PRA considers that it is desirable to define exposure classes for all exposures so that it is clear which approaches firms are permitted to use.
  • Treatment of exposures to the Federal National Mortgage Association (Fannie Mae) and the Federal Home Loan Mortgage Corporation (Freddie Mac): two respondents requested that the PRA clarify the appropriate exposure class for exposures to Fannie Mae and Freddie Mac. The PRA has decided that it would not be appropriate to clarify the appropriate exposure class for specific obligors. The PRA considers that the exposure class definitions that it proposed are appropriate, and that it is desirable for firms to consider all exposures against those definitions rather than introduce specific treatments for individual exposures.
  • Suitability of the ‘corporates’ exposure class for exposures not falling into any other class: one respondent argued that the PRA’s proposed ‘corporates’ exposure class is defined too broadly by including any exposure which does not fall into any other class. The PRA has decided not to make any changes to this exposure class because it considers that it captures sufficiently similar exposures and that it is appropriately divided into exposure sub-classes.

PRA objectives analysis and ‘have regards’ analysis

3.49 The PRA considers that its revision to its proposals regarding the criteria for SMEs to be treated as retail will mean that more SME obligors will be treated as retail than under the CP16/22 proposals. The PRA considers that while the revised criteria are less conservative than the CP16/22 proposal, the definition of retail SMEs remains consistent with its primary objective when considered in the round. The PRA considers that its ‘have regards’ and objectives analysis in CP16/22 remains appropriate given this revision.

3.50 Aside from this, the PRA considers that its near-final rules on IRB exposure classes and sub-classes, incorporating the minor changes and clarifications described in this section, are materially aligned with those proposed in CP16/22. Therefore, the PRA considers its analysis of its objectives and ‘have regards’ in CP16/22 remains appropriate.

Restrictions on IRB modelling

3.51 The PRA proposed to implement a number of restrictions on IRB modelling, including:

  • to remove the IRB approach for exposures to central governments and central banks;
  • to remove the AIRB approach for exposures to institutions (including quasi-sovereigns), large corporates and financial corporates; and
  • to remove the IRB approach for equity exposures.

3.52 The PRA received ten responses to this section in CP16/22. The substantive issues raised by respondents are set out below.

Removal of the IRB approach for exposures to central governments, central banks and quasi-sovereigns

3.53 The PRA proposed to prohibit the modelling of central government and central bank exposures under the IRB approach and to require these exposures to be risk-weighted under the SA. The PRA proposed this because it considered that firms face considerable challenges when modelling these exposures, including a lack of default data. The PRA also considered that firms have an insufficient comparative advantage relative to the regulator in assessing the risk of these exposures.

3.54 In contrast, the PRA proposed to define quasi-sovereign exposures as an exposure sub-class within the ‘institutions’ exposure class and to continue to permit firms to model these exposures under the FIRB approach. The PRA considered that firms may have an information advantage over the regulator in relation to assessing the risk of these exposures.

3.55 Three respondents raised concerns about modelability of quasi-sovereign exposures on the basis that they can be just as challenging to model as central government exposures. They argued that these exposures are often guaranteed by or inextricably linked with central governments. One of those respondents argued that the PRA should require exposures to quasi-sovereigns to be risk-weighted under the SA. Two respondents suggested retaining the option to use either the SA or the FIRB approach to determine the risk weights of these exposures. On the basis that the IRB approach would continue to be permitted in respect of quasi-sovereign exposures, one respondent requested that the PRA clarify that quasi-sovereign modelling approaches where central government ratings are material risk drivers would be acceptable. Two respondents argued that estimates of the credit risk of central government exposures are an essential input to firms’ estimates of the credit risk of quasi-sovereign exposures and therefore that the PRA’s proposal that exposures to central governments should not be modelled, but exposures to quasi-sovereigns should, is not workable in practice.

3.56 Having considered these responses, the PRA acknowledges the arguments of respondents that modelling quasi-sovereign exposures presents similar challenges to modelling central government exposures, and is often based heavily on firms’ central government models. Therefore, the PRA has amended the draft rules to prohibit modelling of quasi-sovereign exposures and to require that all quasi-sovereign exposures should be risk-weighted under the SA.

3.57 The PRA accepts that this decision could mean that some firms who would have been able to develop a robust modelling approach for quasi-sovereigns would not be permitted to adopt it. However, given the responses it received, the PRA does not consider this a likely outcome and therefore considers that prohibiting modelling for these exposures is a proportionate and appropriately prudent decision. The PRA considers that this approach advances its primary objective by restricting the use of modelling to calculate capital requirements where this cannot be done in a robust and consistent manner. It will also ensure a level playing field between firms for risk-weighting quasi-sovereign exposures.

3.58 The PRA notes that the impact of this decision on capital requirements will be mixed as it will depend on the differences in risk weights between the SA and those produced by firms’ IRB models. The PRA is concerned that SA risk weights for exposures to central governments, central banks and quasi-sovereigns can potentially result in an underestimation of RWAs. The PRA therefore intends to consult in the future, as part of the wider Pillar 2 review discussed in Chapter 6 – Pillar 2, on a potential Pillar 2 methodology to help ensure the adequate capitalisation of these exposures.

3.59 One respondent raised concerns about the PRA’s proposed withdrawal of the IRB approach for exposures to central governments and central banks given their use to generate the ‘sovereign ceiling’ for exposures to institutions and corporates. The respondent argued that the use of an external rating to generate the sovereign ceiling would be inconsistent with the principle that firms should undertake their own assessments. The PRA clarifies that using an external central government rating as an input into a firm’s institution and corporate rating systems is permissible in principle.

Removal of the advanced IRB approach for exposures to institutions, large corporates and financial corporates

3.60 The PRA proposed to withdraw the AIRB approach for exposures to institutions, large corporates and financial corporates on the basis that firms face challenges developing robust modelling approaches for estimating LGD and EAD for these exposures. The PRA proposed to define large corporates as those which, among other things, have consolidated annual sales of at least £440 million, as noted in ‘IRB exposure classes and sub-classes’, above. This is aligned with the Basel 3.1 standards.

3.61 One respondent argued that the PRA’s proposed threshold would mean that nearly all clients to whom firms provide revolving credit facilities would be defined as large corporates, and therefore firms would no longer be permitted to apply the AIRB approach to those exposures. They argued that FIRB CFs are higher than the CFs that they typically estimate for exposures to revolving credit facilities and that their RWAs for these exposures would increase, reducing their capacity to lend to those customers.

3.62 Having considered this response, the PRA has decided to maintain its proposal. The PRA considers that firms face considerable challenges developing robust modelling approaches for estimating the LGDs and EADs for exposures to large corporates. The PRA observes that firms’ estimated CFs for revolving credit facilities vary significantly based on firms’ processes and modelling approaches, so the impact of the PRA’s near-final policy will vary. However, as noted in Chapter 2 – Credit risk – standardised approach, the PRA has decided to reduce its CF for ‘other commitments’, except those relating to UK residential real estate, from 50% to 40%. The PRA considers that this reduction will limit any impact of its proposals while continuing to advance its primary objective and remaining aligned with the Basel 3.1 standards.

PRA objectives analysis

3.63 The PRA considers that its decision to prohibit the modelling of quasi-sovereign exposures advances its primary objective because it ensures that firms are only permitted to model exposures where they can do so in a robust manner.

3.64 The PRA considers that its decision to prohibit modelling of quasi-sovereign exposures will also advance its secondary objective to facilitate effective competition as all firms will be required to risk weight these exposures under the SA, resulting in less variation in risk weights across firms.

‘Have regards’ analysis

3.65 In developing these near-final rules, the PRA has had regard to its framework of regulatory principles and the matters to which it is required to have regard when proposing changes to CRR rules. The PRA considers its analysis of its ‘have regards’, as presented in CP16/22, remains appropriate, subject to the following updates:

1. Relative standing of the UK as a place for internationally active firms to operate and competitiveness:

  • The PRA considers that its decision to withdraw the IRB approach for exposures to quasi-sovereigns will not materially impact the competitiveness of the UK. The PRA considers that its decision will likely result in lower RWAs for some exposures and higher RWAs for other exposures relative to those jurisdictions that continue to permit the IRB approach. Where other jurisdictions permit firms to permanently apply the SA to these exposures, this impact will be reduced.

2. Relevant international standards:

  • The PRA notes that the Basel 3.1 standards grant discretion to regulators on the extent to which they implement modelled approaches. The PRA does not therefore consider that withdrawing the IRB approach for quasi-sovereign exposures impacts the UK’s alignment with the Basel 3.1 standards.

Roll-out, permanent partial use, and reversion

3.66 The PRA proposed to introduce a number of roll-out classes and to introduce revised rules relating to the permanent partial use of the SA for firms with permission to use the IRB approach. The PRA’s proposals were intended to reduce barriers to entry for smaller firms adopting the IRB approach while limiting the opportunity to ‘cherry-pick’ approaches in order to reduce RWAs without a corresponding reduction in risk. The PRA also proposed to restate the existing criteria for firms to receive permission to revert to less sophisticated approaches in its rules and supplement them with expectations that it would consider one-off costs arising from implementing the requirements proposed in CP16/22 as a relevant factor when it considers whether those criteria are met.

3.67 The PRA received 13 responses to this section in the CP. The substantive issues raised by respondents are set out below.

Permanent partial use of the standardised approach

3.68 The PRA proposed that firms would be required to roll out the IRB approach to all exposures unless the PRA has granted the firm permission to permanently apply the SA. The PRA would grant a firm permission to permanently apply the SA either for all exposures in a roll-out class or for some exposures in a roll-out class.

3.69 For permanently applying the SA for all exposures in a roll-out class, the PRA would grant permission where:

  • the application of the SA does not result in significantly lower capital requirements than if the IRB approach were used;
  • the firm cannot reasonably model the exposures; or
  • the exposures are immaterial (less than 5% of total group credit risk RWAs).

3.70 For permanently applying the SA to some exposures in a roll-out class, the PRA would grant permission provided that the SA was not permanently applied to more than 50% of exposures within a roll-out class, and where:

  • the firm cannot reasonably model the exposures; or
  • the exposures are immaterial (less than 5% of total group credit risk RWAs for that roll-out class).

3.71 The PRA’s proposal to include modelability as a criterion for these permissions was new. The PRA therefore proposed a lower materiality threshold for exposures within a roll-out class of 5% of total group credit risk RWAs for that roll-out class, relative to the existing expectations, in order to ensure that firms apply the SA only in exceptional circumstances given that some additional exposures would be subject to the SA for reasons other than immateriality.

3.72 Five respondents argued that a 5% materiality threshold would be too low for roll-out of the SA within a roll-out class. They argued that it would require some firms to develop new IRB models for portfolios that are currently on the SA, but which they consider to be immaterial.

3.73 Having considered these responses, the PRA acknowledges the issue raised by firms. The PRA has therefore decided to amend its draft rules and redefine the 5% materiality threshold as a cumulative threshold to be applied across all roll-out classes for which firms have permission to use the IRB approach, rather than being restricted to 5% of credit risk RWAs for each roll-out class. This means the total exposures classed as immaterial under this exemption will be limited to 5% of credit risk RWAs for all of the roll-out classes for which the firm has permission to use the IRB approach.

3.74 The PRA considers that this will reduce the operational burden raised by respondents because firms will have more flexibility regarding which portfolios to roll out the IRB approach to, relative to the proposed approach. For example, under the PRA’s proposal, a firm could have two portfolios of the same size in different roll-out classes where one of the portfolios would be deemed immaterial and the other not, as a result of the size of the roll-out classes that those portfolios are in. The PRA’s revised policy will put both of these portfolios on an equal footing, irrespective of the size of the roll-out classes they are in.

3.75 The PRA considered whether increasing the materiality threshold within a roll-out class would be an appropriate alternative method of addressing respondents’ concerns. However, it considered that this approach would not address the fact that the size of portfolios that have to be rolled out would vary depending on the size of the roll-out classes they are in. The PRA was also concerned that setting a higher materiality threshold within roll-out classes may increase opportunities for ‘cherry picking’ of credit risk approaches to an undesirable extent. The PRA therefore considers that its decision to redefine the materiality threshold as a cumulative threshold across all roll-out classes for which a firm has permission to use the IRB approach addresses respondents’ concerns in a manner that is proportionate and in line with its primary objective.

Reversion to less sophisticated credit risk approaches

3.76 The PRA proposed to retain the requirement that firms must obtain permission from the PRA before reverting to less sophisticated approaches (either from the IRB approach to the SA or from the AIRB approach to the FIRB approach). The PRA proposed to retain the following criteria for granting that permission:

  • that the reversion was not proposed in order to reduce capital requirements;
  • that it is necessary on the basis of the nature and complexity of the firm’s total exposures of that type; and
  • that it would not have a material adverse impact on the solvency of the firm or its ability to manage risk effectively.

3.77 One respondent requested that the PRA revise the proposed criteria. They requested that the primary criterion is whether the firm can reasonably model the exposures in question. They requested that the PRA make the same change to its proposed criteria for using the SA under permanent partial use in order to be consistent.

3.78 Having considered this response, the PRA has decided to maintain its proposals because it considers the criteria are important to ensure firms do not switch between credit risk approaches in order to optimise capital requirements over time, or where doing so would have a detrimental effect on solvency or risk management capability. The PRA notes that a firm’s ability to model exposures is a relevant factor when considering whether reversion is necessary on the basis of the nature and complexity of the firm’s total exposures of that type.

Other issues relating to roll-out, permanent partial use, and reversion

3.79 Other substantive responses focused on a range of technical issues:

  • Permanent partial use of SA by roll-out class: one respondent commented on the PRA’s proposal that firms would be permitted to permanently apply the SA for a roll-out class where the exposures in question are immaterial, where an ‘immaterial’ roll-out class was defined as one where total SA RWAs do not exceed 5% of total group credit risk RWAs. The respondent argued that the PRA’s proposal would lead to disproportionate effort for firms to monitor their portfolios and manage exposures to remain below the 5% threshold. The respondent requested that the PRA introduce a lower threshold at which firms would be required to notify the PRA and a higher threshold that would represent a breach of the PRA’s requirements. The PRA notes that, under its near-final rules, remediation plans presented to the PRA following a breach of the materiality threshold could include reducing the exposures in question or applying to the PRA to include the exposures in the firm’s roll-out plan. The PRA has therefore decided not to make any changes to the CP16/22 proposals as it does not consider that the respondent’s request would achieve a material additional benefit for firms.
  • Corporates roll-out class: one respondent requested that the corporates roll-out class should be split in two, with one roll-out class for non-specialised lending exposures to large corporates and financial corporates, and another roll-out class for non-specialised lending exposures to all other corporates. The respondent argued this was necessary because firms would not be permitted to apply the AIRB approach to exposures to large corporates and financial corporates under the PRA’s proposals. The PRA considers that this distinction is not necessary because the PRA’s near-final rules do not distinguish between exposures on the FIRB approach and the AIRB approach for the purpose of rolling out the IRB approach to and within roll-out classes. Exposures subject to both the FIRB and AIRB approaches would therefore count towards a firm’s compliance with the requirement that the firm rolls out the IRB approach to exposures in accordance with the PRA’s near-final rules. As such, the PRA has decided that non-specialised lending exposures to large corporates and financial corporates should be in the same roll-out class as those to all other corporates.
  • Qualifying revolving retail exposures roll-out class: one respondent requested that the PRA revise its proposed requirement for firms to assess the immateriality of qualifying revolving retail exposures with reference to the RWAs of those exposures under the SA for the purpose of the PRA’s proposed roll-out rules. The respondent argued that this approach would be burdensome because firms are not required to distinguish between exposures which are qualifying revolving retail exposures and those which are not under the SA. The PRA considers the definition of qualifying revolving retail exposures to be relatively simple to apply and that it is reasonable, and consistent with its primary objective, to require firms with IRB permissions to assess the materiality of exposures subject to the SA to avoid ‘cherry picking’ of credit risk approaches. The PRA has therefore decided to maintain its proposal.
  • Clarity of near-final rules: one respondent commented that the PRA’s draft rules regarding the conditions for granting permission for a firm to permanently apply the SA for a roll-out class and within a roll-out class were unclear. Having considered the respondent’s feedback, the PRA has amended its draft rules on PPU conditions to make them clearer. To do this, the PRA has made several significant structural changes to the draft PPU requirements by moving all requirements and conditions into one article. The PRA considers that this will make the requirements easier to follow. The PRA has also set out more clearly the different circumstances in which firms might be permitted to permanently apply the SA and the interaction between the PRA’s near-final rules regarding roll-out, permanent partial use of the SA, and sequential roll-out plans. The PRA considers that these changes will enhance firms’ ability to implement the PPU requirements.

PRA objectives analysis

3.80 The PRA considers that its near-final policy on permanent partial use of the SA within a roll-out class, including the amendment to apply the 5% materiality threshold as a cumulative threshold to be applied across all roll-out classes for which firms have permission to use the IRB approach, is materially aligned with its proposals in CP16/22 from the perspective of the PRA objectives. Therefore, the PRA considers analysis of its objectives in CP16/22 remains appropriate.

‘Have regards’ analysis

3.81 In developing the near-final rules, the PRA has had regard to its framework of regulatory principles and the matters to which it is required to have regard when proposing amendments to CRR rules. The PRA considers its analysis of its ‘have regards’, as presented in CP16/22, remains appropriate, subject to the following update:

1. Proportionality:

  • The PRA considers that its approach to permanent partial use of the SA within roll-out classes will reduce the operational burden for firms as they will have greater flexibility as to which portfolios they roll out the IRB approach to. This increases the proportionality of its approach.

Calculation of risk-weighted assets and expected loss

3.82 The PRA proposed to make a number of changes to the calculation of RWAs and EL, including to:

  • remove the 1.06 scaling factor;
  • revise the obligors to which the 1.25 multiplier to the coefficient of correlation (the ‘multiplier’) would apply;
  • remove the SME support factor;
  • remove the infrastructure support factor;
  • clarify the approach to excesses of specific provisions over EL amounts for defaulted exposures; and
  • replace the existing expectation for calculating an ‘unrecognised exposure adjustment’.

3.83 The PRA received 19 responses to this section in CP16/22. The substantive issues raised by respondents are set out below.

Small and medium-sized enterprise support factor

3.84 The PRA proposed to align with the Basel 3.1 standards by removing the SME support factor under the IRB approach. The PRA received 15 responses to the proposal in addition to responses received in respect of the similar SME support factor proposal for the SA, outlined in Chapter 2 – Credit risk – standardised approach. Respondents argued that the proposal:

  • would reduce lending to SMEs and increase the lending rates which SMEs pay to borrow, because changes in the cost of capital would be passed on to obligors;
  • was not justified on prudential grounds, due to a lack of evidence that the SME support factor leads to SME exposures being undercapitalised in a stress; and
  • would damage the international competitiveness of UK firms.

3.85 Respondents also argued that the rationale for removing the SME support factor for SA exposures does not apply to the IRB approach because the PRA did not propose to make corresponding changes to the IRB approach to offset the impact of its removal.

3.86 Respondents argued that the PRA should instead retain the SME support factor. Four requested that it is phased out gradually, or for it to be retained for exposures existing at implementation date (until final repayment of the exposures).

3.87 Having considered these responses, the PRA has decided to maintain its proposal to remove the SME support factor from Pillar 1. The PRA, acknowledging the arguments made by firms, and the potential impact on growth and UK competitiveness, has however decided to apply a firm-specific structural adjustment to reduce Pillar 2A capital requirements (the ‘SME lending adjustment’) to ensure that its removal does not result in an increase in overall capital requirements for SME exposures (see further detail in Chapter 2 – Credit risk – standardised approach).

3.88 As set out in CP16/22, the PRA considers that IRB modelled risk weights should be appropriately risk-sensitive without the SME support factor. They are based on firms’ own data and therefore historic experience of lending to SMEs, and the IRB corporate risk weight function already includes an obligor size adjustment in the correlation parameter which reduces capital requirements for exposures to corporate SMEs relative to larger corporates. Additionally, exposures to SMEs in the IRB ‘retail’ exposure class are risk-weighted using the retail IRB risk weight function which produces lower risk weights for retail SME exposures than for equivalent corporate exposures. The PRA did not receive quantitative or qualitative evidence that persuasively challenged its view that the SME support factor should be removed based on a consideration of the risk alone. Therefore, the PRA continues to consider that the SME support factor should not be kept in Pillar 1 capital requirements.

3.89 While the PRA has concluded that, in the context of wider changes to the SME package, the SME support factor should be removed under Pillar 1, it acknowledges respondents’ concerns on the potential impact of even limited changes in capital requirements for some firms with certain types of SME exposures. To minimise any potential disruption to SME lending and therefore growth resulting from the removal of the SME support factor, the PRA will apply the SME lending adjustment to ensure that the removal of the SME support factor under Pillar 1 does not result in an increase in overall capital requirements for SME lending. The PRA will apply the SME lending adjustment where firms choose to submit the necessary data to the PRA. The PRA considers that this strikes an appropriate balance in delivering a prudent but proportionate approach. Further information on the SME lending adjustment is set out in Chapter 2 – Credit Risk – standardised approach.

3.90 In addition to the SME lending adjustment, the PRA has included a revised definition of SME in the near-final rules. The new definition will require firms to assess the turnover of entities on the basis of the approach to accounting consolidation in the obligor’s jurisdiction. This will result in more exposures being eligible for treatment as SMEs and enable firms to assess whether obligors meet the definition of SME more easily, eliminating the requirement for firms to undertake a bespoke assessment of linked firms beyond that which is captured by accounting consolidation. This PRA has decided also to make this change for the SA, as set out in Chapter 2 – Credit risk – standardised approach.

