PS10/24 – Review of Solvency II: Reform of the Matching Adjustment

Policy statement 10/24

Published on 06 June 2024

1: Overview

1.1 This Prudential Regulation Authority (PRA) policy statement (PS) provides the PRA’s feedback to responses received to the consultation paper (CP) 19/23 – Review of Solvency II: Reform of the Matching Adjustment. Published in September 2023, CP19/23 was the second major PRA consultation to deliver significant reforms to Solvency II and implement the conclusions of the Government’s Review of Solvency II: Consultation - Response (‘November 2022 statement’). The CP set out the PRA’s proposed reforms that will enable broader and quicker investment by insurers in their matching adjustment (MA) portfolios, while improving responsiveness to risk and enhancing firms’ responsibility for risk management.

1.2 This PS also contains the PRA’s final policy as follows:

a new Matching Adjustment Part of the PRA Rulebook (Appendix 1);
amendments to the Technical Provisions, Conditions Governing Business and Glossary Parts of the PRA Rulebook (Appendix 1);
updated supervisory statement (SS) – SS7/18 ‘Solvency II: Matching adjustment’ (Appendix 3);
updated SS8/18 – Solvency II: Internal models – modelling of the matching adjustment (Appendix 4);
updated SS3/17 – Solvency II: Illiquid unrated assets (Appendix 5);
updated SS1/20 – Solvency II: Prudent Person Principle (Appendix 6);
updated SS11/16 – Solvency II: External audit of, and responsibilities of the governing body in relation to, the public disclosure requirement (Appendix 7);
a new statement of policy (SoP) – Solvency II: Matching Adjustment Permissions (Appendix 8) (‘MA SoP’); and
minor amendments to the PRA’s existing SoP – The PRA’s approach to the publication of Solvency II technical information (Appendix 9) (‘TI SoP’).

Additionally, this PS contains updated near-final rules amending the Reporting Part of the PRA Rulebook (Appendix 2).^{footnote [1]}

1.3 This PS is relevant to all UK Solvency II firms, the Society of Lloyd’s and its members and managing agents, and insurance and reinsurance undertakings that have a UK branch (third country branch undertakings) where they are applying or have applied to use the MA. This PS will refer to these collectively as ‘insurers’ or ‘firms’ unless otherwise specified.

Background and developments since CP19/23

1.4 Chapter 1 (Overview) of CP19/23 set out the background to the MA, the key benefits envisaged by the CP19/23 reforms and the PRA’s approach to MA permissions. As explained in CP19/23, the new UK prudential regime for insurers will eventually be known as ‘Solvency UK’. However, for clarity and consistency the PRA will continue to refer to the regime as Solvency II until such time as all references to Solvency II can be changed across all relevant materials. Therefore, this PS will continue to make reference to Solvency II, where relevant.

1.5 The Government’s legislation The Insurance and Reinsurance Undertakings (Prudential Requirements) Regulations 2023, (‘IRPR Regulations’) was made on 7 December and then laid before Parliament on 8 December 2023, and will come into force for firms on 30 June 2024. These regulations set out the Government’s reforms to the Solvency II MA. This legislation is consistent with the position described by the PRA in Chapter 1 of CP19/23 and reflects discussions between the PRA and the Government on this issue. In particular, the legislation maintains most of the existing methodology and calibration of the fundamental spread (FS), broadens the MA eligibility criteria to allow assets with highly predictable (HP) cash flows, reforms the powers of the PRA where firms breach MA eligibility conditions and creates space for the PRA to make rules on other specific MA policy aspects.

1.6 The final policy in this PS will implement and work alongside the Government’s MA reforms and will come into force for firms on 30 June 2024. This final policy is intended to improve the way that the MA supports investment and to maintain a high level of prudential standards for the insurance sector and protection of insurance policyholders, within the framework of the legislation on the MA.

1.7 On 28 February 2024, the PRA published PS2/24 – Review of Solvency II: Adapting to the UK insurance market and on 29 February 2024, the PRA published PS3/24 – Review of Solvency II: Reporting and disclosure phase 2 near-final. These set out the PRA’s final policy in the form of near-final rules and updated near-final policy materials on the other key areas covered by the Solvency II Review. In particular, PS2/24 provides an update to the PRA’s overall consultation plans for reforms to Solvency II that were in CP12/23 – Review of Solvency II: Adapting to the UK insurance market.

1.8 In advance of the PRA’s final policy set out in this PS, the PRA published a statement^{footnote [2]} (‘April statement’) on 15 April 2024 to support firms in their preparations for implementation of the MA reforms. The PRA’s intention in this statement was to help firms to make the most of new investment opportunities facilitated by the reforms in an efficient way.

1.9 Specifically, the April statement provided clarification on:

the PRA’s expectation that, as mentioned in previous Government statements, existing MA approvals will continue to be valid under the reformed MA regime;^{footnote [3]}
the scope of the PRA’s proposals in CP19/23 regarding assets with HP cash flows, and that changes to the PRA’s policy on assets that are considered to have fixed cash flows prior to the new MA regime were not intended to be in scope of CP19/23; and
additional details relating to the implementation of some of the new requirements.

1.10 On 18 April 2024, the Government made a statutory instrument^{footnote [4]} (SI) under section 138BA of the Financial Services and Markets Act 2000 (FSMA), which will enable the PRA to exercise the power in section 138BA in relation to most of its rules made under FSMA. The power in section 138BA allows the PRA, on application or with the consent of a firm that is subject to PRA rules, to give the firm a permission that enables it not to apply the rules, or to apply the rules with a modification specified in the permission. The SI will come into force on 30 June 2024. As proposed in CP19/23, the PRA will use this power to grant permissions to apply the MA.

1.11 On 22 April 2024, the PRA published CP5/24 – Review of Solvency II: Restatement of assimilated law setting out proposals to restate the remaining firm-facing Solvency II requirements from assimilated law (which is being revoked by the Government under its Smart Regulatory Framework (SRF)) to the PRA Rulebook and other policy materials.

1.12 The Government’s legislation The Insurance and Reinsurance Undertakings (Prudential Requirements) (Transitional Provisions and Consequential Amendments) Regulations 2024 was made on 1 May 2024. This legislation provides that MA approvals granted under Regulation 42 of the Solvency 2 Regulations 2015 prior to 30 June 2024 will become a permission granted by the PRA under section 138BA, ensuring that firms with existing MA approvals can continue to apply the MA after 30 June 2024. The PRA therefore does not expect firms to reapply for permission to apply the MA when such permission has been provided prior to 30 June 2024, unless required due to changes to their MA portfolios to document compliance with the new regime at the point of implementation. Firms with in-flight MA applications should discuss these bilaterally with their usual PRA supervisory contact.

1.13 In CP19/23 the PRA proposed the following reforms to the Solvency II MA:

Improving business flexibility

Widening the range of investments that firms may hold in MA portfolios, by providing a clear framework to permit the inclusion of assets that do not have fixed cash flows. The proposals were designed to increase the incentives for insurers to invest in a wider range of long-term, productive assets, including assets with construction phases.
Expanding the types of insurance business that may claim MA, to permit more insurance liabilities to benefit from the MA.
Removing the limit on the amount of MA that may be claimed from sub-investment grade (SIG) assets, to facilitate more investments close to and below the boundary between investment and SIG assets.

Being more responsive to the level of risk

Establishing a streamlined MA application process for a range of suitable assets, proportionate to risk. The proposals were designed to improve the efficiency of some MA applications, allow firms to move more quickly when investment opportunities arise and reduce the regulatory burden.
Making the regulatory treatment of breaches of MA conditions more proportionate, to provide for more flexible and proportionate consequences.
Increasing the granularity of the FS, where appropriate, to reflect differences in credit quality of firms’ assets by rating notch and to improve the risk sensitivity of the FS used to calculate technical provisions (TPs) while being pragmatic and proportionate by giving firms some flexibility of approach.

Enhancing firms’ responsibility for risk management

Introducing an attestation process for the amount of MA benefit being claimed, to ensure that firms are accountable for the MA, with an FS that is sufficient for the risks in their own portfolio of assets.
Clarifying expectations around the risk management of SIG assets, to promote good risk management and facilitate greater investment freedom.
Formalising the data submitted to the PRA by firms on the assets and liabilities in their MA portfolios, to gather more structured regular information on the type of assets and the quantum of the MA benefit arising from them, through a new Matching Adjustment Asset and Liability Information Return (MALIR).
Converting expectations on internal credit assessments to requirements, to complement and reflect the new structure of the legislation in this area.
Introducing an MA eligibility condition for firms to be able to demonstrate compliance with the Prudent Person Principle (PPP), to show how firms have assessed the suitability and the risks within the assets held in MA portfolios.

Summary of responses

1.14 The PRA received 24 responses to the CP. The names of respondents to the CP who consented to their names being published are set out in Appendix 10. A summary of the responses is below, and further details and the PRA’s feedback can be found in the individual chapters that follow.

1.15 Respondents were generally supportive of the PRA’s proposals in CP19/23, with comments welcoming its general direction of travel and agreeing that it met the Government’s Solvency II objectives outlined in the November 2022 statement.

1.16 Respondents to the CP were appreciative of the PRA’s pre-consultation engagement, in particular the helpfulness of the Subject Expert Groups (SEGs). However, respondents also raised concerns with the tight timelines for implementation of the MA reforms.