Infrastructure support factor

3.91 The PRA proposed to align with the Basel 3.1 standards by removing the infrastructure support factor under both the SA and the IRB approach. The PRA received ten responses to its proposal, of which six respondents opposed the removal, three neither opposed nor supported the removal, and one supported the removal. Respondents made the following arguments which are relevant to the IRB approach:

  • six respondents argued that removing the infrastructure support factor would negatively impact the international competitiveness of the UK;
  • five respondents argued that the infrastructure support factor has increased the amount of lending to infrastructure projects and that removing it would increase the price of lending and decrease lending volumes;
  • four respondents requested that if the PRA does not maintain the infrastructure support factor, legacy instruments or transitional arrangements should be put in place in order to limit the impact on existing exposures;
  • six respondents argued that the PRA’s proposal would impact the financing of green infrastructure projects such as the development and operation of renewable energy sources. One respondent requested that alternative measures be put in place to support green and sustainable financing; and
  • one respondent requested that the proposed draft rules align with the ‘Solvency II Corporate Infrastructure Exemption’.

3.92 Having considered these responses, the PRA has decided to make material amendments to its draft policy. It has decided to maintain its proposal to remove the infrastructure support factor under the IRB approach but to apply a firm-specific structural adjustment to reduce Pillar 2A capital requirements (the ‘infrastructure lending adjustment’). It has also decided to introduce a new substantially stronger category in the slotting approach for project finance, which will reduce risk weights relative to the PRA’s proposals (see ‘Specialised lending’).

3.93 As set out above for the SME support factor, the PRA considers that modelled risk weights reflect firm data and should be risk-sensitive and reflect an appropriate level of risk without a support factor. That is the basis on which model approval is given. The PRA did not receive persuasive evidence that modelled risk weights for infrastructure exposures would be inappropriately high in the absence of the infrastructure support factor and does not consider there to be prudential justification for maintaining the infrastructure support factor. The removal of the infrastructure support factor from Pillar 1 capital requirements also aligns with international standards.

3.94 For exposures subject to the slotting approach, the PRA considers, based on evidence provided by respondents and supervisory data available to the PRA, that a more risk-sensitive treatment of project finance is warranted (see ‘Specialised lending’). The PRA considers that this is an important change that will reduce the Pillar 1 impact of removing the infrastructure support factor while ensuring that the criteria for assigning exposures to the slotting categories will result in appropriate risk weights.

3.95 The PRA acknowledges the role that some infrastructure projects could have in supporting the transition to net zero. However, as set out in the 2021 Climate Adaptation Report, the capital framework is not the right tool to address the causes of climate change. Capital requirements should consider the prudential risks associated with such exposures, and current measures of ‘greenness’ are not necessarily reflective of financial risk.

3.96 However, the PRA recognises concerns raised by respondents on the potential impact on growth and UK competitiveness of even limited increases in capital requirements on infrastructure lending due to the removal of the infrastructure support factor. To support growth, competitiveness and minimise any potential disruption to infrastructure lending resulting from the removal of the infrastructure support factor, the PRA will therefore apply the infrastructure lending adjustment to ensure that the removal of the infrastructure support factor under Pillar 1 does not result in an increase in overall capital requirements for infrastructure exposures. The PRA will apply the infrastructure lending adjustment where firms choose to submit the necessary data to the PRA. The PRA considers that this strikes an appropriate balance in delivering a prudent but proportionate approach. Further information on the optional infrastructure lending adjustment is set out in Chapter 2 – Credit Risk – standardised approach.

Unrecognised exposure adjustment

3.97 The PRA currently expects that firms using the AIRB approach should calculate RWAs on a portfolio basis to reflect products or relationships that are not intended to result in a credit exposure, but where there is an EAD nonetheless. This expectation applies where such amounts are material. An example of these products are current accounts without an overdraft facility that are nonetheless permitted to be overdrawn.

3.98 In CP16/22, the PRA proposed to replace and expand the existing expectation for calculating an ‘unrecognised exposure adjustment’. The PRA proposed that firms using the FIRB approach or the AIRB approach would be required to calculate an unrecognised exposure adjustment for exposures that would not otherwise be captured by the EAD framework because either: (a) they were not intended to result in credit exposures; or (b) they are not classified as off-balance sheet items. Point (a) reflects the expectation that currently applies to firms using the AIRB approach. Point (b) was new for all firms.

3.99 The PRA received seven responses to its proposals. They argued that the proposals would create a deviation between the definitions of commitment used in the IRB approach and the SA. They argued that this would create operational complexity and would put firms that apply the IRB approach at a competitive disadvantage. They argued that it was not clear what exposures the PRA intended to capture under ‘credit facilities that would not otherwise be captured as off-balance items’ and that it was not clear how to assess the full range of risks that would lead to these credit facilities being captured. One respondent argued that this lack of clarity made the adjustment complex and that it therefore could be implemented inconsistently across firms.

3.100 Having considered these responses, the PRA acknowledges the arguments raised by respondents on the operational burden and complexity of the proposals. It has therefore decided not to implement the CP16/22 proposals relating to the unrecognised exposure adjustment (other than to restate the existing expectation into rules). This means that firms will only be required to assess whether to apply an unrecognised exposure adjustment to non-credit facilities for exposures that would be subject to the AIRB approach only if they were recognised as an exposure. The PRA considers this is a proportionate requirement given the responses received, and that it will broadly maintain the existing level of safety and soundness achieved.

1.25 multiplier to the coefficient of correlation

3.101 The PRA proposed the following, which it considered was broadly aligned with the Basel 3.1 standards:

  • that the ‘corporates’ exposure class would be divided into exposure sub-classes for ‘specialised lending exposures’, ‘financial corporates and large corporates’ and ‘other general corporates’;
  • that firms would not be permitted to apply the AIRB approach to exposures in the ‘financial corporates and large corporates’ exposure sub-class;
  • to retain the requirement that firms must multiply the coefficient of correlation in the IRB risk weight function by 1.25 for exposures to large financial sector entities and unregulated financial sector entities;
  • to amend the definitions of those types of entities; and
  • to introduce differentiated LGDs in the FIRB approach for exposures to financial corporates and non-financial corporates.

3.102 The PRA received five responses relevant to these proposals:

  • all five respondents requested that the 1.25 multiplier for the coefficient of correlation should not apply to exposures to the treasury entities of non-financial corporate groups and to certain funds;
  • two respondents argued that the proposed withdrawal of the AIRB approach for exposures to large corporates and financial corporates should be extended to all funds even if they do not meet those definitions; and
  • two respondents argued that firms should continue to be permitted to apply the AIRB approach to exposures to ancillary service undertakings.

3.103 Having considered these responses, the PRA has decided to revise its proposals. In respect of treasury entities of non-financial corporate groups, the PRA considers that it is appropriate in some cases for firms to treat these as non-financial corporates. For exposures to entities where the entity’s sole purpose is to provide treasury or risk management services to its internal group, and it does not provide financial services to third parties, the PRA considers it appropriate for firms not to apply the 1.25 multiplier to the coefficient of correlation. Further, in that case, the PRA considers it appropriate for the AIRB approach to remain available for these exposures (subject to the other restrictions on the application of the AIRB approach in the PRA’s near-final rules). The PRA has amended its draft rules accordingly.

3.104 In respect of funds and ancillary service undertakings, the PRA has decided to maintain its proposals. The PRA considers that it is appropriate to treat these obligors as financial corporates where they meet the definition of financial sector entities to which the PRA’s proposals referred in CP16/22. The PRA did not receive any evidence that the correlation of defaults with respect to the economic cycle is any lower for these entities than for other financial corporates, or that firms do not face the same challenges in modelling LGD and EAD for those exposures.

3.105 The PRA considers that its near-final rules, having made the changes described above, are consistent with its primary objective as they better align the treatment in the IRB approach to the underlying risk. The PRA further considers that it is appropriate and proportionate to permit firms to continue to apply the AIRB approach for exposures where firms are reasonably capable of developing compliant modelling approaches. The PRA’s decision regarding treasury entities of non-financial corporate groups is consistent with the international competitiveness and growth of the UK as it will permit firms to apply lower risk weights to certain treasury entities while still maintaining the risk sensitivity of capital requirements.

Other issues relating to the calculation of RWAs and EL

3.106 Other substantive responses focused on a range of technical issues:

  • EL treatment for fair valued items: one respondent requested that EL for fair valued items should be set to zero because expected gains and losses are already reflected through firms’ earnings. The PRA did not propose any changes to the EL treatment of fair valued items as part of CP16/22 and considers that the matter is not impacted by the PRA’s implementation of the Basel 3.1 standards. The PRA has therefore decided not to make any changes to the EL treatment for fair valued items at this time.
  • Excess specific provisions for defaulted exposures: one respondent commented on the PRA’s proposal to restrict the use of excess specific provisions for defaulted exposures to cover the EL of other exposures. The respondent argued that this proposal would be overly conservative and should be removed. Having considered this response, the PRA continues to consider that its proposal would result in an appropriately prudent framework because it would ensure that firms’ CET1 capital reflects the level of risk. The PRA has therefore decided not to make any changes to its proposals on excess specific provisions for defaulted exposures.

PRA objectives analysis

3.107 The PRA considers that its near-final policy relating to the calculation of RWAs and EL, incorporating the amendments to its draft rules as described above, remains appropriate and advances its objectives. The PRA notes the following updates to the analysis it presented in CP16/22 to reflect its decisions:

  • The revisions to its proposed policy regarding the unrecognised exposure adjustment will mean that fewer exposures are captured by the unrecognised exposure adjustment than under the CP16/22 proposals. However, the PRA considers that the revised scope broadly maintains the existing level of prudence and therefore is consistent with its primary objective.
  • The revisions to the unrecognised exposure adjustment will lead to greater alignment between the SA and the IRB approach in terms of what firms are required to capture for their capital requirements. This will help to advance the PRA’s secondary objective.
  • The revision to permit firms not to apply the 1.25 multiplier to exposures to the treasury entities of non-financial corporate groups will support the safety and soundness of firms by aligning risk weights to the risk of exposures. The PRA does not consider that this decision materially impacts its secondary objective.
  • The decision to proceed with the removal of the SME support factor and infrastructure support factor, and to implement the SME lending adjustment and the infrastructure lending adjustment to ensure the removal of the support factors does not result in an increase in overall capital requirements, will advance its primary objective in a more proportionate manner. The PRA’s decision will ensure that Pillar 1 capital requirements are prudent and risk-sensitive, minimise any disruption to SME and infrastructure lending, and maintain an appropriate overall level of capitalisation. Also, the decision to implement the SME lending adjustment and the infrastructure lending adjustment is consistent with its secondary objective given they will be available to all firms with eligible exposures, regardless of the credit risk approach used.

‘Have regards’ analysis

3.108 In developing these near-final rules, the PRA has had regard to its framework of regulatory principles and the matters to which it is required to have regard when proposing changes to CRR rules. The PRA considers its analysis of its ‘have regards’, as presented in CP16/22, remains appropriate, subject to the following updates:

1. Proportionality:

  • The PRA considers that its revisions to its proposal relating to the unrecognised exposure adjustment make the adjustment materially aligned with the existing expectation that applies to firms. The PRA considers that this represents a proportionate requirement for firms.

2. Relative standing of the UK as a place for internationally active firms to operate and competitiveness:

  • The PRA’s revisions to the unrecognised exposure adjustment support the relative standing of the UK by reducing the scope of what the adjustment captures relative to the proposals in CP16/22. This is more in line with other major jurisdictions who are not expected to implement similar adjustments.
  • The PRA considers that its decision to permit firms to apply lower risk weights to ‘substantially stronger’ project finance exposures (see ‘Specialised lending’), will support the competitiveness of UK firms relative to the CP16/22 proposals and further mitigate any perceived negative impact that the removal of the support factors might have on the relative standing of the UK. The PRA’s decision to implement the SME lending adjustment and infrastructure lending adjustment will also support the competitiveness of UK firms by striking an appropriate balance between delivering prudent Pillar 1 capital requirements that align with international standards and maintaining an appropriate overall level of capitalisation.

3. Finance for the real economy:

  • The PRA considers that the decision to maintain the proposed removal of the SME and infrastructure support factors from Pillar 1 capital requirements will have limited impact on the provision of finance for SME and infrastructure lending. However, the SME lending adjustment and infrastructure lending adjustment will support growth and minimise any potential disruption to the provision of SME and infrastructure lending, and therefore the impact on growth, from the removal of the SME and infrastructure support factors.

General requirements for use of the IRB approach

3.109 The PRA proposed enhancements to minimum data requirements for parameter estimation, data maintenance and IRB model governance and validation. The PRA received seven responses raising a range of technical issues regarding the proposals. The substantive issues raised by respondents are set out below:

  • Use of pooled data: two respondents requested that the PRA clarify whether firms would be permitted to use pooled data. In addition, one respondent requested that the PRA clarify whether the use of pooled data would be permitted when firms make adjustments to obligor grade assignment. The PRA notes that its draft rules and near-final rules permit the use of pooled data for PD estimation. Regarding adjustments to obligor grade assignments, the draft rules and near-final rules state that firms should use ‘all relevant data’, which includes pooled data.
  • Margins of conservatism to address unrepresentativeness of data: five respondents requested that the PRA incorporate paragraph 34 of the EBA Guidelines on PD estimation, LGD estimation and treatment of defaulted assets into its near-final SS4/24. That paragraph notes that firms should apply margins of conservatism to address unrepresentativeness of data. The PRA clarifies that it considers that the principle in paragraph 34 is well-established in the CRR, the PRA’s near-final rules and the near-final SS4/24. The PRA’s proposal was not intended to represent any change in policy. However, for the avoidance of doubt, the PRA has decided to incorporate that paragraph into the near-final SS4/24.
  • Role of IRB rating systems in internal processes: one respondent requested that the PRA clarify its proposal that internal ratings should play an essential role in risk management, decision-making processes, credit approval, internal capital allocation and corporate governance functions. They requested that the PRA clarify whether this would preclude a firm from using rating approaches designed for decision-making and risk management purposes as a basis for an IRB rating system. The PRA clarifies that this would not necessarily be the case.
  • Approving differences between established procedure and actual practice: three respondents requested that the PRA clarify its proposed rule that senior management should approve material differences between established procedure and actual practice. They noted that it was not clear how this differed from the proposed requirement that senior management provides notice of any such difference to the management body or a designated committee thereof. The PRA considers that the two proposed rules are distinct, but it has amended its near-final rule to clarify that an appropriate level of senior management must approve material differences between established procedure and actual practice.
  • Application of margins of conservatism (MoCs): one respondent requested the requirement to apply a MoC due to data and modelling limitations be disapplied for challenger banks. The PRA has considered this request and decided that the requirement for an MoC should apply to all firms. Ensuring appropriately robust capital requirements for all firms subject to the IRB approach is necessary to support safety and soundness of firms.
  • Threshold for estimation error: one respondent requested a materiality threshold for the MoC that firms would be expected to apply for general estimation error (the ‘Category C’ MoC). The PRA has decided to maintain its proposal. The PRA considers that it is unnecessary to set a threshold because its proposed expectations already note that firms can set the Category C MoC to zero when the general estimation error is immaterial.
  • ‘Seasoning’ effects and aging of defaults: three respondents requested that the PRA clarify whether the seasoning assessment for LGD estimation would be related to ‘time since default’ and not ‘time on book’ as the PRA proposed. The PRA agrees with the respondents and has decided to amend its expectations in the draft SS – Internal ratings based approach to that effect.
  • ‘Seasoning’ effects and margins of conservatism: three respondents requested that the PRA clarify that an additional MoC would not be required to account for seasoning effects in LGD and PD estimation if firms are able to demonstrate that seasoning effects are already adequately captured by IRB models. The PRA has decided to amend its draft SS – Credit risk internal ratings based approach to clarify that there is no additional expectation for a separate MoC to capture seasoning effects where firms’ IRB models already capture these effects adequately.
  • Key obligor and facility characteristics: one respondent requested that the PRA clarify its proposed requirement that firms collect and store data on key obligor and facility characteristics. They requested that the PRA clarify whether this would include information relating to characteristics that are not model drivers. The PRA clarifies that it did not intend to limit its proposal only to characteristics that firms use as model drivers and that, therefore, other information could be subject to the requirement in principle.
  • Historic data: one respondent argued that the PRA’s proposal to remove a requirement that firms need not give equal importance to historic data if more recent data are a better predictor of loss rates or drawdowns was inconsistent with the PRA’s proposals on the representativeness of data used to calibrate long-run average PD estimation. The PRA does not share this view. For long-run average PD calibration, the PRA considers that it is appropriate for firms to use data from a long and representative time period and not place greater emphasis on more recent data points. For other purposes, however, including for example rank-ordering, a firm might reasonably decide to place greater emphasis on more recent data points if it is appropriate to do so for those purposes. As such, it has decided to maintain its proposal.

PRA objectives analysis and ‘have regards’ analysis

3.110 The near-final rules on the general requirements for use of the IRB approach, incorporating the minor changes and clarifications described above, are materially aligned with those proposed in CP16/22. Therefore, the PRA considers its analysis of its objectives and ‘have regards’ in CP16/22 remains appropriate.

Definition of default

3.111 The PRA proposed to issue a new SS on the definition of default which would replace existing material related to the definition of default in SS10/13 and SS11/13 as well as the European Banking Authority (EBA) Guidelines on the application of the definition of default which firms are expected to comply with under those SSs. The PRA proposed to make minor changes to existing expectations as part of moving them to the new SS that were either consequential or were intended to enhance the coherence and clarity of the framework.

3.112 The PRA received six responses to this section in CP16/22. The substantive issues raised by respondents are set out below.

  • The use of a ‘months in arrears’ payment allocation scheme for determining obligor default: three respondents requested clarification on whether a ‘months in arrears’ payment allocation scheme is permitted when determining if an obligor is in default. The PRA clarifies that this commonly-used approach remains acceptable under its near-final policy and has updated its draft SS – Credit risk definition of default (Appendix 8) to reflect this.
  • Distressed restructuring and forborne non-performing exposures: five respondents highlighted an apparent contradiction in the draft SS – Definition of default, relating to distressed restructuring. The PRA proposed to expect that firms treat forborne exposures as defaulted only where the relevant forbearance measures are likely to result in a diminished financial obligation. In the draft SS – Credit risk definition of default, the PRA proposed to expect that firms verify on a regular basis that all forborne exposures are treated as defaulted. Having considered these responses, the PRA agrees that setting both expectations would be unclear. It has decided to remove the latter expectation because distressed restructurings that do not result in a materially diminished financial obligation need not necessarily be classified as being in default.
  • The definition of credit agreements and informal arrangements: one respondent requested clarification on the PRA’s proposed expectations relating to credit agreements that explicitly allow the obligor to change the schedule, suspend or postpone payments. The respondent requested that the PRA clarify whether this expectation also applies to informal concessionary arrangements. The PRA clarifies that its near-final expectation only applies where a credit arrangement explicitly allows the obligor to change the schedule, suspend, or postpone the payments under certain conditions, and the obligor acts within the rights granted in the contract.
  • Forbearance measures and distressed restructuring: three respondents argued that a reference to Article 47b of the CRR on forbearance measures in the PRA’s draft rules relating to the definition of default is ambiguous, undesirable, and burdensome to implement. Article 47b lists a number of forbearance measures. The PRA proposed that firms would be required to consider these forbearance measures when assessing whether a distressed restructuring has occurred. The PRA notes that it currently expects that a distressed restructuring should be considered to have occurred when those forbearance measures have occurred as set out in paragraph 49 of the EBA Guidelines on the Definition of Default, and that its proposal is consistent with this existing expectation. The PRA has decided to maintain the substance of its proposal because it considers that these forbearance measures indicate that a distressed restructuring has occurred, and it does not consider that they create any additional burden on firms. The PRA has however decided to specify the forbearance criteria in its near-final SS rather than referring to them in its near final rules to make its regulatory framework easier to navigate.
  • Materiality threshold for non-retail exposures past due: four respondents noted an error in the PRA’s proposed rule regarding the materiality of non-retail exposures past due. The PRA proposed that the materiality threshold for these exposures would be £440 million. The PRA notes that the correct threshold is £440. The PRA has amended the draft rule accordingly.

PRA objectives and ‘have regards’ analysis

3.113 The near-final rules on the definition of default, incorporating the minor changes and clarifications described above, are materially aligned with those proposed in CP16/22. Therefore, the PRA considers its analysis of its objectives and ‘have regards’ in CP16/22 remains appropriate.

Input floors

3.114 The PRA proposed to introduce a number of PD, LGD and EAD input floors to act as a backstop for improving comparability, reducing unwanted variability, and reducing the cyclicality of modelled RWAs.

3.115 The PRA received 12 responses to this section in CP16/22. The substantive issues raised by respondents are set out below.

Probability of default input floors

3.116 The PRA proposed to introduce a 0.1% PD input floor for UK retail residential mortgage exposures. This proposal was more conservative than the 0.05% PD input floor for all retail residential mortgage exposures in the Basel 3.1 standards. This reflects the PRA’s longstanding concern that some IRB UK retail residential mortgage risk weights have been falling over recent years, may be too low, and may not fully reflect the potential for losses in unlikely, but plausible, tail scenarios. The PRA considered that the application of the floor would help backstop against model risk and uncertainty in PD models caused by a lack of historical default data for low-risk exposures (ie low-loan to value mortgage loans). In addition, the input floor would limit the extent to which a fall in default rates in benign economic conditions translates to excessively low risk weights.

3.117 The responses received focused on the potential costs to firms. Five respondents argued that a 0.1% PD input floor would be too conservative and would be inconsistent with approaches taken in other jurisdictions. These respondents requested that the PRA align with the Basel 3.1 standards. Two respondents argued that the proposed input floor is not needed because other measures are sufficient to address issues with declining residential mortgage risk weights. One respondent argued that the proposed input floor is not needed because introducing stricter requirements than the Basel 3.1 standards would imply that the UK retail residential mortgage market is riskier than those in other jurisdictions. One respondent argued that the input floor should be inserted into the SS rather than created as a rule.

3.118 Having considered these responses, the PRA has decided to retain the proposed 0.1% PD input floor for UK retail residential mortgage exposures. The PRA did not receive any persuasive evidence that the proposed input floor is inappropriately calibrated. The PRA considers that its near-final rule best advances safety and soundness by ensuring an appropriate level of conservatism for the estimation of the PD for UK mortgage exposures, while maintaining risk sensitivity in firms’ models. The PRA has decided to retain the PD input floor in its near-final rules to ensure consistent application across firms. The PRA considers that the 0.1% input floor will also advance the PRA’s secondary objective by acting as an additional safeguard against an excessive gap between IRB and SA risk weights for the lowest-risk mortgage exposures.

Other issues relating to input floors

3.119 Other substantive responses focused on a range of technical issues:

  • LGD input floors for unsecured retail exposures with recourse to a broker: one respondent requested that the PRA revise its proposed 30% LGD input floor for certain unsecured retail exposures with recourse to a broker. They argued that the floor would be inappropriately high in that case. The PRA has decided to maintain its proposal on the basis that 30% represents an appropriate backstop for unsecured retail LGD models that broadly reflects the risk of this category of exposure and is in line with the Basel 3.1 standards. The PRA considers that it would be excessively complex to introduce different floors on a granular product basis.
  • LGD input floors for retail exposures secured by residential immovable property: one respondent requested that the PRA clarify whether its proposed 5% LGD input floor for retail residential mortgage exposures would apply to all exposures in the ‘retail exposures secured by residential immovable property’ exposure sub-class. The PRA clarifies that this is the case. The PRA has updated the draft rule accordingly.
  • Interaction of PD floors with 1.25 asset value co-efficient of correlation multiplier: one respondent requested that the PRA disapply the 1.25 multiplier for obligors whose estimated PDs fall below the proposed 0.05% PD floor to avoid having a double impact on highly-rated exposures. The PRA has decided to maintain its proposals. The PRA notes that the 1.25 multiplier is intended to capture the systemic risk associated with exposures to ‘large financial sector entities’ and ‘unregulated financial sector entities’ while the PD floor is intended to address model risk uncertainty and to limit the extent to which a fall in default rates translates to excessively low risk weights. The PRA therefore considers that having both requirements is important to advance its primary objective.
  • PD input floors for qualifying revolving retail exposures: five respondents requested that the PRA correct errors in CP16/22 and draft rules on the value of PD input floors for qualifying revolving retail exposures. In CP16/22, the PRA proposed that the PD input floor for qualifying revolving retail exposures which are transactors would be 0.1%. The PRA has corrected this error and has specified PD input floors of 0.05% for qualifying revolving retail exposures which are transactors and 0.1% for those which are revolvers.
  • Complexity of input floors: two respondents argued that the PRA’s proposed input floors would introduce unnecessary complexity given other proposed backstops such as portfolio-level output floors and the output floor adjustment. One of those respondents argued that the PRA’s proposal not to align with the Basel 3.1 standards in all cases adds to the complexity of the framework. The PRA continues to consider that its proposed approach to input floors is necessary to ensure an appropriate backstop against modelling risk. As such, the PRA has decided to maintain its proposals.

PRA objectives and ‘have regards’ analysis

3.120 The near-final rules on input floors, incorporating the minor changes and clarifications described above, are materially aligned with those proposed in CP16/22. Therefore, the PRA considers its analysis of its objectives and ‘have regards’ in CP16/22 remains appropriate.

Probability of default estimation

3.121 The PRA proposed to make a number of changes to existing approaches relating to PD estimation, including:

  • prohibiting the use of continuous rating scales; and
  • prohibiting the use of adjustments to the assignment of obligor grades to recognise support that is not in writing.

3.122 The PRA received seven responses to this section in CP16/22. The substantive issues raised by respondents are set out below.

Use of continuous rating scales

3.123 The PRA proposed to prohibit the use of continuous rating scales in PD models and to require firms to use discrete rating scales instead. The PRA considered that this proposal was aligned with the Basel 3.1 standards. The PRA observed that, for PD estimation, continuous rating scales are used relatively infrequently by firms and that they typically result in lower RWAs than discrete rating scales in a manner that is not always justified.

3.124 The PRA received five responses requesting that the PRA continue to permit the use of continuous rating scales in PD estimation. Three respondents argued that firms should only be required to use discrete rating scales for residential mortgage lending and that the PRA should continue to permit firms to use continuous rating scales for all other exposures. Two respondents argued that the use of discrete rating scales could lead to a material loss of risk sensitivity and that the perceived risk of RWA underestimation is addressed by other measures such as the PD input floor. One respondent argued that firms can achieve robust risk differentiation for low PD exposures using continuous rating scales.

3.125 Having considered these responses, the PRA has decided to maintain its proposal to prohibit the use of continuous rating scales. The PRA continues to consider that the increase in risk differentiation that can be achieved with continuous rating scales is not sufficient to justify the lower risk weights that they generate. The PRA did not receive any persuasive evidence to the contrary.

Obligor grade adjustments

3.126 The PRA proposed that adjustments to obligor grade assignments to reflect support arrangements would only be permitted where the support is in writing. The PRA considered the proposed approach to be broadly aligned with the Basel 3.1 standards. The PRA considered that undocumented support arrangements can be unclear and not robust, that it is difficult for firms to demonstrate a reduction in default risk from these arrangements, and that it is difficult for supervisors to challenge whether these arrangements are appropriately reflected in IRB models.

3.127 Respondents did not support the PRA’s proposals and argued that the PRA should continue to permit firms to make adjustments to obligor grade assignments where there is undocumented support. Two respondents raised concerns regarding international competitiveness since some other jurisdictions do not require that parental support is documented. One respondent requested that the PRA clarify the meaning of ‘documented’ to ensure consistency among firms. One respondent submitted data showing that obligors with undocumented support have lower observed default rates than those without any support arrangements in place but higher default rates than those with documented support in place.

3.128 Having considered these responses, including the data and impact analysis demonstrating the historical effect of undocumented support arrangements, the PRA has decided to amend its proposal and permit undocumented support being reflected in adjustments to obligor grade assignments. The data provided to the PRA were particularly valuable in adjusting the PRA’s proposals. The PRA has decided instead to introduce risk management expectations in its near-final SS4/24, to ensure a robust and broadly consistent method of support recognition across firms. The PRA considers that this will be a more proportionate way to address the prudential risk previously identified and that it will also support the competitiveness of UK firms in third country markets.

Other issues relating to PD estimation

3.129 Other substantive responses focused on a range of technical issues:

  • Notched grading: two respondents requested that the PRA clarify whether its proposal to prohibit the recognition of undocumented support for the assignment of obligors to rating grades means that using notched grading approaches, where firms notch down from the parent rating or notch up from the subsidiary rating, would also be prohibited. Noting the PRA’s decision to not maintain its proposal to prohibit the recognition of undocumented support for the assignment of obligors to rating grades, the PRA nevertheless clarifies that notched grading approaches are acceptable in principle. As part of the risk management expectations relating to the recognition of undocumented support, the PRA has introduced an expectation that firms should produce a standalone rating for obligors subject to a notched approach.
  • Mix of good and bad periods (macroeconomic data): five respondents requested that the PRA clarify whether the use of macroeconomic data is permitted to meet the requirement that the data firms use for parameter estimation is representative of a mix of good and bad periods. The PRA considers that its draft rules are clear that internal, external, and pooled data can all be used and notes that firms can use a combination of these data types to meet minimum data requirements (eg by using external industry-level data to backcast internal default rates). The PRA therefore considers that no further clarification is necessary.
  • Mix of good and bad periods (risk quantification and rank-ordering): one respondent requested that the PRA clarify its proposed requirement that firms should estimate PDs from long-run averages of one-year default rates over a representative mix of good and bad periods. They requested that the PRA clarify whether this requirement would apply just to risk quantification or whether it would apply also to rank-ordering. The PRA clarifies that it proposed the former, which is aligned with the Basel 3.1 standards. The PRA considers that its near-final rules are sufficiently clear on this point.
  • Guidance on representative periods for UK portfolios: two respondents requested that the PRA specify the years that constitute good and bad periods for UK portfolios. The PRA has decided not to specify those periods. The PRA considers that it is important for firms to be required to identify periods that are representative for their own exposures.
  • Count-weighted long-run average PD calculation and default rate calculation: three respondents requested that the PRA clarify its proposed rules regarding calculating the one-year observed default rate and long-run average default rate. The respondents argued that the PRA’s draft rules conflated the calculation of a one-year default rate with the calculation of long-run average default rates. To address any lack of clarity, the PRA has decided to not include the relevant text in its near-final rules as it considers that these concepts are covered elsewhere. The one-year default rate is defined in the CRR Article 4(1)(78) and the long-run average default rate is covered by paragraph 11.15 of the near-final SS4/24.
  • Long-run average PDs for low default portfolios: one respondent argued that the PRA’s proposed requirement for the calculation of count-weighted average PDs would not be feasible for low default portfolios. The PRA does not share this view. The PRA considers that it is appropriate to require firms to calculate long-run average PDs for these portfolios in the same way as for other portfolios. The PRA considers that it is appropriate to expect firms to apply statistical techniques when calculating PDs for low default portfolios in line with its near-final SS4/24.
  • Use of continuous rating scales in the calculation of the long-run average (LRA) default rate: one respondent requested clarification on how firms should calculate LRA PD for unsecured retail exposures as these PDs are typically estimated on a continuous rating scale. The PRA considers that it is not necessary to provide further guidance in this area because it considers its rules are clear, and because it considers that calculating LRA PD for discrete rating systems is a relatively well-established practice among firms.
  • ‘Point in time (PiT) plus buffer’ methodology: four respondents requested clarification on whether the use of ‘point in time plus buffer’ methodology would be permitted, assuming the PRA prohibits the use of continuous rating scales as proposed. The PRA confirms that it has not changed its approach to rating and calibration philosophy from that set out in PS11/20. The PRA continues to consider that for retail exposures other than residential mortgages, firms may choose to develop models that use dynamic recalibration in conjunction with a suitable buffer to achieve a point-in-time approach.

PRA objectives analysis

3.130 The PRA considers that its near-final policy relating to undocumented support arrangements is consistent with its primary objective. The PRA considers that the evidence it received in response to CP16/22 demonstrates that it can be appropriate for firms to reflect these arrangements in underlying risk weights. The PRA’s new risk management expectations will require that firms make a comprehensive assessment of support arrangements before recognising them in an IRB model. This advances safety and soundness without imposing a disproportionate impact on firms, including with respect to their international competitiveness.

3.131 The PRA recognises that firms that apply the IRB approach are permitted to recognise a broader range of third-party support in capital requirements than firms that apply the SA. Therefore, in addition to advancing the PRA’s primary objective, the PRA’s proposal in CP16/22 was intended to advance its secondary objective by aligning more closely the situations in which firms using different credit risk weighting approaches are permitted to recognise third party support. Notwithstanding this, and on the basis of the responses received, the PRA considers that its near-final policy facilitates effective competition by only recognising the impact of undocumented support where it is robust, based on the safeguards set out in the PRA’s rules and expectations for the IRB approach.

3.132 For all other aspects of PD estimation, the PRA considers its analysis of its objectives as presented in CP16/22 remains appropriate.

‘Have regards’ analysis

3.133 In developing these near-final rules, the PRA has had regard to its framework of regulatory principles and the matters to which it is required to have regard when proposing changes to CRR rules. The PRA considers its analysis of its ‘have regards’, as presented in CP16/22, remains appropriate, subject to the following updates:

1. Relative standing of the UK as a place for internationally active firms to operate and competitiveness:

  • The PRA considers that its decision to permit the recognition of undocumented support arrangements supports the international competitiveness ‘have regard’ by more closely aligning the UK approach with that of other jurisdictions.

2. Proportionality:

  • The PRA considers its decision to permit firms to make adjustments to obligor grades to reflect undocumented support, but to apply additional supervisory expectations to ensure it is robust, is proportionate. It ensures that the burden imposed on firms is proportionate to the risks of reflecting the benefit of insufficiently robust arrangements.

Loss given default estimation

3.134 The PRA proposed to make a number of changes to existing approaches relating to LGD estimation, including to:

  • reduce the applicable LGD under the FIRB approach for unsecured exposures to corporates that are senior claims from 45% to 40%;
  • clarify the eligibility of collateral in the LGD modelling collateral method and introduce an alternative methodology;
  • clarify the eligibility requirements for recognising unfunded credit protection (UFCP) in the LGD adjustment method;
  • clarify the applicable LGD when applying the parameter substitution method; and
  • remove the wholesale LGD framework.

3.135 The PRA received seven responses to this section in CP16/22. The substantive issues raised by respondents are set out below.

Loss given default modelling collateral method for firms using the AIRB approach

3.136 The PRA proposed that firms applying the LGD modelling collateral method would be required to exclude recoveries from ineligible and disregarded eligible collateral when calculating unsecured LGD. The PRA also proposed that firms record the source of recoveries. The rationale behind its proposal was that if cashflows from ineligible and disregarded eligible collateral were included in LGD estimates, it would effectively negate the classification of the collateral as ineligible or disregarded, and consequently bias LGD estimates.

3.137 Three respondents argued that exclusion of cash flows associated with ineligible and disregarded collateral would bias the real recovery amount and hence increase LGD estimates. Respondents argued that, while these recoveries come from ineligible collateral, they are still realised and not reflecting them in LGD estimates would be overly conservative. The PRA did not receive responses on the proposed requirement to record the source of recoveries.

3.138 Having considered these responses, the PRA has decided to broadly retain the proposals in CP16/22. The PRA considers that recoveries from ineligible collateral are less certain and there is prudential risk in permitting recoveries from them to be reflected in LGD and RWA calculations. However, the PRA also notes that firms may not always be able to identify the source of recoveries. The PRA has therefore decided to proceed with its proposal to require firms to exclude recoveries from ineligible and disregarded collateral when calculating unsecured LGD. The PRA has also decided to introduce a new expectation that firms’ policies on the treatment and allocation of unidentified cash flows should consider whether cash flows are classified as arising from recognised collateral or unfunded credit protection, unrecognised collateral or unfunded credit protection, or from other sources, and should do this in a way that results in LGDs being estimated in a conservative manner.

Proposed introduction of an alternative methodology

3.139 The PRA proposed to require the use of an alternative methodology if a firm is using the LGD modelling collateral method to recognise the effect of collateral and does not have sufficient data to model the effects of the collateral in a particular jurisdiction. The PRA proposed to set an expectation that the data would be considered insufficient where the firm has fewer than 20 relevant data points for any non-financial collateral that the firm wishes to recognise in its LGD models.

3.140 Three respondents commented on the PRA’s proposal. Respondents noted that the expectation was omitted from the draft SS - Credit risk internal ratings based approach, in error. The PRA has corrected this omission in the near-final SS4/24. The near-final SS4/24 also includes expectations regarding whether recoveries in a different jurisdictions are relevant for the purpose of the 20 relevant data points. The PRA has made minor drafting amendments to improve the clarity of its near-final rules relating to the cases where the alternative methodology applies.

Other issues relating to LGD estimation

3.141 Other substantive responses focused on a range of technical issues:

  • Probability of possession given default (PPGD) reference points: two respondents requested changes to the PRA’s expectations relating to the reference points that it considers are appropriate when estimating the PPGD of UK residential mortgage exposures where firms have limited data. The PRA did not propose any changes to these reference points as part of CP16/22 and considers that their suitability is not impacted by the PRA’s implementation of the Basel 3.1 standards. The PRA has therefore decided not to make any changes to these reference points at this time.
  • Assuming zero costs and recoveries for incomplete workouts: three respondents requested clarification on whether firms are permitted to assume zero costs and recoveries for incomplete workouts when estimating Best Estimate of Expected Loss (BEEL). The PRA clarifies that an assumption of zero costs and recoveries should not be used when estimating BEEL because conservatism should not be used in the calculation of BEEL, and the inclusion of this assumption would add an element of conservatism.
  • Recovery periods for incomplete workouts: three respondents requested that the PRA clarify whether its proposed retention of an existing expectation that firms should not estimate costs and recoveries for incomplete workouts after the maximum period of the recovery process for the purpose of long-run average LGD estimation is consistent with the statements it made at the IRB mortgages roundtable that it held on 5 October 2020. The PRA understands that this comment relates to an observation made by the PRA that firms may make a conservative assumption in PPGD modelling that all exposures unresolved at the end of the outcome period enter possession. The PRA therefore considers that there is no inconsistency between its proposals and the material discussed at the mortgage roundtable.
  • Long-run average LGD estimation: one respondent requested that the PRA permits calibration at the calibration segment level for discrete LGD model estimates when only a portfolio level calibration target is available. The PRA agrees that this is an appropriate approach and has amended its draft SS – Credit risk internal ratings based approach, accordingly.
  • Recognising collateral: one respondent requested clarification on the PRA’s draft rule that firms using the LGD modelling collateral method to recognise collateral must have internal requirements that are generally consistent with those of the foundation collateral method. The PRA did not propose any substantive changes to this requirement as part of CP16/22 and considers that its operation is not materially impacted by the PRA’s implementation of the Basel 3.1 standards. The PRA has therefore decided not to provide further clarification of this requirement.
  • Separation of general expectations and model development expectations: two respondents highlighted that the chapter on model development in the draft SS – Credit risk internal ratings based approach, includes expectations that should apply to both model development and model calibration. The PRA has adjusted the section headings in its near-final SS to clarify this point. The PRA considers that this adjustment in the naming of sections will provide further clarity to users navigating the near-final SS.
  • The use of house price indices as an economic indicator: one respondent requested clarification on the PRA’s proposal to retain requirements regarding the inclusion of house price indices as a relevant economic indicator when estimating LGD and EAD. The PRA has decided not to provide further clarification in its near-final rules as it considers that they are sufficiently clear.
  • Appropriate discounting rate for cashflows in downturn LGD calculation: one respondent noted an apparent error in the PRA’s proposed expectations. The PRA proposed that, when estimating downturn LGD, firms would be expected to apply a discounting rate of at least the higher of 9% and the Sterling Overnight Index Average (SONIA) plus 5%. The PRA agrees that this was an error and has corrected it and clarified that, for the purpose of downturn LGD, the PRA expects firms to use a discount rate of at least 9%.
  • Supervisory LGD for covered bonds: one respondent requested that the PRA extend the application of the 11.25% supervisory LGD that it proposed for certain covered bonds issued in the UK to covered bonds issued outside the UK. The PRA has decided to maintain its proposal. The PRA considers that this approach means that only high-quality covered bonds are eligible for the lower LGD, which is important for safety and soundness and is consistent with the approach described in Chapter 2 – Credit risk – standardised approach.
  • Treatment of funded credit protection securing unfunded credit protection: as set out in Chapter 4 – Credit risk mitigation, the PRA has decided to specify in its near-final rules the treatment of funded credit protection securing unfunded credit protection obligations. As a result, the PRA has made minor amendments to the draft IRB rules to address how firms applying the AIRB approach may treat this type of credit risk mitigation in their LGD models. The PRA has also introduced some expectations on this situation in its near-final SS4/24. The PRA considers that these changes improve the clarity, and will support a consistent implementation, of its policy.

PRA objectives and ‘have regards’ analysis

3.142 The near-final rules on LGD estimation, incorporating the minor changes and clarifications described above, are materially aligned with those proposed in CP16/22. Therefore, the PRA considers its analysis of its objectives and ‘have regards’ in C16/22 remains appropriate.

Exposure at default estimation

3.143 The PRA proposed to make a number of changes relating to EAD estimation, including to:

  • align the CFs in the FIRB approach with the proposed revised CFs that would apply under the SA;
  • restrict the scope of EAD modelling to revolving commitments in the form of revolving loan facilities only;
  • prohibit modelling of EAD under the slotting approach;
  • require that firms use a 12-month fixed-horizon approach for EAD modelling;
  • permit firms to recognise post-default additional drawings in either EAD or LGD for non-retail exposures as well as for retail exposures, remove an existing PRA expectation that additional drawings beyond a 12-month time horizon need not be incorporated in model estimates, and clarify that pre-default additional drawings would be required to be reflected in EAD estimates; and
  • remove the wholesale EAD framework.

3.144 The PRA received seven responses to this section in the CP. The substantive issues raised by respondents are set out below.

Exposure at default modelling of on-balance sheet items

3.145 Three respondents requested clarification that firms can model EAD for certain on-balance sheet exposures, such as fully drawn down retail mortgage loans under the PRA’s proposals. The PRA clarifies that its near-final rules do not permit EAD modelling for any on-balance sheet exposures which are unconnected to a revolving facility. The PRA considers that this will lead to greater consistency and simplicity in the EAD framework because firms will not be required to switch from not modelling non-revolving commitments before they are drawn down to modelling the exposure once it moves onto the balance sheet. It also ensures a consistent approach to modelling by product type.

Requirement for a 12-month ‘fixed-horizon’ approach for exposure at default modelling

3.146 Four respondents commented on the PRA’s proposal to prohibit the use of the ‘cohort approach’ to EAD modelling and to require use of the fixed horizon approach. Three of these respondents argued that the PRA’s proposal may cause inconsistency between PD, LGD and EAD estimation which could impact RWAs and estimates of expected loss. One respondent argued that the PRA’s proposals would be difficult to implement.

3.147 Having considered these responses, the PRA has decided to maintain its proposal. It considers that the benefits of standardising the time horizon for EAD modelling across firms domestically and internationally in reducing risk weight variability among firms and aligning with the Basel 3.1 standards are justified despite the concerns raised by respondents.

Exposures for which exposure at default or a conversion factor must be estimated

3.148 Four respondents commented on the proposed revisions to the CRR definition of CF set out in HMT’s consultation – Implementation of the Basel 3.1 standards, and highlighted their concern that the proposed definition, when read in conjunction with the PRA’s draft rules, may imply that CFs would be multiplied by an off-balance sheet amount calculated with reference to the unadvised limit, rather than the advised limit, which could result in materially higher capital requirements. Respondents also argued that calculating capital requirements with reference to unadvised limits would be inconsistent with the PRA’s proposed definition of commitment, given the proposed definition refers to contractual facilities which are offered and accepted by the obligor.