1.17 Respondents raised some general points in relation to CP19/23. Those general points, as well as those relating to Chapter 1 (Overview) of CP19/23, are addressed in Chapter 12 of this PS. The responses relevant to specific proposals are addressed in the relevant chapters of this PS.

1.18 A summary of the overall responses to each chapter in CP19/23 is set out below.

Respondents were supportive of certain areas of the proposals in the investment flexibility chapter, such as the proposals to expand the asset eligibility criteria, the design of the additional matching tests and the design of the standard approaches to FS additions. However, some respondents considered that the overall controls framework for assets with HP cash flows was too onerous, and that the PRA should have proposed wider criteria for asset eligibility. Some respondents were also concerned that changes to existing policy could reclassify a material proportion of their existing MA eligible assets (which were previously deemed to have fixed cash flows) as assets with HP cash flows, and reduce firms’ scope to invest in new productive assets.
The proposed expansion of the liability eligibility criteria was generally welcomed. Several respondents commented that the proposed PRA rules would lead to only individual in-payment income protection policies being eligible for inclusion into MA portfolios, leaving group policies ineligible. Some respondents proposed that the eligibility criteria should permit group dependant annuities (GDAs) in MA portfolios.
Most respondents were supportive of the proposed removal of the cap on MA benefit for SIG assets but comments or requests for clarification were made regarding the proposed expectations in respect of risk and investment management, the PPP and the treatment of SIG assets in firms’ internal models.
Respondents generally welcomed and supported the internal credit assessment proposals. Most of the responses related to the need for proportionality in applying the rules and expectations and requests for more detailed guidance on external assurance of their internal credit ratings.
Most respondents welcomed the proposals in the MA permissions, breaches and consequential rule changes chapter, but requested additional changes and clarifications. In particular, several respondents sought clarification over the treatment of the Solvency Capital Requirement (SCR) for firms in breach of MA eligibility conditions and asked that the PRA further simplify the application process, including documentation requirements.
Respondents were supportive of some elements of the MA attestations proposal, including removal of liquidity premium from the attestation wording, and the proportionate approach to reviewing the FS. However, the majority of respondents requested a less onerous attestation approach for corporate bonds and asked that offsets between assets should be taken into account when determining the size of any FS additions.
Several respondents supported the overall objective of publishing the assumptions underlying the MA to improve consistency and transparency. Some concerns were expressed around application of the assumptions to non-corporate bond assets, consistency between the assumptions and their use in setting capital add-ons.
The formalisation of the MALIR data collection was supported by several respondents. However, many expressed concerns over the burden of the MALIR on firms and suggested some ways to reduce this.
Respondents welcomed the notching proposals but requested several areas of clarification. Respondents raised particular concerns around the timings associated with the implementation of notching.
Although some respondents agreed with the PRA’s cost benefit and have regards analysis, several respondents suggested that the PRA’s estimate of the ongoing compliance costs was potentially underestimated, with costs being considered disproportionate for firms with smaller MA portfolios. Many also commented on the impact of proposals on the PRA’s secondary objectives.

1.19 Responses to the CP also suggested that the PRA should go further than the proposals in CP19/23 by adopting one or more regulatory ‘sandbox’ ideas, which would permit assets to be included up to a limited proportion of an MA portfolio without (or prior to) receiving permission from the PRA. Proposals of this nature were not within scope of CP19/23 and so are not covered in this PS. The PRA also considers that, before taking any such approach, it would need to consider in more detail how the proposals would advance the PRA’s objectives. In addition, any such proposals would have to be subject to future consultation, in order to allow all stakeholders to provide their views. In the meantime, the PRA has convened a SEG with a panel of insurers to explore these issues further and to help inform the PRA’s thinking on any future proposals. Further details are provided in Chapter 5 of this PS.

Changes to draft policy

1.20 Where the final rules differ from the draft in the CP in a way that is, in the opinion of the PRA, significant, FSMA^{footnote [5]} requires the PRA to publish:

details of the differences together with an updated cost benefit analysis (CBA); and
a statement setting out in the PRA’s opinion whether or not the impact of the final rules on mutuals is significantly different from: (a) the impact that the draft rule would have had on mutuals; or (b) the impact that the final rule will have on other PRA-authorised firms.