3.149 The PRA did not intend to require firms using non-modelled approaches to apply CFs to unadvised limits. The PRA notes that HMT has stated that it will remove the concept of unadvised limits from the CRR definition of CF, if it is to remain in legislation at the point the PRA’s near-final rules take effect (such that the updated definition would also take effect at this point).footnote [47]

3.150 The PRA has considered the position for exposures where EAD or CFs would be modelled under the AIRB approach. For these exposures, the PRA clarifies that it expects firms’ models to capture the ‘balance at default’ regardless of whether this arises from an advised or unadvised limit. Therefore, while CFs should be measured as a proportion of the advised limit only, the entire balance at default should be reflected in realised and modelled EADs and CFs regardless of whether the balance arose from an advised or unadvised limit. This may result in CFs exceeding 100%. The PRA considers this is appropriate as modelling under the AIRB approach is intended to be more risk-sensitive than the SA. The PRA has amended its draft SS – Credit risk internal ratings based approach accordingly.

3.151 As set out in Chapter 2 – Credit risk – standardised approach, six respondents requested that the PRA revise its proposed definition of commitment to exercise the national discretion in the Basel 3.1 standards to permit firms to exempt from the definition of commitment certain arrangements in relation to corporates and SMEs.footnote [48] As set out in Chapter 2 – Credit risk – standardised approach, the PRA has decided to maintain its proposed definition of commitment. The PRA has, however, decided to amend its draft SS – Credit risk internal ratings based approach, to delete a reference relating to the circumstances in which an EAD or CF should be estimated, which was inconsistent with this definition of commitment.

PRA objectives and ‘have regards’ analysis

3.152 The near-final rules on EAD estimation are materially aligned with those in CP16/22 and therefore the PRA considers its analysis of its objectives and have regards in CP16/22 remains appropriate.

Maturity

3.153 The PRA proposed a number of clarifications to its rules regarding the calculation of maturity. The PRA received six responses to this section in the CP. The substantive issues raised by respondents are set out below.

Calculation of maturity

3.154 The PRA proposed to require firms applying the FIRB approach to calculate effective maturity rather than applying fixed parameters. This proposal was supported by one respondent. Three respondents requested that the PRA clarify the treatment that would apply to certain exposures with indeterminate tenor. The PRA can clarify that these would be subject to the general five-year maturity cap because it does not consider that it would be prudent to apply a lower maturity cap for these exposures.

Other issues relating to maturity

3.155 Other substantive responses focused on a range of technical issues:

  • Purchased receivables: four respondents recommended that purchased receivables should be eligible for the proposed one-day maturity floor that would apply to exposures meeting certain conditions, rather than being subject to a one-year floor in every case. The PRA notes that the proposed change to the maturity floor for purchased receivables was not intended to override the provisions surrounding the application of the one-day floor. In cases where purchased receivables meet the relevant conditions, they will still be subject to a one-day floor. The PRA has made minor drafting changes in the near-final rules to clarify this.
  • Scope of one-day floor: two respondents requested that the PRA extend the scope of the one-day floor. They requested in particular that this should apply to inter-bank deposits and nostro accounts. The PRA has decided not to change the scope of the one-day floor because it is not intended to apply to all short-term exposures, and the PRA considers that even if a firm is permitted to withdraw an inter-bank or nostro deposit, it may choose not to do so for business or practical reasons, meaning it would be imprudent to apply a one-day floor.
  • Secured lending transactions: two respondents commented that the PRA’s proposed amendments relating to secured lending transactions, which referred to daily margining, appeared to be inconsistent with the proposed definition (in the Credit Risk Mitigation Part) of a secured lending transaction as a transaction without daily margining. The PRA has amended the near-final rules to address this inconsistency by removing the tailored maturity treatment for secured lending transactions.

3.156 In preparing the near-final rules, the PRA has made other minor changes to improve the clarity of the requirements, in particular relating to the calculation of effective maturity for exposures subject to a master netting agreement and the interaction between the provisions regarding the maturity calculation for different transaction types and the provisions regarding the application of a one-day floor.

PRA objectives and ‘have regards’ analysis

3.157 The near-final rules on maturity, incorporating the minor changes and clarifications described above, are materially aligned with those proposed in CP16/22. Therefore, the PRA considers its analysis of its objectives and ‘have regards’ in CP16/22 remains appropriate.

Specialised lending

3.158 In CP16/22, the PRA proposed to make a number of changes to the IRB approach for specialised lending exposures including the slotting approach. This included changes to the definition of specialised lending and changes to introduce additional risk-sensitivity to the slotting approach.

3.159 The PRA received 13 responses to this section of the CP. The substantive issues raised by respondents are set out below.

High volatility commercial real estate

3.160 The PRA proposed to introduce a new specialised lending category of high volatility commercial real estate (HVCRE), which would be subject to higher risk weights than other specialised lending categories and would align with the Basel 3.1 standards. The PRA proposed to apply a definition of HVCRE that was based on the definition in the Basel 3.1 standards, with some adjustments, with the intention of reducing the scope for inconsistent interpretations across firms. The PRA considered that the introduction of an HVCRE category would promote the safety and soundness of firms by increasing the risk sensitivity of the specialised lending framework.

3.161 Five respondents opposed the introduction of the HVCRE category and ten respondents provided feedback on the proposed definition of HVCRE. Regarding the proposed introduction of an HVCRE specialised lending category, respondents argued that not introducing HVCRE would maintain alignment with the expected EU approach, thereby supporting international competitiveness and simplicity for firms that operate in both jurisdictions. Respondents argued that the additional complexity would not be justified by its impact on capital requirements and argued that the increased risk associated with HVCRE exposures could be reflected by capturing the risk drivers of HVCRE in the slotting criteria.

3.162 Regarding the proposed definition of HVCRE, respondents argued that the definition is complex and subjective. They argued that the application of the HVCRE treatment to certain exposures, in particular residential real estate exposures, would not be appropriate. Respondents argued that if the PRA does introduce the HVCRE category, then it should simplify the definition, align it more closely with the definition in the Basel 3.1 standards, and permit firms to recognise other regulators’ definitions of HVCRE.

3.163 Having considered these responses, the PRA has decided to proceed with the introduction of an HVCRE category, but to amend the definition to align more closely with the Basel 3.1 standards in line with the responses received. The revised definition clarifies that HVCRE only includes commercial real estate and real estate under development where the ultimate use will be commercial. The PRA has introduced further expectations on the application of this definition in the near-final SS4/24, including the types of exposure that the PRA considers exhibit higher volatilities in the UK and in other jurisdictions with similar commercial real estate markets and planning systems. The PRA continues to consider that the introduction of HVCRE will advance safety and soundness by increasing the risk sensitivity of the specialised lending framework. The PRA considers that the revised definition addresses concerns raised by respondents, in particular by improving consistency with the way HVCRE is defined in other jurisdictions.

HVCRE risk weights

3.164 Two respondents noted an error in the PRA’s proposed risk weights that would apply to certain HVCRE exposures in the good category. The PRA stated in its draft rule that those exposures would be assigned a risk weight of 90% under the slotting approach. The PRA has corrected the near-final rules to require firms to apply a risk weight of 95%. This is in line with the PRA’s intention, as set out in CP16/22, to align the treatment of HVCRE with the Basel 3.1 standards.

Introduction of additional risk sensitivity in the slotting approach

3.165 The PRA proposed several changes to the slotting approach to increase risk- sensitivity, including:

  • introducing an additional condition to the application of the maturity-based lower risk weights for exposures in the good and strong categories;
  • introducing a lower risk weight for income producing real estate (IPRE) exposures that are substantially stronger than the criteria specified for the ‘strong’ rating grade; and
  • aligning the treatment of EL with the lower risk weight treatments for all specialised lending categories, except for HVCRE.

The application of lower risk weights based on maturity

3.166 Under the CRR framework, firms are permitted to apply a lower risk weight in the slotting approach to exposures in the strong and good categories if the remaining maturity of the exposures is less than 2.5 years. The PRA proposed to restrict the exposures that benefit from this approach to those where the firm reasonably considers that the obligor would be able to refinance the exposure in a severe but plausible stress in the refinancing market. This would have prevented exposures that would likely have a longer maturity in practice from being assigned a lower risk weight where that would be inappropriate. This additional restriction is not part of the Basel 3.1 standards.

3.167 One respondent supported this proposal, while 11 respondents opposed it. Those opposed argued that the additional restriction adds unnecessary complexity, subjectivity and would be operationally difficult to apply. Six respondents additionally argued that the risk that the PRA sought to address with the additional restriction is already covered by the slotting approach criteria. One respondent argued that refinancing risk should be captured in Pillar 2B rather than Pillar 1.

3.168 Having considered these responses, the PRA considers that the risk it was seeking to address would be adequately mitigated if firms correctly reflected the refinancing risk arising from an obligor’s contractual obligations at the point of maturity in their assessment of the exposure against the slotting factors. The PRA has therefore decided to not proceed with its proposal and to instead introduce an expectation in its near-final SS4/24 that firms should, when assessing exposures against the ‘stress analysis’ slotting subfactor, assess the obligor’s ability to meet its obligations without the firm refinancing the exposure to extend it beyond its current maturity or the firm otherwise providing any forbearance.

Substantially stronger IPRE exposures

3.169 The PRA proposed that IPRE exposures in the strong slotting category could receive a lower risk weight where they are substantially stronger than the criteria for the strong category. The PRA proposed criteria for this treatment which it considered were aligned with the Basel 3.1 standards and would lead to a more risk-sensitive slotting approach. The PRA proposed to introduce an expectation into its draft SS – Credit risk internal ratings based approach, that ‘substantially stronger’ exposures would broadly correspond to an EL-based credit rating of BBB+.

3.170 While respondents generally supported the PRA’s proposals, they requested that the PRA revise the proposed criteria. Respondents argued that the proposed criteria are too conservative and would result in very few exposures receiving the lower risk weight. Four respondents argued that the PRA’s proposed criterion relating to the leverage of the obligor was circular and would not achieve the PRA’s policy intent of capturing the highest-quality exposures. Two respondents argued that the PRA’s proposal relating to the source of cashflows was unclear and would be difficult to apply. One respondent argued that the three proposed criteria overlap and that exposures should not have to meet all three to be eligible for the lower risk weight.

3.171 Having considered these responses, the PRA notes that its proposals could be read as introducing a stricter requirement than the PRA intended. In order to potentially qualify as substantially stronger, the PRA’s intention was to propose that exposures would need to be assigned to the ‘strong’ category in all of the ‘financial strength’, ‘asset characteristics’, ‘strength of sponsor/developer’ and ‘security package’ slotting factors, rather than every slotting sub-factor. The PRA considers that, with this clarification, the PRA’s proposal was less stringent than was clear to respondents.

3.172 Notwithstanding this, and having considered the responses received, the PRA has decided to amend its proposed criteria to make them clearer, easier to apply by firms, more aligned with the Basel 3.1 standards, and more appropriately calibrated to the risk of IPRE exposures. The PRA has made the following adjustments to the proposed criteria in the near-final rule:

  • replaced the proposed criterion that the exposure receives a ‘Strong’ grade in all slotting factors with a requirement that the risk characteristics of the exposure and the firm’s underwriting standards are ‘substantially stronger’ than specified in the slotting criteria. The PRA considers that this criterion is operationally easier for firms to apply and aligns the scope of exposures to which the lower risk weights apply with the Basel 3.1 standards.
  • revise the proposed criterion that the leverage of the obligor is substantially below the market norm. Under the near-final rules, the leverage of the obligor is required to be ‘very low’ in order to receive the lower risk weight. The PRA considers that this addresses respondents’ comments that the PRA’s proposed criterion was circular and would not achieve the PRA’s policy intent.
  • revise the proposed criterion regarding the quality of the income stream of the exposure. Under the revised near-final rules, the income stream is required to be consistent with one that a firm would expect for an investment-grade exposure, and 100% of the obligor’s debt service obligations must be covered by tenant income, in order to receive the lower risk weight. The PRA considers that this addresses respondents’ comments that the PRA’s proposed criterion was unclear and difficult to apply.

3.173 The PRA has decided to adjust its proposed expectation that exposures which receive the lower risk weight broadly correspond to an external credit rating of BBB+ to link it to the first criterion above. Under its near-final rules, the PRA expects that, when firms consider whether the exposure’s risk characteristics and the firm’s underwriting standards are substantially stronger than specified in the slotting criteria, this broadly corresponds to an external credit rating of BBB+.

3.174 The PRA considers that these revisions better align the lower risk weights with the risk of exposures, are easier for firms to implement, and align with the scope of exposures to which the lower risk weights apply in the Basel 3.1 standards.

Application of lower risk weights to commercial ADC exposures

3.175 The PRA has reassessed its overall policy on specialised lending in light of the adjustments above. The PRA notes that the Basel 3.1 standards do not permit the lower risk weights to apply to commercial acquisition, development, and construction (ADC) exposures that are classified as IPRE rather than HVCRE. The PRA has decided to implement this restriction in its near-final rules. Following the changes set out above to the conditions for lower risk weights on the basis of maturity or being ‘substantially stronger’, the PRA considers that, given the risks involved in commercial ADC exposures, it would not be prudent to apply lower risk weights to these exposures where they are in the IPRE category. This will also support competition with firms applying the SA, where such exposures would be subject to a 150% risk weight.

Project finance

3.176 The PRA did not propose any changes specific to the risk weights for project finance exposures on the slotting approach. However, these exposures would be impacted by the changes to the slotting approach proposed more generally and the proposal to remove the infrastructure support factor.

3.177 In response to the proposal to remove the infrastructure support factor, as discussed in ‘Calculation of RWAs and EL’, one respondent requested that the PRA introduce more risk sensitivity in the slotting approach for project finance by introducing a ‘super strong category’. The respondent argued that risk weights would otherwise be too high for high quality project finance exposures compared to both SA risk weights and the default risk of this type of exposure.

3.178 Having considered this response, the PRA has decided to introduce a new ‘substantially stronger’ category for project finance exposures on the slotting approach. This category would include exposures where the risk characteristics of the exposure and the firm’s underwriting standards are substantially stronger than specified in the slotting criteria. The PRA has also decided to introduce an expectation in its near-final SS4/24 that this corresponds to exposures that are broadly equivalent to an EL-based credit rating of BBB+ or better.

3.179 The majority of project finance exposures subject to the slotting approach are currently in the strong category, leading to limited risk differentiation for these exposures. The PRA considers that this change will therefore improve risk differentiation of project finance exposures, and that it will be in line with the underlying risk of the exposures. The PRA considers that this change will partially offset some of the impact of removing the infrastructure support factor for exposures on the slotting approach. The PRA considers that this approach aligns with the Basel 3.1 standards that explicitly permit a ‘substantially stronger’ treatment for these exposures.

Object finance

3.180 In CP16/22, the PRA proposed to introduce a 25% LGD floor for senior unsecured exposures to corporates under the AIRB approach. The PRA proposed to introduce an LGD of 40% for senior unsecured exposures to non-financial corporates under the FIRB approach. Firms using the FIRB approach would be permitted to apply a 25% LGD to the secured portion of an exposure following the application of haircuts as appropriate. This is in line with the Basel 3.1 standards. Five respondents argued that this would result in LGDs that are inappropriately high for some object finance exposures. Those respondents cited public data on observed loss rates for maritime and aviation finance exposures to argue that lower LGDs should be permitted.

3.181 Having considered these responses, the PRA has decided to maintain its proposal. The PRA considers that the evidence provided is not sufficient to address the PRA’s concerns as it does not sufficiently reflect a downturn scenario. The idiosyncratic nature of object finance collateral and its generally poor resilience to stress, which respondents recognised, mean that it is appropriate that firms apply LGDs that are appropriately calibrated to a downturn scenario.

Other issues relating to specialised lending

3.182 Other substantive responses focused on the following issues:

  • Use of the slotting approach for IPRE and HVCRE exposures: one respondent argued that the PRA’s proposal to require firms to apply the slotting approach to IPRE and HVCRE exposures puts firms in the UK at a disadvantage to firms in jurisdictions that permit the use of modelled IRB approaches. The PRA has decided to maintain its proposal to no longer permit firms to apply the AIRB and FIRB approaches to IPRE and HVCRE exposures. The PRA continues to consider that this decision will contribute to improving the robustness of RWAs, given the persistent modelling challenges observed for these exposures. The PRA considers that this decision will not result in significant RWA changes, given that the PRA already expects that these exposures are risk-weighted under the slotting approach.
  • Additional risk sensitivity in the slotting approach: one respondent requested that the PRA increase the risk sensitivity of the slotting approach by increasing the number of slotting categories. They requested that the PRA includes LTV as a driver of lower risk weights. They alternatively requested that the PRA redesigns the slotting approach such that firms use it to estimate PDs of specialised lending exposures rather than risk weights. The PRA has decided not to implement these suggestions. The PRA considers that the slotting approach is less risk-sensitive than the modelled IRB approaches by design, and that the near-final rules on the slotting approach are consistent with international standards.
  • Definition of specialised lending against specific transaction structures: one respondent requested that the PRA provide further guidance on how to allocate exposures with specific transaction structures between the general corporate and specialised lending exposure sub-classes. The PRA considers that the allocation of exposures will have to be done on a case-by-case basis and it has decided to not provide further guidance on specific transaction structures.

PRA objectives analysis

3.183 The PRA considers that its near-final policy relating to specialised lending advances its objectives. The PRA notes the following updates to the analysis it presented in CP16/22 to reflect its decisions.

3.184 The PRA considers that introducing a substantially stronger category for project finance exposures, together with the changes made to the application of the substantially stronger category for IPRE and the maturity-based lower risk weights, will increase risk sensitivity in the slotting approach. The PRA considers the changes will advance its primary objective by better aligning risk weights with the risk of exposures and ensuring an appropriate degree of consistency between firms’ approaches.

3.185 The PRA considers that its decisions advance its secondary objective. It considers that the increased risk sensitivity for project finance lending will benefit firms with exposures on the slotting approach relative to firms modelling using the FIRB and AIRB approaches, and relative to firms applying the SA.

‘Have regards’ analysis

3.186 In developing these near-final rules, the PRA has had regard to its framework of regulatory principles and the matters to which it is required to have regard when proposing changes to CRR. Where the proposed new rules are CRR rules, the PRA has also taken into consideration the matters to which it is required to have regard to when proposing changes to CRR rules. The PRA considers its analysis of its ‘have regards’, as presented in CP16/22, remains appropriate, subject to the following changes:

1. Relative standing of the UK as a place for internationally active firms to operate and competitiveness:

  • The PRA considers that its decisions will advance the competitiveness of UK firms. The PRA considers that the introduction of a ‘substantially stronger’ category for project finance exposures, and the revisions to its proposed scope of the ‘substantially stronger treatment’ for IPRE exposures, will make it easier for UK firms to compete for those types of business.

2. Relevant international standards:

  • The PRA considers that the changes to the definition of HVCRE presented in this PS, and the approach to applying lower risk weights for certain specialised lending exposures, are better aligned with the Basel 3.1 standards than the proposed definition in CP16/22.

3. Climate change:

  • The PRA considers that climate change considerations are nuanced and complex for the near-final rules on specialised lending. The near-final rules should apply a more risk-sensitive approach to the financing of environmental infrastructure projects where these are low risk and eligible for the ‘substantially stronger’ slotting category. Respondents raised concerns that the removal of the infrastructure support factor could hinder investment in environmental projects. The PRA acknowledges that the potentially wider applicability of the infrastructure support factor compared to the project finance substantially stronger treatment may impact lending to ‘green’ projects. However, exposures currently benefitting from the infrastructure support factor are small in aggregate and the criteria do not require that the project in question contributes to environmental objectives.

4: Credit risk mitigation

Introduction

4.1 This chapter provides feedback to responses to chapter 5 of consultation paper (CP) 16/22 – Implementation of the Basel 3.1 standards, which set out proposals to implement the Basel 3.1 standards for credit risk mitigation (CRM), and to amend the Prudential Regulatory Authority’s (PRA) expectations in respect of CRM. This chapter also sets out the PRA’s near-final policy on CRM following the consultation.

4.2 Throughout this chapter, the PRA refers to the following CRM methods outlined in Table 6 and Table 7 below:

Table 6: CRM methods referenced in this chapter applying to Funded Credit Protection (FCP)

CRM method

Description

On-balance sheet netting

A method for recognising on-balance sheet netting under all approaches to credit risk, which the PRA is restricting to recognition through exposure value only.

Financial collateral simple method (FCSM)

A method for recognising financial collateral by reducing risk-weights, which can only be used by firms applying the standardised approach (SA).

Financial collateral comprehensive method (FCCM)

A method for recognising financial collateral by reducing exposure values, which will only be available for (a) exposures that give rise to counterparty credit risk (other than derivatives) under all credit risk approaches, (b) exposures that do not give rise to counterparty credit risk under the SA and, as set out in this near-final PS, (c) exposures subject to the slotting approach. Firms are currently able to model the volatility adjustments used within this method if they have permission from the PRA; however, the PRA is withdrawing this option.

Foundation collateral method

A new method for recognising financial and non-financial collateral, which the PRA is introducing for firms using the foundation internal ratings based (FIRB) approach and which will replace existing similar methods.

Other funded credit protection (OFCP) method

A bespoke method for recognising other funded credit protection (pledged cash and cash-assimilated instruments, pledged life assurance policies, and instruments issued by third-party institutions that will be repurchased on request) under the SA and the FIRB approach which the PRA is retaining.

Loss given default (LGD) modelling collateral method

A method for firms using the advanced internal ratings based (AIRB) approach to recognise the effects of financial and non-financial collateral in LGD estimates.

Securities financing transactions value at risk (SFT VaR) method (previously known as the ‘internal models approach for master netting agreements’)

A method for calculating the exposure value of SFTsfootnote [49] which firms may apply subject to PRA permission. The method currently applies to exposures covered by Master Netting Agreements (MNAs) only; however, the PRA is extending it to also cover single transactions.

Internal models method (IMM)

A method for modelling exposure value for derivatives and SFTs in accordance with counterparty credit risk requirements.footnote [50]

Table 7: CRM methods referenced in this chapter applying to Unfunded Credit Protection (UFCP)

CRM method

Description

Risk weight substitution method

A method that involves substituting the risk weight of the exposure with that of the protection provider to reflect the effect of UFCP. It will apply to all exposures subject to the SA, and to exposures subject to the FIRB and AIRB approaches where comparable direct exposures to the protection providerfootnote [51] would be subject to the SA. The PRA is extending this method to exposures subject to the slotting approach in certain circumstances.

Parameter substitution method

A method that involves substituting probabilities of default (PDs) and, optionally, FIRB LGD values, of the exposure with those of the protection provider to reflect the effect of UFCP. This method is applied by firms using the FIRB and AIRB approaches where they are not applying the risk weight substitution method or, for AIRB approaches, the LGD adjustment method.

LGD adjustment method

A method that involves firms making adjustments to modelled LGD values to reflect the effect of UFCP. The PRA is restricting this approach to exposures subject for the AIRB approach where comparable direct exposures to the protection provider are also subject to the AIRB approach.

4.3 In CP16/22, the PRA proposed amendments which it considered would reduce complexity, improve coherence, and provide greater clarity to firms regarding the availability of CRM methods. Key proposals included:

  • under the SA, removing certain methods for calculating the effects of FCP and amendments to the methods that would remain available;
  • amendments to existing methods for calculating the effects of FCP under the FIRB approach, including new supervisory LGD values and collateral volatility adjustments;
  • restrictions on existing UFCP methods where firms adjust probabilities of default (PDs) and/or obligor grades in internal ratings based approach (IRB) models; and
  • new restrictions on recognising and modelling UFCP which would depend on the credit risk approach applicable to comparable direct exposures to the protection provider.

4.4 The PRA received 31 responses to its proposals on CRM. Comments focused on:

  • collateral eligibility for repurchase transactions, securities borrowing and lending transactions, and margin lending transactions in the trading book;
  • eligibility of collateral where there is material positive correlation between the collateral value and the exposure value for limited recourse securities;
  • the risk weight treatment and eligibility criteria applied under the risk weight substitution method;
  • CRM for exposures subject to the slotting approach; and
  • the risk weight treatment applied to credit insurance as a form of UFCP.

4.5 Having considered the responses, the PRA has decided to amend the draft rules and draft amendments to its supervisory statement in certain areas. This chapter describes the comments and amendments that the PRA considers are more material. As described in Chapter 1 – Overview, the PRA has also made a number of less material amendments and clarifications to the draft rules, which are not described in this chapter. These amendments are reflected in the near-final PRA Rulebook: CRR Firms: (CRR) Instrument [2024] in Appendix 2. Please refer to the document titled Comparison of Draft PRA Rulebook (CRR) Instrument [2023] against Near-final PRA Rulebook: CRR Firms (CRR) Instrument [2024], which contains a comparison of the near-final rules with the draft rules as set out in CP16/22 for ease of identifying all of the changes made.

4.6 The appendices to this near-final policy statement (PS) contain the PRA’s near-final policy, which will:

  • introduce a new Credit Risk Mitigation (CRR) Part of the PRA Rulebook (Appendix 2) to replace the CRR articles HM Treasury (HMT) intends to revoke;
  • insert an additional provision into the Counterparty Credit Risk (CRR) Part of the PRA Rulebook;
  • amend SS17/13 - Credit risk mitigation (Appendix 10); and
  • amend SS12/13 – Counterparty credit risk (Appendix 11).

4.7 The sections below have been structured broadly along the same lines as chapter 5 of CP16/22, covering the main areas where the PRA received comments from respondents as follows:

  • Methods for recognising CRM;
  • FCP; and
  • UFCP.

Methods for recognising credit risk mitigation

4.8 The PRA proposed to implement the Basel 3.1 standards while introducing additional clarifications on the application of the CRM framework, in order to reduce excessive variability of risk weighted assets (RWAs) and reduce complexity. The PRA received 11 responses to this section of CP16/22. The substantive issues raised by respondents are set out below.

FCP securing UFCP obligations

4.9 The PRA proposed to require that where a firm recognises both FCP and UFCP covering the same exposure, it should do so in an appropriate manner and in a way that does not double count the effects of the credit protection.

4.10 Two respondents argued that the PRA’s proposed approach would lead to an inappropriately conservative treatment in the case where a firm receives FCP that secures the obligations of an UFCP provider made in respect of an exposure, for example for collateralised guarantees. These respondents argued that firms should be permitted to recognise FCP where the UFCP has been provided by a protection provider that is not included in the list of eligible protection providers set out in the Credit Risk Mitigation (CRR) Part Article 201, but where the other UFCP eligibility criteria are met.

4.11 Having considered the responses, the PRA has decided to amend its draft rules to include a decision tree and related provisions to specify the treatment of FCP securing UFCP obligations. The PRA has also added further guidance on the treatment of FCP securing UFCP obligations in supervisory statements (SS)17/13 – Credit risk mitigation and SS4/24 – Credit risk internal ratings based approach (as set out in Chapter 3 – Credit risk – internal ratings based approach).

4.12 The PRA recognises that the credit quality of the UFCP provider does not impact the risk-mitigating effect of the FCP, and it has therefore decided to permit recognition of eligible FCP where the UFCP has been provided by a CRM-ineligible protection provider as long as all other UFCP eligibility criteria have been met. The PRA has decided to cap the amount of FCP that may be recognised at the value of the UFCP and firms will not be able to take the UFCP itself into account in this scenario (because the protection provider is not eligible).

4.13 The PRA has also decided that if both the UFCP and the FCP meet all relevant eligibility criteria, then firms will be permitted to take both into account. The amount of FCP that may be recognised will again be capped at the value of the UFCP.

4.14 The PRA has decided that firms should be permitted to recognise FCP securing UFCP obligations in line with the decision tree below. This is a simplified version of a decision tree included in the near-final rules which sets out in more detail how the calculations will work in each case.

Chart 2: Summary of the framework for recognition of FCP securing UFCP obligations

Chart 2 is a decision tree which summarises how firms can recognise funded credit protection securing unfunded credit protection obligations. This decision tree is a simplified version of that which is presented within the near-final rules.

FCP recognition for exposures subject to the slotting approach

4.15 The PRA proposed that, for exposures subject to the slotting approach, collateral would not be recognised via the CRM framework but would instead continue to be reflected in the assignment of exposures to slotting categories.

4.16 One respondent opposed this proposal because it would mean that a fully cash-collateralised exposure could not receive a risk weight lower than 50%. This respondent argued that this risk weight treatment would be overly punitive for cash-collateralised exposures and suggested that either FCP should be recognised via the CRM framework where exposures are subject to the slotting approach, or that the PRA should explain how cash collateral should be reflected within the slotting approach itself.

4.17 Having considered the response, the PRA has decided to amend its draft rules to permit firms to apply the FCCM to exposures on the slotting approach, subject to restrictions to prevent collateral being double counted. Where an exposure is collateralised, the PRA agrees that a minimum 50% risk weight with no adjustment to the exposure value would be disproportionately conservative. The PRA considers it justifiable from a risk perspective that firms should be able to recognise any FCCM eligible collateral in their calculation of an exposure value, and then apply the standard slotting risk weights to the reduced exposure value. In such cases, the collateral will not be recognised in the slotting assessment, to avoid double counting.

Other issues regarding methods for recognising CRM

4.18 Other substantive responses focused on a range of technical issues:

  • Application of CRM to modelled approaches: three respondents suggested clarifying the instances that the CRM rules apply to firms using the IMM. The PRA has amended its draft rules to clarify that certain provisions in the Credit Risk Mitigation (CRR) Part of the PRA Rulebook apply to firms using the IMM in accordance with the relevant cross-references.
  • CRM decision trees: two respondents supported the use of decision trees in the draft rules that set out available CRM methods as they considered that these would provide greater clarity on the CRM approach and methodology to be used; however, they requested the PRA includes rule references within each box of the decision tree. The PRA has decided not to add rule references into the decision trees because the PRA considers that the near-final rules already provide sufficient clarity about the scope of each article.
  • Recognition of collateral under different CRM methods: one respondent requested that the PRA clarify what approach should be used for receivables, other physical collateral and collateralised lease exposures when not applying the foundation collateral method. The PRA clarifies that the only alternative method for recognising these types of collateral is the LGD modelling collateral method, which is only available where a firm is permitted to use the AIRB approach.
  • Use of automated valuation processes: four respondents requested that the PRA clarify that automated valuation models would be permitted under the foundation collateral method. The PRA has clarified in its near-final rules that valuations can be provided by a suitably robust statistical model, which could include an automated valuation model.

PRA objectives analysis

4.19 The PRA considers its analysis of its objectives, as presented in CP16/22, remains appropriate, subject to the updates below.

4.20 The PRA considers that it is prudentially appropriate for firms to recognise FCP securing a UFCP obligation regardless of the guarantor, provided all other UFCP and FCP eligibility criteria are met, for the reasons set out above. The PRA considers that the revised approach broadly maintains the current level of prudence of the capital requirements framework and is consistent with its primary objective to promote the safety and soundness of the firms it regulates.

4.21 The PRA considers that its decision to permit firms to apply the FCCM to exposures on the slotting approach will improve the risk sensitivity of the framework while containing appropriate safeguards to prevent collateral from being double counted, and therefore is consistent with its primary objective.

‘Have regards’ analysis

4.22 In developing these near-final rules, the PRA has had regard to its framework of regulatory principles and the matters to which it is required to have regard when proposing changes to CRR rules. The PRA considers its analysis of its ‘have regards’, as presented in CP16/22, remains appropriate, subject to the following updates:

1. Relative standing of the UK as a place for internationally active firms to operate and competitiveness:

  • While there remains uncertainty about the final approach that will be taken by other regulators, the PRA has not identified any adverse impact on the competitiveness of the UK arising from the changes to the PRA’s proposals as set out in this section. The PRA considers that the amendments to its draft rules may potentially benefit UK competitiveness to the extent that other regulators take a less permissive approach.

2. Relevant international standards:

  • The PRA considers that its near-final policy remains broadly aligned with international standards. As set out in CP16/22, there is ambiguity regarding the availability of CRM methods in the Basel 3.1 standards, and the PRA has sought to achieve an implementation of the standards that will reduce unwarranted variation in risk weights and minimise uncertainty for firms.

3. Efficient and economic use of PRA resources:

  • The PRA considers that its updated approach for FCP securing UFCP obligations clarifies its original proposals and will support a more consistent treatment across firms. The PRA considers the additional clarity will contribute towards efficient and economic use of PRA resources.

Funded credit protection

4.23 The PRA proposed to implement the Basel 3.1 standards while introducing additional clarifications relating to the FCP framework, including:

  • a series of changes to the FCCM formula and supervisory volatility adjustments; and
  • changes to the collateral eligibility requirements for SFTs in the trading book.

4.24 The PRA proposed to maintain the CRR treatment in a number of areas, including:

  • collateral eligibility requirements for own-issued securities; and
  • identification of cases with material positive correlation.

4.25 The PRA received 12 responses to this section of CP16/22. The substantive issues raised by respondents are set out below.

The financial collateral comprehensive method

4.26 The PRA proposed to make the volatility adjustments that apply under the FCCM (to reflect possible future volatility in the value of the collateral) more risk-sensitive, and to introduce a revised formula for reflecting the effect of eligible MNAs across multiple SFTs.

4.27 Responses focused on a number of areas:

  • Volatility adjustments under the FCCM: two respondents noted conflicting language in the draft rules regarding volatility adjustments applied to secured lending transactions and capital market-driven transactions, which could be read as implying that these overlap. The PRA has made minor changes in the near-final rules to reflect that there is no overlap. The PRA considers that the clarification removes any ambiguity in the rules.
  • Margin frequency under the FCCM: the PRA proposed to not retain CRR Article 225(2)(c), which provides a formula for scaling up or down volatility adjustments to reflect shorter or longer liquidation periods, because it relates to the own estimates approach which the PRA proposed to withdraw. Two respondents requested that the formula in CRR Article 225(2)(c) is added back into the PRA’s rules. They argued it was relevant to other situations where supervisory haircuts are applied for non-standard liquidation periods. The PRA has amended its draft rules to introduce a formula for when collateral has a non-standard liquidation period.
  • The FCCM formula for MNAs: two respondents provided suggestions to amend the proposed FCCM formula for MNAs. The PRA has made minor drafting changes to its draft rules to: (i) ensure consistent use of the term ‘groups of securities’; (ii) clarify where ineligible groups of securities should be excluded from the terms of the formula; and (iii) correct the calculation of Enet.
  • Recognition of FCCM-eligible collateral: one respondent requested that the collateral eligibility requirements under the standardised approach to counterparty credit risk (SA-CCR) regime be brought in line with the FCCM. This response related to concerns that were outside the scope of the proposals in CP16/22. Consequently, the PRA has decided to make no changes to its proposed policy.

Collateral eligibility for SFTs in the trading book

Extended collateral eligibility for certain trading book exposures

4.28 Firms applying the Counterparty Credit Risk chapter of the CRR to repurchase transactions and securities or commodities lending or borrowing transactions in the trading book are currently permitted to treat all financial instruments and commodities that are eligible to be included in the trading book as eligible collateral, even if the firm does not currently trade them and they are not held in the trading book. This extends the set of eligible collateral beyond that which could be otherwise recognised under the CRM framework.

4.29 The PRA proposed to limit this treatment to financial instruments and commodities that are held in a firm’s trading book. This proposal aligned with pre-existing Basel standards and was intended to prevent the recognition of collateral that firms would not always be able to promptly liquidate in practice because they are not currently trading it.

4.30 Six respondents opposed this proposal. These respondents raised concerns over the operational burden of the proposed requirement and argued that collateral inventories would need to be monitored on a daily basis and that changes in the stock of inventory would cause volatility in capital requirements and large exposures limits.

4.31 Having considered the responses, the PRA has amended its draft rules to remove the requirement that the financial instruments or commodities must be included in the firm’s trading book and to revert to the existing requirement that they must be eligible for inclusion in the trading book. The PRA has however decided to introduce three additional eligibility requirements, that the firm: (i) has assessed the market liquidity of securities received as collateral and is able to demonstrate sufficient depth within the market to exit the position; (ii) has the legal and operational capabilities to trade the financial instruments; and (iii) has the capability to risk manage and value the financial instrument within the trading book. The PRA considers that it is appropriate to add these requirements in order to mitigate the risk that firms recognise additional types of collateral which cannot easily be liquidated and traded in a timely manner.

Application of extended collateral eligibility to margin lending transactions

4.32 The extended collateral eligibility criteria for certain trading book exposures referred to above apply only to ‘repurchase transactions and securities or commodities lending or borrowing transactions’ in the trading book. However, the PRA’s proposals describing this treatment incorrectly referred to ‘securities financing transactions’ (SFTs) in the trading book. SFTs are defined in the CRR as ‘repurchase transactions and securities or commodities lending or borrowing transactions, and margin lending transactions’, and as a result, margin lending transactions in the trading book were inadvertently included in the set of exposures to which extended collateral eligibility criteria would apply in the PRA’s draft rules.

4.33 Four respondents agreed with the inclusion of margin lending transactions in the scope of this treatment, arguing there is no intrinsic difference between the securities received as collateral and the related market liquidity of those securities, or the firm’s ability to trade them upon a default event, in any type of SFT. However, some of these respondents noted that the Basel standards refer to ‘repo style transactions’, which does not include margin lending, and asked for clarification of the PRA’s intention.

4.34 The PRA notes that ‘securities or commodities borrowing or lending transactions’ and margin lending transactions are not mutually exclusive concepts. The PRA considers that margin cash loans that entail transfer of legal title of the securities collateral to the lender can currently fall within the scope of securities borrowing transactions on the part of the margin lender, and can therefore fall within scope of this treatment, provided they also satisfy the provision’s other conditions.

4.35 The PRA considers that a key difference between title-transfer transactions and non-title transfer transactions is the length of time it might take the firm to liquidate the position. Where the firm does not have legal title to the collateral, it must take steps to take legal ownership following a counterparty default. This increases the length of time over which the value of the collateral might change, and therefore the risk that the eventual sale price is not sufficient to cover the loss on the position. This is not the case for transactions where the firm holds legal title from the time the collateral is posted. The PRA also notes that ‘repurchase transactions and securities or commodities lending or borrowing transactions’ will only include title-transfer transactions, whereas margin lending may or may not involve title-transfer transactions.

4.36 Having considered the responses, the PRA has decided to amend its draft rules to clarify that the extended collateral eligibility criteria do not apply to all SFTs, but only to repurchase transactions and securities or commodities lending or borrowing transactions. This means not all margin lending is automatically in scope. Consistent with the analysis above, margin lending is in scope if the transactions involve title transfer and the provision’s other conditions met.

Collateral eligibility requirements

Own-issued securities

4.37 In line with the CRR eligibility criteria, for firms using the FCSM, the FCCM, the foundation collateral method and the SFT VaR method, the PRA proposed that firms would not be permitted to: (i) recognise collateral where there is material positive correlation between the value of the collateral and the credit quality of the obligor; and (ii) recognise own-issued securities as collateral. The PRA proposed to retain an existing carve-out from the prohibition on recognising own-issued securities as collateral for qualifying covered bonds where they are posted as collateral for repurchase transactions and there is no material positive correlation.

4.38 Three respondents argued that there are examples of own-issued securities other than qualifying covered bonds without material positive correlation which may still be an effective risk mitigant. Additionally, respondents argued that the PRA should widen the definition of qualifying covered bonds to include non-UK covered bonds (see Chapter 2 – Credit risk – standardised approach, exposures to institutions and covered bonds and Chapter 3 – Credit risk – internal ratings based approach, LGD estimation). These respondents argued that restricting the scope of FCP-eligible own-issued securities to only qualifying covered bonds is overly prudent and widening the scope to other specific securities would be consistent with the Basel 3.1 standards.

4.39 Having considered the responses, the PRA has decided to maintain its proposals regarding the eligibility of own-issued securities. The PRA considers that the general prohibition on recognising own-issued securities as eligible collateral is appropriate because these are less likely to provide effective risk mitigation relative to other eligible types of collateral. The PRA considers that extending the existing carve-out from this requirement would not align with international standards.

Guidance on identifying cases with material positive correlation

4.40 The PRA proposed to set an expectation that any financial collateral asset whose value has a material positive correlation with the total value of all of the assets to which the lender has legal recourse (including collateral posted by the obligor and any other assets to which the firm has legal recourse) would not be considered eligible collateral on the basis of the material positive correlation prohibition, as described above in own-issued securities.

4.41 Three respondents argued that this expectation was inappropriate for certain derivative transactions where the exposure value is a function of the value of the collateral, for example, sold covered call options. In such cases, if the value of the collateral asset declines, the collateral value would still cover the exposure to the obligor as the exposure value will have reduced in line with the collateral value. As such, these respondents argued that the assessment of material positive correlation should not solely consider the value of the collateral asset and the total value of all assets to which a firm has legal recourse but should also consider the correlation of the collateral to the value of the exposure.

4.42 Having considered the responses, the PRA acknowledges that in the situations described by the respondents it would be appropriate to consider the collateral to be eligible. The PRA has therefore amended its draft amendments to SS17/13 to introduce a transaction-specific carve-out from this prohibition, exempting transactions that are structured so that the exposure value directly depends on the collateral value in such a way that the proportion of the exposure value that is collateralised does not fall if the value of the collateral falls.

Treatment of FCP in RWA calculations

4.43 The PRA proposed to retain an existing requirement to apply a 100% conversion factor (CF) for the purpose of determining the portion of an exposure that is covered by FCP even if a lower CF applies for the purpose of determining the exposure value.footnote [52]

4.44 One respondent made three arguments against the use of a 100% CF for determining the CRM coverage:

  • the requirement is not part of the Basel 3.1 standards;
  • applying a 100% CF would be inconsistent with the purpose of applying CFs, which is to reflect the likelihood of the relevant undrawn exposure being drawn prior to default; and
  • applying a 100% CF would make the effective cost of such protection disproportionately higher for a firm than would be the case for the same protection provided in respect of an on-balance sheet item. This is because the same amount of protection is required to fully cover the exposure in each case despite the actual risk being lower for the off-balance sheet item.

4.45 Having considered the responses, the PRA has decided to retain the proposed 100% CF for the purpose of determining the portion of the exposure that is covered by CRM. The PRA considers that using a lower CF could result in imprudent outcomes.

4.46 For example, if an exposure were subject to a 40% CF, then under the FCCM a firm holding collateral valued at 40% of the exposure value after haircuts would have a zero-capital requirement for the exposure. The realised CF could be higher or lower than 40% in practice due to statistical variation. For the purpose of calculating exposure value, these effects can be expected to average out across a portfolio, but for the purpose of calculating collateral coverage, the effects are asymmetric. This is because if the realised CF is more than 40% then the firm would have a positive exposure at default (potentially leading to a loss), but if the realised CF is less than 40%, excess collateral would be returned to the counterparty (leading to a zero gain). Across a portfolio there would on average be a positive exposure leading to a positive loss which should be capitalised.

4.47 The PRA therefore considers that use of a 100% CF for the purpose of calculating CRM coverage is consistent with its primary objective as it mitigates the risk that insufficient capital is held against off-balance sheet exposures.

Other issues regarding funded credit protection

4.48 Other substantive responses focused on a range of technical issues:

  • Liquidation periods under the FCCM and SFT VaR method: the PRA proposed that a 5-day liquidation period be applied to repurchase transactions in the FCCM and SFT VaR method. Three respondents argued the 5-day liquidation period should also be applied to all SFTs (and not only repurchase transactions), given all SFTs benefit from daily re-margining and prompt liquidation or set-off of collateral upon the event of default. The PRA has decided to maintain its proposal that a 5-day liquidation period shall be applied to repurchase transactions and a 10-day liquidation period shall be applied to other capital market transactions, on the basis that it is prudent to apply a longer liquidation period to non-title transfer transactions, for the reasons set out in paragraph 4.35. The PRA also considers its proposal to be consistent with the Basel 3.1 standards.
  • The look through approach (LTA) for collective investment undertaking (CIU) collateral: four respondents argued that under the PRA’s proposals, most CIU collateral has little value because it is only eligible to the extent that the CIU is required by its mandate to invest in eligible collateral, rather than considering the current amount of eligible collateral the CIU holds as investments. This means the PRA’s proposal to apply the LTA for calculating volatility adjustments for CIU collateral would provide limited benefit. The PRA agrees with respondents that this approach would be overly conservative and has amended the eligibility criteria for CIU collateral in its draft rules so that where a firm is applying the LTA, a portion of a CIU exposure can be considered eligible collateral, reflecting the extent that the CIU currently invests in assets that are eligible collateral. The PRA has also amended the draft rules so that the value of CIU collateral that can be recognised reflects the proportion of CIU shares that are pledged as collateral, in line with the existing credit risk treatment for direct exposures to CIUs.
  • Physical inspection of inventories: in respect of collateral eligibility requirements which would apply under the foundation collateral method,footnote [53] the PRA proposed that other non-financial collateral would only be eligible where the periodic revaluation process includes physical inspection of the collateral. One respondent requested that the PRA clarify what constitutes a physical inspection. The PRA clarifies that this would entail a physical inspection by a suitably qualified valuer.
  • Settlement currency: for firms using the FCCM, the PRA proposed that in the case of over the counter (OTC) derivatives transactions covered by netting agreements recognised by the PRA under CRR Articles 295 to 298, firms would apply a volatility adjustment reflecting currency volatility when there is a mismatch between the collateral currency and the settlement currency. One respondent requested that the PRA clarify the definition of settlement currency. The PRA clarifies that settlement currency refers to the termination currency.
  • Valuation of immovable property: for firms using the foundation collateral method, the PRA proposed to retain the existing treatment where, when determining the value of immovable property collateral, the value is first reduced to account for higher priority claims on the property and then the applicable haircuts are applied. Having considered the treatment of these types of exposures under the SA, as set out in Chapter 2 – Credit risk – standardised approach, the PRA has also reviewed the treatment under the foundation collateral method. The PRA has amended its draft rules so that the value is first reduced by the applicable haircuts and then allocated to the charges in order of priority. The PRA considers this treatment is aligned with the Basel 3.1 standards and ensures the approach will be consistent with that used under the SA.
  • Maturity mismatches in on-balance sheet netting: the PRA proposed to clarify that maturity mismatch requirements apply where firms recognise on-balance sheet netting. The PRA notes that there is a potential ambiguity in its draft rules relating to the calculation of the maturity of the credit protection for on-balance sheet netting which could lead to inconsistent application by firms. The PRA has decided to clarify in its near-final rules that the maturity date of the credit protection should be taken as the earlier of the termination date of the netting agreement and the date on which the off-setting balances may be withdrawn. This means that on-balance sheet netting arrangements applied to instant access deposits will be deemed to have a residual maturity of zero days for the purpose of the maturity mismatch near-final rules.
  • Equities traded on a recognised exchange: for firms using the FCCM, the PRA proposed that equities or convertible bonds traded on a recognised exchange would be considered eligible collateral. One respondent noted an inconsistency compared to the terminology used when setting out volatility adjustments, where in one case the PRA referred to other equities or convertible bonds listed on a recognised exchange. They recommended using the term ‘traded’ rather than ‘listed’ consistently. The PRA’s proposal to refer to traded securities when determining collateral eligibility came from a desire to use standardised language, but the intention was not to change the scope relative to the current framework. The PRA notes that a security may be traded on an exchange but not listed there, so the PRA’s proposal would result in wider collateral eligibility. The PRA considers this may not be prudent, and notes this would be inconsistent with the Basel standards which refer to securities listed on a recognised exchange. The PRA has amended its draft rules to consistently refer to other equities or convertible bonds listed on a recognised exchange.

PRA objectives analysis

4.49 The PRA considers its analysis of its objectives, as presented in CP16/22, remains appropriate, subject to the below updates.

4.50 The PRA considers that its near-final policy relating to collateral eligibility for SFTs in the trading book is consistent with its primary objective, by permitting trading book collateral to be eligible only if firms have the necessary risk management capabilities and a credible process to instigate trading if, and when, required.

4.51 The PRA considers that its near-final policy relating to valuation of immovable property is consistent with its primary objective as it will appropriately capture the risks associated with second charge mortgages. The PRA also considers its approach to facilitate the secondary competition objective as the treatment under the FIRB approach is now aligned with that under the SA.

‘Have regards’ analysis

4.52 In developing these near-final rules, the PRA has had regard to its framework of regulatory principles and the matters to which it is required to have regard when proposing changes to CRR rules. The PRA considers its analysis of its ‘have regards’, as presented in CP16/22, remains appropriate, subject to the following updates:

1. Relevant international standards:

  • While the PRA recognises that its decision to base extended collateral eligibility for certain trading book exposures on a firm’s ability to risk manage and liquidate the collateral differs from the approach in the Basel standards, the PRA considers this approach should deliver broadly similar prudential outcomes to the Basel standards.

2. Proportionality:

  • The PRA considers that the near-final rules on collateral eligibility requirements for SFTs in the trading book provide a proportionate approach to addressing the risk that a firm may have difficulty liquidating collateral which is not currently in the trading book.

Unfunded credit protection

4.53 The PRA proposed to implement the Basel 3.1 standards relating to the UFCP framework while introducing additional clarifications, including on the eligibility criteria for recognising UFCP and the application of UFCP methods. The PRA received 24 responses to this section of CP16/22. The substantive issues raised by respondents are set out below.

Eligibility criteria for recognising unfunded credit protection

The slotting approach – risk weight substitution method

4.54 The PRA proposed to extend the use of the risk weight substitution method to exposures subject to the slotting approach in certain circumstances. The PRA also proposed that the eligibility criteria for UFCP would depend on the CRM method being used, because it considered it appropriate for all firms to be subject to the same eligibility criteria when they employ a given UFCP methodology.

4.55 The effect of these proposals was that firms would only be able to recognise UFCP for exposures subject to the slotting approach where the eligibility criteria applicable under the risk weight substitution method are met, including that corporate protection providers must be externally rated.

4.56 Three respondents supported the PRA’s proposed extension of the risk weight substitution method to exposures subject to the slotting approach. These respondents argued however that the PRA should also permit recognition of UFCP provided by corporates which are internally rated using the IRB approach, as this can be recognised for exposures subject to other IRB approaches because the parameter substitution method is available.

4.57 Having considered the responses, the PRA has decided to maintain its overall proposal that the eligibility criteria for UFCP applied to exposures subject to the slotting approach should depend on the CRM method that is being used. The PRA considers that it is important to maintain this principle to reduce complexity in the CRM framework and improve the consistency and comparability of the application of the CRM framework across firms.

Application of unfunded credit protection methods

Risk weight substitution method: revised formula

4.58 The PRA proposed that under the risk weight substitution method firms would substitute the risk weight of the exposure with that of the protection provider as calculated under the SA, even if direct exposures to the protection provider are risk weighted under the IRB approach.

4.59 Eight respondents argued that the PRA should permit substitution of the risk weight applicable to the protection provider under the IRB approach, where available. Respondents argued that the ability to take account of parental guarantees is an integral part of firms’ client management framework for global banking business. Where UFCP is being used in respect of a parental guarantee, permitting substitution of an IRB risk weight would allow firms to recognise a CRM benefit for parental support where exposures to the parent are subject to the IRB approach and exposures to the subsidiaries are subject to the SA approach.

4.60 Having considered the responses, the PRA has decided to maintain its proposal to require the risk weight of the protection provider that can be substituted to be calculated under the SA. The PRA considers that if firms were permitted to substitute risk weights calculated under the IRB approach, then the reduction in capital requirements due to the risk weight substitution method would often be driven more by the difference in risk weights between the IRB and SA approaches, rather than by the risk reduction offered by the UFCP. The PRA considers this would not be appropriate, as the intention of recognising CRM in capital requirements is to reflect its effect as a risk mitigation tool. The PRA also considers that permitting substitution of the risk weight calculated under the IRB approach would not be consistent with the Basel Committee on Banking Supervision’s intention in developing the Basel 3.1 standards to restrict modelling in the SA regime or with requirements in other jurisdictions.

4.61 The PRA also notes that permitting substitution of an IRB risk weight would mean that a firm that applies the SA to some exposures and the IRB approach to other exposures could apply lower risk weights to its exposures on the SA relative to a firm that applied the SA to all exposures. Permitting substitution of an IRB risk weight would therefore negatively impact the PRA’s secondary objective.

Parameter substitution method: revised formula

4.62 The PRA proposed a revised formula for calculating risk weights under the parameter substitution method. Under this formula, the risk weight that would be applied to the protected portion of the exposure is calculated using the PD and risk weight function applicable to the protection provider in conjunction with either the LGD applicable to the exposure (as if there was no UFCP) or the FIRB approach LGD applicable to the protection provider. One respondent argued that the risk weight formula applicable to the underlying obligor rather the protection provider should instead be used to calculate RWAs. Having considered the responses, the PRA has decided to retain its proposal because it considers that the risk weight formula for the protection provider will produce risk weights that better reflect the risk of the protection. The PRA also notes that this is aligned with the Basel 3.1 standards.

Requirements for applying UFCP

4.63 For firms using the LGD adjustment method, the risk weight substitution method, or the parameter substitution method, the PRA proposed to introduce an explicit requirement that UFCP would only be eligible if it does not contain any clause which would allow the protection provider to change the credit protection unilaterally to the detriment of the lender. Four respondents argued that there are already a number of eligibility criteria, and it was not clear what further value this clause would add. They noted that it would be operationally burdensome to check all existing arrangements for compliance. Respondents requested that if this additional criterion was introduced, the PRA should also introduce a carve out such that evidence of compliance is not required for existing UFCP arrangements, in order to reduce the operational burden.

4.64 Having considered the responses, the PRA has decided to maintain its proposals in CP16/22 for new UFCP arrangements, on the basis that it is prudent to prevent recognition of protection which is potentially ineffective if the provider changes the terms of protection. In view of the responses received on operational burden, the PRA has however decided to introduce a transitional provision in the Credit Risk: General Provisions (CRR) Part of its near-final rules so that the new criterion will not apply to existing UFCP arrangements (those agreed prior to 1 January 2026) until 1 July 2028.

Calculating RWAs under the SA

4.65 The PRA proposed a new formula to clarify how expected loss (EL) is calculated under the risk weight substitution method. The purpose of this formula was to ensure there is no ‘EL minus provision’ adjustment for the protected part of the exposure. The formula would achieve this by setting the EL amount for the protected part equal to a pro-rated share of the specific provisions associated with the exposure.

4.66 Four respondents questioned whether this new formula is necessary, and asked how EL should be reported for the protected part of the exposure, given that the protected part of the exposure is subject to an SA risk weight, and there is no concept of EL in SA.

4.67 Having considered the responses, the PRA has decided to amend its draft rules so that EL is not calculated for the protected part of the exposure, and the provisions relating to the protected part of the exposure are not included in the ‘EL minus provision’ adjustment.

Credit insurance as UFCP

4.68 The PRA proposed to remove the AIRB approach for risk weighting exposures to financial corporates and to only permit use of the LGD modelling collateral method where both the exposure subject to UFCP and direct exposures to the protection provider are subject to the AIRB approach. The PRA proposed this because it shares the Basel Committee on Banking Supervision’s concerns regarding the robustness of LGD and exposure at default (EAD) models for exposures to financial corporates. Therefore, under the PRA’s proposals, firms using the IRB approach would be required to use prescribed FIRB LGDs under the parameter substitution method to determine the risk weight where the protection provider is on the FIRB approach. Under the PRA’s proposals, the prescribed FIRB LGD for senior unsecured claims on an insurer would be 45%. Therefore, under the parameter substitution method this would be the lowest LGD that could be substituted in the case of an exposure with credit insurance.

4.69 Sixteen respondents commented on the PRA’s proposals for the treatment of credit insurance as UFCP. Respondents argued that the 45% FIRB LGD was excessive for policy claims on insurance firms for reasons including that:

  • insurance policyholders rank senior to all other creditors in the creditor hierarchy;
  • the insurance industry is less positively correlated with the economic cycle than the banking industry;
  • insurance firms are subject to separate prudential regulation; and
  • the PRA’s proposals could reduce the benefit of using credit insurance and as such incentivise the use of riskier CRM tools.

4.70 Respondents also argued that it would be aligned with international standards to permit firms to use a lower FIRB LGD for policy claims on insurance firms, because the Basel 3.1 standards state that firms may use an LGD which takes ‘into account’ the seniority of the guarantee, which in this case would be the super-seniority of the policy claim compared to a direct exposure to an insurer.

4.71 Having considered the responses, the PRA has decided to maintain its proposals. The PRA was not persuaded by respondents’ arguments for a lower FIRB LGD given the lack of evidence that losses on exposures protected by credit insurance justify a LGD below 45%. The PRA considers that some of the arguments made by respondents regarding the low-risk nature of credit insurance relate more to PD (which firms can still model under FIRB) than LGD. Indeed, the lack of default data for exposures to insurers is a key reason why the Basel Committee on Banking Supervision has removed the AIRB approach for these exposures as the lack of data points makes it difficult to robustly estimate LGD. Therefore, the PRA considers that allowing a lower FIRB LGD for exposures protected by credit insurance would not be consistent with its primary objective.

4.72 While the PRA acknowledges that insurance policy claims would rank more senior in the creditor hierarchy than a direct exposure to an insurer, the majority of most insurers’ liabilities are policy claims. Therefore, the PRA does not consider that the super-senior status of policyholder claims in the creditor hierarchy would necessarily justify a lower LGD than for senior claims against other financial sector entities (for example, bank guarantees) in the event of a default.

4.73 The PRA does not agree with respondents’ claim that the Basel 3.1 standards permit applying a FIRB LGD for credit insurance policies that is lower than 45%. The Basel 3.1 standards prescribe two potential FIRB LGDs for unsecured exposures to financial corporates: 45% for senior claims and 75% for subordinated claims. The Basel 3.1 standards also state that the 45% LGD may be used if the guarantee represents a senior claim, regardless of whether the exposure is a senior claim. The PRA therefore considers that its near-final rules are consistent with the Basel 3.1 standards.

4.74 Given the use of credit insurance as UFCP is not exclusive to the UK, the PRA considers that should further evidence on the risk mitigation provided by credit insurance emerge which could justify a different approach for credit insurance, it would be preferable for this to be agreed internationally in order to avoid excessive inconsistency across jurisdictions.

Other issues regarding unfunded credit protection

4.75 Other substantive responses focused on a range of technical issues:

  • Credit default swaps (CDS): under the risk weight substitution method and the parameter substitution method, the PRA proposed to retain the CRR treatment and only permit full recognition of CDS as UFCP if restructuring which results in a credit loss is included as a credit event. The Basel 3.1 standards include a national discretion to not apply this requirement where a 100% vote is needed to amend the maturity, principal, coupon, currency or seniority of the protected exposures and the legal domicile in which the corporate exposure is governed has a well-established bankruptcy code that allows for a company to reorganise/restructure and provides for an orderly settlement of creditor claims. One respondent requested that the PRA implement this national discretion, noting that while it is common for CDS on European corporate and financial entities to include restructuring clauses, that was not the case for CDS on North American corporate and financial entities. The respondent argued this was the type of arrangement which was intended to be captured by the national discretion. Having considered the response, the PRA has decided to amend its draft rules to implement this national discretion. The PRA considers, based on the evidence it received, this is consistent with its primary objective and with the requirements implemented in other jurisdictions.
  • Indirect counter guarantees: aligned with the Basel 3.1 standards, the PRA proposed to restrict the eligibility of indirect counter-guarantees, under the risk weight substitution method and the parameter substitution method, to those provided by central governments and central banks. One respondent argued that an exception should be made to this requirement where UFCP is provided by a special purpose entity (SPE), which then benefits from a counter-guarantee from an entity which is an eligible protection provider. The PRA has decided to maintain its proposal to restrict the eligibility of indirect counter-guarantees to those provided by central governments and central banks. The PRA considers that this is appropriate because the complexity of indirect credit arrangements such as that described by the respondent means that they are in general less likely to be effective than direct credit protection arrangements.
  • UFCP provided by central counterparties: the PRA proposed that qualifying central counterparties may be considered eligible providers of UFCP under the risk weight substitution method and the parameter substitution method. One respondent requested that the PRA clarify whether the reference to calculation of RWAs in Credit Risk: General Provisions (CRR) Part Article 107(2) was intended to cover the scenario of a firm calculating the effect of UFCP provided by a central counterparty. The PRA clarifies that the reference in Article 107 does not cover cases where the central counterparty is providing a guarantee and the exposure itself is not to a central counterparty.
  • Treatment of UFCP in RWA calculations: as set out in paragraph 4.45 for FCP, the PRA has decided to maintain the proposed 100% CF for the purpose of determining the portion of the exposure that is covered by CRM as it considers that using a lower CF would result in imprudent outcomes.
  • Presentation of risk weight calculations: as set out in Chapter 8 – Reporting, credit risk standardised approach (SA), respondents raised concerns about the reporting of exposures subject to the risk weight substitution method. The PRA has made minor amendments to the presentation of the risk weight calculations to be performed under the risk weight substitution method and parameter substitution method to clarify that firms should separate each exposure subject to these methods into a covered part and an uncovered part. The PRA considers that this improves the clarity of the rules and will facilitate the reporting of these exposures.

PRA objectives and ‘have regards’ analysis

4.76 The amendments noted in this section, including on how EL is calculated under the risk weight substitution method and on CDS, are not expected to materially affect the outcomes of the UFCP framework. Therefore, the PRA considers the near-final rules are materially aligned with those in CP16/22 and its analysis of its objectives and ‘have regards’ in CP16/22 remain appropriate.

5: Output floor

Introduction

5.1 This chapter provides feedback to responses to chapter 9 of consultation paper (CP) 16/22 – Implementation of the Basel 3.1 standards, which set out proposals to implement the Basel 3.1 standards for the output floor with respect to firms’ calculation of own fund requirements. This chapter also sets out the Prudential Regulation Authority’s (PRA) near-final policy on the output floor following the consultation.

5.2 In CP16/22, the PRA proposed:

  • to introduce a floor on risk-weighted assets (RWAs) that would require firms in scope of the output floor, with internal model (IM) permissions, to calculate RWAs as the higher of: (i) the total RWAs calculated using all approaches that they have supervisory approval to use (including IM approaches); and (ii) 72.5% of RWAs calculated using only standardised approaches (SAs) (where the latter is called ‘the output floor’ or ‘floored RWAs’);
  • to apply the requirement to UK firms that are not part of a group headquartered overseas. For those firms, the output floor would be applied on a consolidated basis at the level of the UK consolidation group where such a group exists, or on an individual basis where the firm is not part of a group. In addition, where a firm is a ring-fenced body (RFB), the output floor would be applied on a consolidated basis at the level of the ring-fenced sub-group, or on an individual basis where the RFB is not part of a ring-fenced sub-group;
  • to require those firms to apply floored RWAs in the calculation of all own funds requirements and buffers when bound by the output floor;
  • to require that those firms apply the PRA’s proposed implementation of the SA in the same manner as firms without permission to use IMs; and
  • to apply transitional arrangements for the output floor, consistent with the Basel 3.1 standards regarding the length of the transitional period, beginning on 1 January 2025.

5.3 The PRA received 49 responses to its proposals on the output floor. Comments generally supported the overall approach. Respondents focused on:footnote [54]

  • the implementation of the output floor, in particular the treatment of accounting provisions;
  • its scope and levels of application, including for international subsidiaries; and
  • the application of the output floor to minimum requirements and buffers.

5.4 Having considered the responses to the consultation, the PRA has decided to amend the draft rules relating to the treatment of accounting provisions and the transitional period. This chapter describes the comments and amendments that the PRA considers are more material. These amendments are reflected in the near-final PRA Rulebook: CRR Firms: (CRR) Instrument [2024] in Appendix 2. The document titled Comparison of Draft PRA Rulebook (CRR) Instrument [2023] against Near-final PRA Rulebook: CRR Firms (CRR) Instrument [2024] contains a comparison of the near-final rules with the draft rules as set out in CP16/22 for ease of identifying all of the changes made.

5.5 The appendices to this near-final policy statement (PS) contain the PRA’s near-final policy, which will:

  • introduce a new Required Level of Own Funds (CRR) Part of the PRA Rulebook to replace CRR articles that HM Treasury (HMT) plans to revoke.

5.6 The sections below have been structured broadly along the same lines as chapter 9 of CP16/22, covering the main areas where the PRA received comments from respondents as follows:

  • Implementation of an output floor;
  • Scope and levels of application;
  • Application of the output floor to minimum requirements and buffers;
  • Application of the output floor to securitisation exposures; and
  • Transitional arrangements.

5.7 Unless specified otherwise, the near-final rules are consistent with those in CP16/22 and therefore the PRA considers its analysis of its objectives and ‘have regards’ in CP16/22, with respect to the areas mentioned in paragraph 5.5 above, remains appropriate.

Implementation of an output floor

5.8 The PRA allows firms to measure risk and calculate associated RWAs in two different ways, depending on the risk concerned: the SAs, in which the PRA defines the risk weights that should be applied to different exposures and risks; and IM approaches, where the PRA allows firms with the requisite permissions to model certain parameters. Aligned with the Basel 3.1 standards, the PRA proposed to implement an output floor. The output floor, a central new element in the Basel 3.1 standards, would ensure that RWAs for firms with IM permissions do not fall below a defined percentage of the RWAs calculated under the SAs.

5.9 A firm within the scope of the proposed output floor would be required to calculate RWAs, for the purposes of compliance with own funds requirements and buffers, as the higher of: (i) the total RWAs calculated using all approaches which it has supervisory approval to use (including IM approaches); and (ii) 72.5% of RWAs calculated using only standardised approaches.

5.10 The proposed output floor was constructed to reduce excessive cyclicality in RWAs, enhance comparability of RWAs among firms, and protect against model risk, promoting the credibility of the risk-weighted regulatory capital framework.

5.11 The PRA received 17 responses to this section of CP16/22. 16 respondents indicated that the proposed floor the PRA consulted on may result in an increase in capital requirements. The PRA considers that the proposed output floor could increase capital requirements where it addresses previously uncaptured model risk.

5.12 A number of respondents also discussed the calculation of the output floor itself, in particular the treatment of accounting provisions, as set out below.

Calculation of the output floor

5.13 The PRA proposed the following equation for the calculation of total RWAs under the output floor:

Total RWA = max [RWAs (all approaches), 0.725 *RWAs (SAs only)]

5.14 The proposed output floor calculation did not include any adjustments for the differing treatments of accounting provisions under the credit risk SA and internal ratings based (IRB) approach. Instead, CP16/22 noted that firms using IM approaches ‘would not be required to make any adjustments to the treatment of accounting provisions with respect to own funds to align with SAs, regardless of whether they are bound by the output floor’.

5.15 For exposures subject to the IRB approach, firms are required to compare the total amount of accounting provisions (including various adjustments) (P) with the total expected loss (EL) amount and adjust capital resources as follows:footnote [55]

  • where the EL amount exceeds P, firms must deduct the difference from Common Equity Tier 1 (CET1) capital;
  • where P is greater than the EL amount, firms may add the difference to Tier 2 (T2) capital up to a cap; and
  • if specific provisions exceed EL amounts for defaulted exposures, a separate calculation is required, resulting in both a CET1 deduction and a T2 add-on.

5.16 For exposures subject to the credit risk SA, firms are not required to calculate regulatory EL or make any equivalent comparison with P. Firms may however add General Provisions (where relevant under the applicable accounting framework) to T2 capital, up to a cap (specific provisions are reflected in credit risk SA RWAs as the exposure value used is net of specific provisions).

5.17 The Basel 3.1 standards for the output floor do not specify an approach for reflecting differences between EL amounts and accounting provisions, or General Provisions. In CP16/22, the PRA proposed to make no adjustment because it considered that, although this arguably resulted in a less comparable basis for the output floor, the cost was outweighed by the benefit of applying a simpler approach, consistent with the output floor acting as a backstop.

5.18 The PRA received eight responses on the proposed treatment of accounting provisions in the calculation of the output floor. Seven, including industry representative bodies, expressed a strong preference for the PRA to place more weight on consistency of the treatment of firms that use the IRB approach and firms that do not under the output floor rather than on simplicity, and therefore argued to include an adjustment for accounting provisions. Four respondents argued that the approach proposed in CP16/22 could also introduce some double counting into the framework.

5.19 The PRA has considered the responses and on balance has decided to introduce an adjustment in its near-final rules, which adjusts for EL and accounting provisions, in order to better approximate the capital ratio that would have applied if all exposures had been subject to the SAs. The PRA considers that the revised equation remains aligned with the Basel 3.1 standards, but takes into account respondents’ preference for a calculation that brings output floor and SA methodologies into closer alignment.

5.20 The output floor calculation will be as follows:

Total RWA = max [RWAs (all approaches), 0.725 *RWAs (SAs only) + Output Floor Adjustment]

Where:

  • Output Floor Adjustment = 12.5 * (IRB T2 – IRB CET1 – GCRA + SA T2);
  • IRB T2 is the P greater than EL that may be added to T2 up to a cap;
  • IRB CET1 is the EL greater than P that may be deducted from CET1;
  • General Credit Risk Adjustments (GCRA) are the General Provisions (GP) which may be added back to T2 up to a cap in the SA framework, calculated assuming all exposures are subject to the SA, as per the expectations of the output floor; and
  • SA T2 is the GP already reflected in T2 due to exposures on the Credit Risk SA.

5.21 The new adjustment converts the effect of the difference between EL and P on capital resources into equivalent RWAs by multiplying by 12.5, so that they can be directly taken into account in the calculation of the output floor. This is to ensure the approach is as straightforward to implement as possible. Both IRB-related adjustments (the IRB T2 - IRB CET1 terms) and SA-related adjustments (the GCRA + SA T2 terms) are included to provide the most comparable basis for the calculation.

PRA objectives analysis

5.22 With the exception of the above adjustment for accounting provisions, the PRA considers that the near-final output floor rules remain aligned with the proposals in CP16/22 and, therefore, the PRA considers its analysis of its objectives and have regards for those areas in CP16/22 remains appropriate.

5.23 With respect to the PRA’s decision to introduce an adjustment for accounting provisions, the PRA considers that the approach continues to advance the PRA’s primary objective by prescribing a transparent and comparable approach to calculating the output floor. As the adjustment is applied via a consistent methodology for all firms that use the IRB approach, and brings floor-bound and SA methodologies closer in alignment, the PRA considers the adjustment to be consistent with its secondary competition objective.

‘Have regards’ analysis

5.24 In developing these near-final rules, the PRA has had regard to its framework of regulatory principles and the matters to which it is required to have regard when proposing changes to CRR rules. The PRA considers its analysis of its ‘have regards’, as presented in CP16/22, remains appropriate, subject to the following updates:

1. Relevant international standards:

  • The PRA considers that the adjustment for accounting provisions is aligned with international standards.

2. Proportionality:

  • The PRA considers that the additional complexity of the new adjustment for accounting provisions is justified in light of consultation feedback supporting the feasibility of a more complex calculation.

Scope and levels of application

5.25 The PRA proposed the output floor would apply to firms in scope of the PRA’s CRR requirements in the following way:

  • on a consolidated basis only, at the UK consolidation level (the ultimate UK group level) of UK-headquartered groups;
  • on an individual basis to UK stand-alone firms; and
  • on a sub-consolidated basis for RFB sub-groups, or individual basis where the RFB is not part of a ring-fenced sub-group.

5.26 The proposed approach would exclude UK-based subsidiaries of banking groups headquartered overseas (‘international subsidiaries’), where they are subject to group consolidation outside the UK.footnote [56]

5.27 The PRA received 37 responses relating to the scope and levels of application of the output floor. Seven respondents expressed support for application at the consolidated level as set out in CP16/22. The most substantive comments related to the treatment of international subsidiaries, and the impact of applying the output floor to mutuals and other firms specialising in retail lending as described below.

5.28 In CP16/22, the PRA recognised that it may, in the future, consider there is a prudential case to review the scope of the output floor. The PRA recognises that clear expectations about capital requirements benefit safety and soundness. Any future change would need to be supported by evidence and be consistent with the PRA’s objectives and ‘have regards’. Any proposed change to the policy would be subject to further consultation. However, the PRA is aware that jurisdictions may implement final rules at different times. Therefore, to support the consistent application of requirements, the PRA has introduced a permission it may grant for firms to be treated as international subsidiaries, where the PRA is satisfied that there is a public commitment by the relevant authorities to implement an output floor (as per Article 92 of the Required Level of Own Funds Part of the PRA Rulebook). The PRA will shortly provide details of how to apply for this permission on its Permissions (CRR firms) webpage.

International subsidiaries

5.29 The PRA proposed that international subsidiaries would not be subject to the output floor requirement. CP16/22 also noted that the PRA may still request firms excluded from the scope, including international subsidiaries, to participate in ad hoc data gathering exercise(s) on the impact of the output floor.

5.30 The PRA received a range of comments from respondents:

  • four respondents supported the approach to international subsidiaries;
  • two respondents argued against the exclusion of international subsidiaries from the scope of the output floor, citing concerns about competition from EU-headquartered groups;
  • two respondents requested transparency around any expected role of equivalence in the scope of the output floor; and
  • two respondents considered the ad hoc data gathering to be potentially burdensome.

5.31 Having considered these responses, the PRA has decided not to amend the proposed treatment of international subsidiaries. The output floor is an aggregate backstop, aiming to address issues of variability, accuracy, and consistency in RWAs at a broad level, rather than activity by activity. As such, the PRA considers that the output floor is applied most effectively at the consolidated level, to allow the recognition of diversification between risks, and minimise potential impact on specific business activities.

5.32 The PRA has considered the concerns of respondents on the impact of the scope of application of the output floor on competition between UK-headquartered firms and international subsidiaries, including those that would be part of EU-headquartered groups. The PRA recognises these concerns; however, it considers that its approach to only exclude international subsidiaries from the scope of the output floor in cases where they are subject to the floor at consolidated level (and not subject to ring-fencing at the sub-consolidated level), should minimise any competition effects. With regard to the role of equivalence in determining the scope of the output floor, the PRA does not currently envisage a direct link to equivalence determinations. International subsidiaries will however continue to be subject to PRA supervision, which recognises the quality of home jurisdiction capital regulation (including its equivalence), as well as associated supervisory powers. Over time, the PRA will monitor the implementation of international standards across home jurisdictions in the context of its supervision of UK-regulated firms.

5.33 With regards to ad hoc data gathering for international subsidiaries, these exercises will be limited to the UK group consolidation level and apply only to firms with permission to use IMs. The PRA will aim to minimise items collected, to limit operational disruption and burden. The PRA considers such data collections necessary to support the legal requirement to carry out impact assessment and evaluation.

Impact on retail lenders

5.34 The PRA proposed that the output floor would apply to mutuals and other firms, including those that specialise in retail lending. In CP16/22, it acknowledged that mutuals with IM permissions may experience a relatively higher impact from the output floor because of their lack of diversification. The size of this effect would depend on the extent to which the IRB approach continues to produce lower average risk weights relative to the Basel 3.1 SA following changes including the introduction of hybrid modelling,footnote [57] the IRB roadmapfootnote [58] and the lowering of some SA risk weights in the Basel 3.1 standards.

5.35 The PRA also proposed to implement the output floor at the level of ring-fenced sub-groups (or individual RFBs if there is no sub-group). This was consistent with the prudential function of the ring-fence (treating the ring-fenced sub-group as equivalent to the UK consolidation level, within the ring-fence). The PRA considered this important in supporting its secondary competition objective, as it would promote a level playing field with the specialised mutuals in the provision of finance to the UK domestic retail market.

5.36 The PRA received ten responses on the impact of the output floor on non-RFBs specialised in retail lending, including mutuals. Respondents expressed concern that the potential impact of the output floor could be disproportionate.

5.37 The PRA received five responses supporting the application of the output floor to RFBs, and four suggesting the impact may be disproportionate. Of the responses that supported application at the RFB level, three specifically cited consistency of competition in the UK domestic retail market.

5.38 Having considered these responses, the PRA has decided to maintain the approach set out in CP16/22, which includes applying the output floor to mutuals and ring-fenced sub-groups. Specialisation in retail lending, and in particular residential mortgages, does not preclude a firm from having model risk, and may amplify it, due to less diversified model use. The PRA considers application of the output floor at the RFB level to be consistent with the prudential function of the ring-fence. The PRA also considers that this level of application supports fair competition between RFBs, mutuals and other retail lenders. The PRA notes that the impact of the output floor on all firms focused on the UK domestic retail market may also be smaller when considered alongside the combined impact of any other applicable elements of the capital framework, including Basel 3.1, hybrid modelling, the IRB roadmap and the leverage ratio.

5.39 The PRA recognises the role that applying the output floor to RFBs plays in supporting fair competition between retail lenders, and the concern raised by respondents regarding the potential impact of any change to the ring-fencing regime. Should a change in the regulatory landscape result in material risk to the PRA’s objectives, the PRA would assess the case for addressing any material concerns through routine supervisory powers or, if deemed necessary, consult on a revised scope of the output floor.

PRA objectives and ‘have regards’ analysis

5.40 The near-final rules are materially consistent with those in CP16/22. In addition, further information including from other jurisdictions made available since the publication of CP16/22 has not materially altered our judgements with regards to competitiveness. Therefore, the PRA considers its analysis of its objectives and have regards in CP16/22 remains appropriate.

Application of the output floor to minimum requirements and buffers

5.41 The PRA proposed that when the output floor is ‘activated’ (when 72.5% of RWAs calculated using SAs exceed RWAs calculated using IM approaches), ‘floored’ RWAs would be used as the applicable RWAs wherever relevant in all elements of the capital stack.

5.42 In particular, the ‘floored’ RWAs would be used as the basis for the calculation of buffers, where relevant, including:

  • the capital conservation buffer (CCoB), set at 2.5% of RWAs, and countercyclical capital buffer (CCyB);
  • higher loss absorbency requirements for systemic firms, ie for firms identified as global systemically important banks (G-SIBs) or other systemically important institutions (O-SIIs); and
  • the PRA buffer (also referred to as Pillar 2B).

5.43 Firms within scope of the leverage ratio framework would also remain subject to leverage ratio requirements.

5.44 The PRA received 12 responses regarding the application of floored RWAs through the capital stack:

  • eight respondents raised concerns about the application of floored RWAs to Pillar 2 and buffers. The substantive comments related to concerns about double-counting, where the same risk is fully captured twice by two different requirements, and the potential for a competitive disadvantage with other jurisdictions. The respondents also requested further clarification on the expected approach where granular ie portfolio or asset level RWAs, rather than total RWAs, are used, such as in the calculation of credit concentration risk, as set out in statement of policy (SoP) – The PRA’s methodologies for setting Pillar 2 capital.
  • four respondents specifically referred to potential interactions of the output floor with the PRA buffer, via stress testing, in particular raising concern that the phase-in of the output floor may be effectively brought forward by stress testing and resultant PRA buffer calculations.

5.45 Having considered these responses, the PRA has decided to retain the proposed approach of applying the output floor to the whole capital stack. The PRA has not identified any double-counting between the output floor and other elements of the capital stack, including through consideration of evidence provided as part of responses to CP16/22. Any refinements related to the Basel 3.1 standards which improve risk capture and could produce double-counting in Pillar 2A, will be addressed in an off-cycle review of firm-specific capital requirements, as set out in Chapter 6 of PS17/23 – Implementation of the Basel 3.1 standards near final part 1.

5.46 The PRA recognises that the introduction and phase-in of the output floor may have operational impacts on the calculation of Pillar 2, as set out in PS17/23. To promote a smooth and consistent transition, the PRA has decided:

  • to rebasefootnote [59] firms’ Day 1 Pillar 2 (both variable parts of Pillar 2A and the PRA buffer) as part of the off-cycle review; and
  • to rebase firms’ variable Pillar 2A requirements during the output floor transitional period, so that any changes to RWAs resulting from the output floor multiplier do not impact Pillar 2A capital requirements where the relevant risk level has not changed.

5.47 In response to questions raised during the consultation, the PRA is able to clarify the expected interaction between the output floor and Pillar 2A. The output floor applies at a firm-wide level (or RFB), and not to individual asset classes, models or risk weights. As such, the PRA does not expect or require firms, when becoming bound by the output floor, to change the granular risk-weights used for the purposes of calculating individual components of Pillar 2A.footnote [60] The PRA considers this position consistent with the PRA’s proposal in CP16/22 regarding the calculation of CCyB pass-through rates.

5.48 The PRA has also considered feedback regarding the interaction between the PRA buffer and stress testing. The PRA buffer is an amount of capital firms should maintain in addition to their total capital requirement (and the combined buffer). The PRA buffer is generally calibrated to absorb losses that may arise under a severe stress scenario, while avoiding duplication with the combined buffers. Its size is informed by a range of factors including the results of relevant stress-testing and supervisory judgement. The PRA recognises the output floor may increase complexities in the setting of the PRA buffer for some UK firms. Where the output floor impacts a firm’s RWAs, it might in turn affect the extent of a capital drawdown and / or RWA increases in a severe stress scenario. The PRA acknowledges that there will be interactions between Basel 3.1, stress testing and the PRA buffer and will provide firms with more guidance as part of the next concurrent exercise involving firm submissions of stressed projections.

5.49 In setting the PRA Buffer, the PRA will be particularly alert to the intended effect of the transitional arrangements for the output floor, and will also consider any other relevant factors, including the extent to which the output floor is expected to bind during stress.

Application of the output floor to securitisation exposures

5.50 The PRA proposed that when applying the output floor, IM firms would apply the PRA’s proposed implementation of the SA in the same manner as for firms without permission to use IM. This was because the PRA considered that a robust and consistent application of SA methodologies by IM firms subject to the output floor is necessary to achieve the prudential objectives of the output floor and facilitate the secondary competition objective.

5.51 Consistent with this, the PRA proposed that securitisation exposures would be included in the output floor calculations, aligned with the Basel 3.1 standards. The PRA also proposed to engage with firms originating significant risk transfer (SRT) securitisations and market participants, to understand the impact of the proposed use of standardised methodologies for securitisations for the purposes of the output floor.

5.52 The PRA received 13 responses regarding the application of the output floor to securitisation exposures. Respondents expressed concern that the output floor could cause securitisation transactions to become uneconomic, because of the higher risk weights under SEC-SA compared to SEC-IRBA, and this could in turn reduce their lending capacity. Respondents asked that the PRA reconsider the design of the application of the output floor to securitisation, potentially by adjusting the SEC-SA framework. Respondents indicated that they would consider the impact of the output floor at the securitisation transaction level when considering capital allocation within their consolidated group.

5.53 The PRA published discussion paper (DP)3/23 – Securitisation: capital requirements to seek views and evidence from market participants on policy options related to this issue. The PRA will consider industry feedback on this issue with the intention of informing the PRA’s approach in a future consultation, subject to HMT making the necessary legislation.

Transitional arrangements

5.54 The PRA proposed to align with the Basel 3.1 standards by including a five-year transitional period, beginning on 1 January 2025, in line with the PRA’s proposed implementation date. The transitional period was intended to support the implementation of the output floor in an orderly manner, and reduce potential cliff edges in own funds requirements. The PRA proposed not to apply the RWA capfootnote [61] during the transitional period.

5.55 The PRA received two responses regarding the transitional period. One respondent supported the PRA pursuing international consistency where feasible and one respondent supported the proposed start date, but argued that the total length of the transitional period was excessive.

5.56 Four respondents commented on the RWA cap. Three respondents supported the PRA’s proposal, while one suggested a cap would be desirable in the context of different international transitional arrangements.

5.57 As set out in Chapter 1 – Overview, the PRA has decided to amend the implementation date for the Basel 3.1 standards to 1 January 2026. With regards to the output floor transitional period, the PRA has decided to retain the proposed end-date of 31 December 2029. Therefore, the transitional period will begin on 1 January 2026, with 55% becoming the initial multiplier. The following transitional multipliers would remain in line with the PRA’s proposal. The PRA considers that this approach supports consistency in industry expectations, and alignment with other jurisdictions, while guarding against cliff edge effects.

5.58 In line with CP16/22 and industry responses, the PRA is not implementing the transitional cap. The PRA considers that the cap would create a cliff-edge when removed, introducing additional volatility and undermining the benefits of the output floor to safety and soundness.

5.59 As a result of the above, for the transitional period the output floor will be calibrated as set out below:

Date

Transitional multiplier

1 January 2026

55%

1 January 2027

60%

1 January 2028

65%

1 January 2029

70%

PRA objectives and ‘have regards’ analysis

5.60 The near-final rules, incorporating the clarifications with regards to the start date of the transitional period made above, are materially consistent with those in CP16/22 and therefore the PRA considers its analysis of its objectives and have regards in CP16/22 remains appropriate.

6: Pillar 2

Introduction

6.1 This chapter provides feedback to responses to chapter 10 of consultation paper (CP)16/22 – Implementation of the Basel 3.1 standards, which described the implications of the proposed changes to the Pillar 1 risk-weighting framework for the Prudential Regulation Authority’s (PRA) Pillar 2 framework.footnote [62] This chapter sits alongside chapter 6 of PS17/23 – Implementation of the Basel 3.1 standards near-final part 1 and focuses on the responses relating to the Pillar 2A credit risk methodology and the interaction between Pillar 2 and the output floor.

Pillar 2 review

6.2 CP16/22 did not contain any policy proposals for Pillar 2, but it set out the topics the PRA is considering so that Pillar 2 requirements are updated as necessary for the implementation of the Basel 3.1 standards. In particular, it outlined the PRA’s principle that it would not double count capital requirements for the same risks in both Pillar 1 and Pillar 2A.

6.3 Taking responses to CP16/22 into account, in PS17/23 the PRA stated its intention to sequence its Pillar 2 review work by first addressing the consequential impacts of Basel 3.1 Pillar 1 changes on Pillar 2 within the existing methodologies in an off-cycle review of firm-specific requirements, and then reviewing the Pillar 2A methodologies after the PRA’s rules to implement the Basel 3.1 standards are finalised.

6.4 In chapter 6 of PS17/23, the PRA indicated it would provide more details regarding the way forward for the refined methodology to Pillar 2Afootnote [63] when this second near-final PS is published. For this purpose, the PRA issued CP9/24 – Streamlining the Pillar 2A capital framework and capital communications process, which proposes to retire the refined methodology and streamline firm-specific capital communications. For the avoidance of doubt, CP9/24, as well as the off-cycle review of firm-specific capital requirements set out in PS17/23 and updated below (which supports the implementation of the Basel 3.1 standards), do not form part of the PRA’s the future review of the Pillar 2A methodologies.

6.5 In chapter 6 of PS17/23, the PRA also said it would publish proposals on capital-related measures under the Strong and Simple framework in Q2 2024, which would cover simplifications to Pillar 2. The PRA published CP7/24 – The Strong and Simple Framework: the simplified capital regime for Small Domestic Deposit Takers (SDDTs). Please refer to that document for further detail.

Update on the off-cycle review of firm-specific capital requirements

6.6 As set out in PS17/23, the PRA will conduct an off-cycle review of firm-specific Pillar 2 capital requirements using the PRA’s existing Pillar 2 methodologies.footnote [64] The review will address double counting and unwarranted increases or decreases in capital arising from changes in risk-weighted assets (RWAs) as a result of the implementation of the Basel 3.1 standards. The PRA also plans to apply firm-specific structural adjustments to Pillar 2A (the ‘SME lending adjustment’ and ‘infrastructure lending adjustment’) to ensure that overall capital requirements for SME and infrastructure lending do not increase as a result of the removal of the Pillar 1 support factors in this off-cycle review.footnote [65] The PRA is conducting a data collection exercise to inform the PRA’s assessments on the adjustments needed. As set out in PS17/23, the PRA does not expect firms to conduct a full Internal Capital Adequacy Assessment Process (ICAAP) for the purposes of this review. The PRA plans to communicate the adjusted Pillar 2 requirements to firms (ie the outcome of this off-cycle review), ahead of the implementation date of the Basel 3.1 standards on 1 January 2026 (‘day 1’), so that firm-specific requirements will be updated at the same time as the Basel 3.1 standards are implemented.

Pillar 2A – Credit risk

6.7 In CP16/22, the PRA stated that while changes to credit risk methodologies would improve risk capture, some deficiencies may remain and a Pillar 2A methodology for credit risk would likely continue to be required. The PRA did not propose any policy changes within CP16/22 and set out its intention to reflect further on its Pillar 2A methodology as part of its Pillar 2A review.

6.8 17 respondents provided feedback on the PRA’s approach to Pillar 2A credit risk. These included questions on the PRA’s future Pillar 2A credit risk methodology, particularly how the use of internal ratings based (IRB) benchmarks will be reviewed and updated, and how the Pillar 2 framework could be reviewed in light of the changes in the Pillar 1 credit risk approach.

6.9 The PRA intends to consider the Pillar 2A credit risk methodology (including the use of IRB benchmarks), including any necessary updates and changes, as part of the future Pillar 2A methodologies review, which will be subject to a separate consultation. In line with the PRA’s intention to adjust specific elements of firms’ Pillar 2 requirements ahead of implementation via the off-cycle review, the PRA will rebasefootnote [66] Pillar 2A capital add-ons for credit risk.footnote [67] The PRA considers that this is the most pragmatic approach given existing Pillar 2A credit risk add-ons are small relative to firms' overall capital requirements, and typically for idiosyncratic risks that are not captured by the Pillar 1 credit risk approach. Credit risk add-ons are also typically less relevant to firms which use the IRB approach because where there are deficiencies in IRB models, the PRA requires the firm to remediate the shortcomings of the Pillar 1 models rather than setting Pillar 2A capital requirements.

Interaction with the output floor

6.10 As stated in CP16/22, the PRA considers that no material methodological overlap will exist between the output floor and the PRA’s Pillar 2 methodologies given model risk is not captured specifically under Pillar 2. Further, where there are deficiencies in Pillar 1 models, the PRA requires firms to remediate the shortcomings rather than setting additional Pillar 2A capital requirements. The PRA stated it would consider how, or if, the output floor should be taken into account in any revised IRB benchmark and Pillar 2A credit risks add-on.

6.11 The PRA has provided more specific feedback to the responses related to the output floor, and its interaction with the Pillar 2 framework in Chapter 5 – Output floor of this PS. In line with the PRA’s plan to adjust specific elements of firms’ Pillar 2 requirements in the off-cycle review as set out above and in PS17/23, the PRA will take into account the impact of the output floor (if any) by:

  • rebasing firms’ day 1 Pillar 2 (both variable parts of Pillar 2A and the PRA buffer) as part of the off-cycle review; and
  • rebasing firms’ variable Pillar 2A requirements during the output floor transitional period, so that any changes to RWAs resulting from the output floor multiplier do not impact Pillar 2A capital requirements where the relevant risk level has not changed.

7: Disclosure (Pillar 3)

Introduction

7.1 This chapter provides feedback to responses to chapter 11 of consultation paper (CP)16/22 – Implementation of the Basel 3.1 standards, which set out proposals to update the Prudential Regulation Authority’s (PRA) Pillar 3 disclosure requirements to reflect its proposals on calculating Pillar 1 risk-weighted assets (RWAs). This chapter also sets out the PRA’s near-final policy on disclosure (Pillar 3) following the consultation.

7.2 In CP16/22, the PRA proposed to:

  • update disclosure requirements in respect of credit risk, market risk, credit valuation adjustment (CVA) risk, counterparty credit risk (CCR), operational risk and the output floor, as well as capital and risk management summaries;
  • modify, delete existing and introduce new disclosure templates to align with the disclosure requirements under the Basel 3.1 standards and reflect the proposals set out in the CP; and
  • continue to apply the existing proportionality approach set out in the Disclosure (CRR) Part of the PRA Rulebook whereby the frequency of disclosure is varied according to a firm’s size category and listing status.

7.3 The PRA received ten responses to its proposals on disclosure (Pillar 3) relating to:

  • clarifications on the technical detail of the disclosure templates and instructions;
  • identification of minor errors within the disclosure templates and instructions;
  • clarifications of the scope of application of disclosures about subsidiaries;
  • reduction in the frequency of quarterly disclosure templates;
  • disclosures to only be required at the highest level of consolidation for disclosing entities; and
  • out of scope feedback to the proposals set out in CP16/22, such as requests for the PRA to review disclosure requirements for mid-tier banks that would not fall in scope of the PRA’s Small Domestic Deposit Taker (SDDT) criteria for reduced disclosure.

7.4 Having considered the responses, the PRA has amended the Pillar 3 disclosure templates and instructions in certain areas. This chapter summarises the responses to CP16/22 and sets out the material amendments that the PRA had made to Pillar 3 requirements.

7.5 As described in Chapter 1 – Overview, the PRA has also made a number of less material amendments and clarifications to the proposed disclosure templates and instructions, which are not described separately in this chapter. These amendments are reflected in the near-final PRA Rulebook: CRR Firms: (CRR) Instrument [2024] in Appendix 2. Please refer to Comparison of Draft PRA Rulebook (CRR) Instrument [2023] against Near-final PRA Rulebook: CRR Firms (CRR) Instrument [2024], which contains a comparison of the near-final rules with the proposed rules as set out in CP16/22 for ease of identifying all of the changes made. The appendices to this near-final policy statement (PS), contain the PRA’s near-final policy, which will amend the Disclosure (CRR) Part of the PRA Rulebook. The PRA will update the hyperlinks to the underlying workbooks containing the updated disclosure templates and instructions in this near-final PS.

7.6 The sections below have been structured broadly along the same lines as chapter 11 of CP16/22, covering the main areas where the PRA received comments from respondents as follows:

  • Credit risk
  • Market risk
  • Credit Valuation Adjustment (CVA)
  • Counterparty Credit Risk (CCR)
  • Operational risk
  • Output Floor
  • Consequential updates to capital summary disclosures
  • Additional feedback.

7.7 Unless specified otherwise, the near-final rules are consistent with those in CP16/22 and therefore the PRA considers its analysis of its objectives and ‘have regards’ in CP16/22 with respect to the areas mentioned in the paragraph above remains appropriate.

Credit risk

Standardised approach

7.8 The PRA proposed in CP16/22 to amend existing standardised approach (SA) disclosure templates UKB CRD, UKB CR4 and UKB CR5 to broadly align the disclosures with the templates under the Basel 3.1 standards, but with UK-specific modifications to reflect the PRA’s proposed implementation of the SA approach.

7.9 The PRA received two responses to its SA disclosure proposals, which both sought minor clarification of the instructions for template UKB CR5 and its SA template instructions. The PRA has amended the proposed disclosure instructions in order to make corrections and provide further clarity.

Internal ratings based approach

7.10 The PRA proposed to amend existing disclosure templates UKB CRE, UKB CR6, UKB CR6-A, UKB CR7, UKB CR7-A and UKB CR10 to broadly align the existing internal ratings based (IRB) disclosure templates with the templates under the Basel 3.1 standards.

7.11 The PRA received two responses on:

  • minor inconsistencies in referencing in template UKB CR10 against the PRA Rulebook;
  • template UKB CMS2, which was a proposal under the output floor in CP16/22. However, the responses related to credit risk data items and respondents:
    • requested the PRA align the sub exposure classes in this template to those used under the SA and the IRB approach; and
    • sought clarification on disclosure requirements for purchased receivables, institutions that are not quasi-sovereign and the ‘other’ category.

7.12 The PRA has updated the above templates and instructions for UKB CR10 to improve consistency between the exposure class breakdown in other credit risk disclosure templates and updated the template and instructions for UKB CMS2 to improve consistency and clarity. The PRA has introduced additional fields in UKB CR7 and UKB CR7-A to align with its decision to permit certain credit risk mitigation techniques in respect of exposures subject to the slotting approach.

Market risk

7.13. The PRA proposed to introduce five new market risk disclosure templates and delete seven existing disclosure templates. Of the five new templates, two require firms to disclose qualitative information relating to firms’ market risk management framework (UKB MRA) and for Internal Model Approach (IMA) firms, details of the approved trading desks and models used (UKB MRB). The remaining three templates would require firms to provide details in relation to the market risk calculation approach they use, either the alternative standardised approach (ASA) (UKB MR1), IMA (UKB MR2) or simplified standardised approach (SSA) (UKB MR3).

7.14 The PRA received two responses to its market risk disclosure proposals. Both respondents requested clarification on whether IMA exposures should be included in the disclosure template UKB MR1 for the purpose of the output floor requirement.

7.15 The PRA can confirm that firms are not required to disclose the information in disclosure template UKB MR1 and other market risk disclosure templates for the purpose of output floor requirement.

7.16 The PRA has not amended the market risk disclosure templates and instructions and will implement the disclosure requirements as set out in CP16/22.

Credit valuation adjustment

7.17 The PRA proposed to introduce six new disclosure templates on CVA, and to delete the existing UK CCR2 template. The PRA also proposed that the six new templates would align CVA disclosures with the Basel 3.1 standards on disclosure.

7.18 The new templates would require firms to provide high level data on their capital requirements under the basic approach to CVA (BA-CVA), the standardised approach to CVA (SA-CVA), and/or the alternative approach to CVA (AA-CVA) where applicable, as well as qualitative data providing information on the methodology used.

7.19 The PRA received two responses in relation to the CVA disclosure template proposals. These related to a request to clarify the disclosure requirements of BA-CVA in UKB CVA1 and to address the inconsistent use of the term ‘capital requirements’ and ‘RWAs’ in templates UKB CVA2, UKB CVA3 and UKB CV4.

7.20 Both respondents identified a difference in the calculation formula of the systematic and idiosyncratic components for BA-CVA between the disclosure template UKB CVA1 and reporting template CAP 26.02. They noted that the disclosure instructions required firms to include a component not essential for the capital requirement for CVA and requested that the disclosure templates be amended to align to the reporting template. Additionally, one respondent asked the PRA to confirm whether netting sets calculated under the full BA-CVA and disclosed in UKB CVA2 still need to be disclosed under the reduced BA-CVA in template UKB CVA1.

7.21 For the difference in calculation components, the PRA’s proposed disclosure template is consistent with the BCBS disclosure requirements, although it differs from the PRA’s proposed reporting template. For supervisory purposes, the PRA considers the definitions of the systematic and idiosyncratic components in the reporting template to be most appropriate but that this does not warrant a change to the disclosure templates. As such, the definitions contained in UKB CV1, which are consistent with the Basel Committee on Banking Supervision (BCBS) disclosure standards, remain appropriate for disclosure and the PRA has not amended the disclosure templates and instructions for UKB CV1.

7.22 The PRA confirms that only netting sets subject to the reduced BA-CVA are disclosed in CVA1. This approach is consistent with the BCBS disclosure standards.

7.23 The PRA has retained the use of the term ‘capital requirements’, to maintain consistency with the corresponding reporting template.

7.24 The PRA will implement the CVA disclosure templates and instructions as set out in CP16/22.

Counterparty credit risk

7.25 The PRA proposed updating one existing disclosure template, UKB CCR1, to reflect the calculation of non-financial counterparties (NFCs) and pension scheme arrangement under standardised approach to CCR (SA-CCR) and simplified SA-CCR.

7.26 The PRA received no responses to these proposals and will implement these in full as set out in CP16/22. The PRA has also decided to make minor amendments to CCR3 to reflect the near-final credit risk standardised approach rules as set out in this PS.

Operational risk

7.27 The PRA proposed to delete template UK OR1, introduce three new Pillar 3 templates UKB OR1, UKB OR2 and UKB OR3 and amend existing template UK ORA in order to align to the disclosure requirements under the Basel 3.1 standards in terms of content and format.

7.28 The PRA received two responses to its proposals for operational risk disclosures, which challenged disclosure of historical loss data in template UKB OR1, given the proposed separate reporting of this information to the PRA. Respondents also sought clarification on the level of information expected in the accompanying narrative required in UKB OR1 and UKB OR3.

7.29 The PRA considers it important that the proposed operational risk templates and instructions align to the disclosure requirements under the BCBS disclosure standards and is not amending the proposed approach. The PRA has updated the instructions to templates UKB OR1, UKB OR3 and UKB OR2 to provide further clarifications, including the definition of rapid and insurance recoveries, as well as further guidance on the exclusion of loss events and the information firms should disclose in the accompanying narrative. The PRA has also made some amendments to improve clarity in the templates and instructions.

Output floor

7.30 The PRA proposed to introduce two new disclosure templates on the output floor which are aligned with the Basel 3.1 standards. The proposals included two quantitative templates, one for all firms within the scope of the output floor (UKB CMS1), and one for firms with permission to use the IRB approach for calculating credit risk (UKB CMS2).

7.31 The PRA received three responses to these proposals, and all these were in respect of template UKB CMS2. These related specifically to credit risk and the responses and PRA feedback are captured in the credit risk section of this chapter.

Consequential updates to capital summary disclosures

7.32 The PRA proposed to implement consequential amendments under the Basel 3.1 standards relevant to the existing templates UK KM1 and UK OV1 for all firms within the proposed scope of application of the Basel 3.1.

7.33 The PRA received no responses to these proposals and will implement these in full as outlined in CP16/22.

Additional feedback

7.34 The PRA received a number of general comments on disclosures, which were not specific to any template, as well as comments on disclosure topics out of scope of CP16/22. The most substantive responses covered:

  • clarification on whether third-country large subsidiaries are subject to disclosure requirements of the Disclosure (CRR) part of Rulebook based on the CRR large institution definition;
  • requests for disclosure requirements not to apply to subsidiaries;
  • a request for the PRA to consider reducing the frequency of quarterly Pillar 3 requirements for firms that do not publish quarterly financial information to half yearly because they view Pillar 3 disclosures as uninformative without the supporting financial statements;
  • a request for the PRA to review disclosure requirements for mid-tier firms and large groups;
  • acceleration of the implementation of CP4/23 – The Strong and Simple Framework: Liquidity and Disclosure requirements for Simpler-regime Firms, disclosure policy for SDDTs; and
  • comments referring to disclosure requirements out of scope of CP16/22 or not CRR disclosures.

7.35 The PRA did not propose any changes to the scope of disclosure requirements for large subsidiaries in CP16/22. Rule 1.1 of the Disclosure (CRR) Part of the PRA Rulebook sets out the scope of firms subject to the PRA’s Pillar 3 disclosure requirements. A broader review of subsidiary disclosure requirements is beyond the scope of the proposals in CP16/22.

7.36 The PRA has taken a proportionate approach to developing its disclosure requirements and so the quarterly disclosure requirements only apply to the largest firms or those firms that are listed. The proposals to require quarterly disclosures, set out in CP16/22, are aligned to the BCBS disclosure requirements. The broader disclosure frequency of Pillar 3 is out of scope of this near – final PS which is focused on implementing the Basel 3.1 specific changes to Pillar 3. Therefore, the PRA will implement the frequencies for disclosure consulted on in CP16/22.

8: Reporting

Introduction

8.1 This chapter provides feedback to responses to chapter 12 of consultation paper (CP)16/22 – Implementation of the Basel 3.1 standards, which set out proposals for how firms would report on the proposed framework for the calculation of Pillar 1 risk-weighted assets (RWAs) to the Prudential Regulation Authority (PRA). This chapter also sets out the PRA’s near-final policy on reporting following the consultation.

8.2 In CP16/22, the PRA proposed to amend existing and introduce new reporting requirements, as well as delete redundant reporting in order to align reporting with the revised approaches to RWA calculation under the Basel 3.1 standards. CP16/22 proposed to:

  • revise existing common reporting (COREP) templates and instructions on own funds, and own funds’ requirements to reflect the CP proposals;
  • delete certain COREP templates that would become obsolete under the CP proposals;
  • introduce new COREP templates to reflect the proposed new Pillar 1 RWA calculations, and proposed internal model use conditions;
  • delete the FSA005 Market risk template to reflect the proposed discontinuation of the ‘risks not in value-at-risk’ (RNIV) methodology for the calculation of market risk; and
  • revise the Capital+ templates and instructions to reflect the CP proposals.

8.3 The PRA received ten responses to its proposals on reporting. Respondents were generally supportive of the reporting proposals and the responses covered:

  • reduction to the frequency or removal of the historical loss and internal loss multiplier (ILM) reporting requirements;
  • clarification of certain reporting requirements;
  • identification of minor errors within the reporting templates and instructions; and
  • guidance on reporting processes beyond the proposed rules-based policy.

8.4 Having considered the responses, the PRA has decided to amend the draft rules in certain areas. This chapter describes the comments and amendments that the PRA considers are more material. As described in Chapter 1 – Overview, the PRA has also made a number of less material amendments and clarifications to the draft rules, including the reporting templates and instructions, which are not described in this chapter. These amendments are reflected in the near-final PRA Rulebook: CRR Firms: (CRR) Instrument [2024] in Appendix 2. Please refer to the document titled Comparison of Draft PRA Rulebook (CRR) Instrument [2023] against Near-final PRA Rulebook: CRR Firms (CRR) Instrument [2024], which contains a comparison of the near-final rules with the draft rules as set out in CP16/22.

8.5 The appendices to this near-final policy statement (PS), contain the PRA’s near-final policy, which will:

  • amend the Reporting (CRR) Part of the PRA Rulebook (Appendix 2);
  • amend the Regulatory Reporting Part of the PRA Rulebook (Appendix 2);
  • amend the Reporting Pillar 2 Part of the PRA Rulebook (Appendix 2); and
  • amend supervisory statement (SS) 34/15 – guidelines for completing regulatory reports (Appendix 15).

8.6 The sections below have been structured broadly along the same lines as chapter 12 of CP16/22, covering the main areas where the PRA received comments from respondents as follows:

  • Credit risk
  • Market risk
  • Credit valuation adjustment (CVA)
  • Counterparty credit risk (CCR)
  • Operational risk
  • Output floor
  • Capital+
  • Taxonomy implementation
  • Additional feedback

8.7 Unless specified otherwise, the near-final rules are consistent with those in CP16/22 and therefore the PRA considers its analysis of its objectives and ‘have regards’ in CP16/22 with respect to the areas mentioned in paragraph 8.6 above remains appropriate.

8.8 The near-final updated templates and instructions are attached to this near-final PS, (see Appendix 14), alongside amendments to SS34/15 – guidelines for completing regulatory reports (Appendix 15) which include minor restructuring within the annexes for clarity, and the reporting rule instruments (Appendix 2).

Credit risk

Standardised approach

8.9 For firms that apply the standardised approach (SA), the PRA proposed in CP16/22 to amend three existing COREP templates and instructions to align with the proposed changes to the SA.

8.10 The PRA received two responses. The points raised were that:

  • the inclusion of counterparty credit risk within credit risk SA reporting (template CAP 07.00) was burdensome and that counterparty credit risk should be reported separately;
  • the PRA could request a more granular breakdown of retail exposures in template CAP 09.01; and
  • the PRA’s proposed policy for risk weighting exposures subject to unfunded credit protection (UFCP) would be overly burdensome to implement and as a consequence would result in the submission of poor-quality data.

8.11 The PRA considers that the inclusion of the CCR reporting within the 07.00 template continues to be useful for the PRA’s purposes. However, in due course, the PRA plans to review the full range of bank reporting data it collects with a view to making improvements and efficiencies. Similarly, the PRA considers that the risk weight breakdown of retail exposures already provided in template 07.00 is sufficient, without adding further granularity to template 09.01.

8.12 As set out in Chapter 4 – Credit risk mitigation, the PRA has made minor amendments to the calculation of RWAs, relating to exposures subject to UFCP. The PRA, as a result, has amended reporting instructions on covered and uncovered portions to align with this change and to make the data reported more relevant.

8.13 The PRA has also amended credit risk SA reporting to reflect responses received on the SA policy as part of CP16/22, which is explained separately in this near-final PS. The PRA has:

  • introduced a row in CAP 07.00 to report the currency mismatch multiplier;
  • included a column in CAP 07.00 for the reporting of the new conversion factor of 40% for other commitments and for a breakdown of specialised lending project finance exposures; and
  • made several adjustments to template CAP 07.00 relating to the reporting of real estate exposures to reflect structural changes in the underlying policy.

Internal ratings based approach

8.14 For firms that apply the internal ratings based (IRB) approach, the PRA proposed in CP16/22 to amend ten existing COREP templates and instructions to align with proposed changes to the IRB framework.

8.15 The PRA received two responses to its IRB reporting proposals. The responses identified minor errors in the templates and instructions. The PRA has corrected the errors and reflected changes in the near-final reporting templates and instructions.

8.16 The PRA has made amendments to the near-final IRB policy, as set out in this near-final PS. These amendments have resulted in further changes to IRB reporting in respect of templates C08.01, 08.02, 08.03, 08.05, 08.05.01, 08.06, 08.07, 09.02 and 34.07. The PRA has:

  • removed references to quasi-sovereign exposure for which the PRA has decided to withdraw the IRB approach as set out in Chapter 3 – Credit risk – internal ratings based approach; and
  • introduced new columns to allow recognition of funded credit protection (FCP) for exposures subject to slotting and introduced data points to reflect policy which the PRA has decided to introduce in the framework.

8.17 The PRA has made several amendments to template C08.07 and the accompanying instructions, to require firms to report data that reflect their use of ‘permanent partial use’ (PPU) of the SA for some exposures, accurately reflecting the PRA’s requirements. The PRA considers that these amendments will improve the accuracy of firms’ reporting of PPU and its ability to supervise firms’ compliance with PPU requirements.

Market risk

8.18 In CP16/22, the PRA proposed to:

  • introduce ten new reporting templates on the new market risk advanced standardised approach (ASA) and three new reporting templates on the new market risk internal model approach (IMA);
  • introduce a summary template on the size of a firm’s balance sheet and the relevant approaches applied to calculate the firm’s market risk capital requirement;
  • r