1.21 The PRA is grateful for the responses to CP19/23, which it has carefully considered. Having done so, the PRA has identified a number of areas where it will make adjustments to the draft policy, to reflect its consideration of the responses received. The most material changes include:

amendments to the proposed changes to SS7/18 to clarify the PRA’s policy intent about the treatment of existing assets with ‘fixed’ cash flows and to ensure that no unintended changes arise on those assets (see Chapter 2 of this PS);
a more flexible calibration of the additional matching tests in SS7/18 to accommodate additional investment capacity in assets with HP cash flows (see Chapter 2 of this PS);
amendments to the final rules to ensure all in-payment income protection business and in-payment GDAs are MA eligible (see Chapter 3 of this PS);
an additional chapter in the new MA SoP on MA permissions to outline the PRA’s monitoring of the MA permissions framework, including its commitment to publish regular reports on the MA framework covering application review and approval rates (see Chapter 5 of this PS);
removal of references in SS7/18 and the MA SoP to ‘new risks’ triggering variations of MA permissions (see Chapter 5 of this PS);
amendments to the MA SoP for firms submitting MA applications, including reduced documentation requirements (see chapter 5 of this PS);
amendments to SS7/18 for firms' analysis of corporate bond portfolios, where firms will not be expected to consider risks that may not have been adequately captured by the corporate bond historical credit performance data used to calibrate the FS (see Chapter 6 of this PS);
clarification in SS7/18 that firms may perform an initial top-down analysis of assets by homogenous risk groups in MA attestations, followed by an examination of specific assets where necessary, to strike a balance between practicality and granularity (see Chapter 6 of this PS);
removal from the MALIR instructions of the requirement for firms to submit detailed cashflow data extending beyond 50 years within the MALIR reporting (see Chapter 8 of this PS); and
moving back the implementation date for notching rules to 31 December 2024 (see Chapter 9 of this PS).

1.22 Further details on all issues raised in responses, and any related amendments to the policy, are set out in the relevant chapters of this PS. The PRA considers the changes made to the policy are appropriate and improve the final rules, near-final rules and final policy materials, in a manner that aligns with the PRA’s statutory objectives.

1.23 The PRA considers the costs and benefits of the final rules, near-final rules and final policy in this PS do not significantly differ overall from those derived from the draft policy proposed in CP19/23 and, therefore, the aggregated CBA presented in Chapter 10 (Cost benefit analysis) of the CP remains appropriate.

1.24 While there are no substantial changes to the overall CBA, the PRA considers that the changes to policy could bring further or alternative costs or benefits to firms in some areas. For example, the PRA considers that the simplifications to the MA attestation requirements for homogenous risk groups and corporate bond portfolios, simplifications to the MALIR template and widening of the MA liability eligibility criteria to include GDAs would reduce ongoing compliance costs for firms. However, costs are estimated to marginally increase due to a change in policy regarding auditors’ focus on the correct application of mandatory FS additions.

1.25 Further details regarding any potential changes to the CBA are provided in Chapter 10 (Cost benefit analysis) in this PS, alongside the PRA’s feedback to specific points raised regarding the CBA in response to the CP.

1.26 The PRA also considers that additional consequential amendments to the PRA Rulebook and policy materials to those proposed in CP19/23 are required to update relevant cross-references for consistency with upcoming changes in legislation. Regulation 4B of the Solvency 2 Regulations 2015 (‘Regulation 4B’) currently sets out requirements relating to the PRA’s duty to publish Technical Information (TI), which includes that relating to the volatility adjustment (VA).^{footnote [6]} This regulation will be revoked on 30 June 2024,^{footnote [7]} and will be replaced by Regulation 3 of the IRPR Regulations (‘Regulation 3’). Regulation 3 will cover the PRA’s duty to publish VA TI, but at a much less detailed level than is currently set out in Regulation 4B (3) to (7). Furthermore, the PRA has already published proposals relating to the treatment of paragraphs (3) to (7) of Regulation 4B within CP5/24.^{footnote [8]} However, the PRA has proposed that those changes, subject to the consultation process, would apply from 31 December 2024. Therefore, for the intervening period, the PRA intends to maintain its current approach for the production of VA TI (ie in line with the requirements in paragraphs (3) to (7) of Regulation 4B).

1.27 In order to provide certainty for firms that use the VA between 30 June and 31 December 2024, the PRA has made minor consequential amendments to Rules 8.1 and 8.4 of the Technical Provisions Part of the PRA Rulebook and to the definition of ‘volatility adjustment’ in the PRA Rulebook Glossary, to replace existing references to Regulation 4B with references to new Regulation 3, which is scheduled to come into force from 30 June 2024.

1.28 Furthermore, the PRA has made minor consequential changes to its TI SoP (Appendix 9) to replace existing references to Regulation 4B with references to Regulation 3, which is scheduled to come into force from 30 June 2024. This change will take effect from that date, while the additional changes to the TI SoP proposed by the PRA in CP5/24, subject to the consultation process, are intended for implementation from 31 December 2024.

1.29 The PRA does not consider that the impact of the final policy, near-final rules and final rules, in this PS would have a significantly different impact on mutuals relative to the impact of the draft policy and rules on mutuals,^{footnote [9]} or on other PRA-authorised firms.

Accountability framework

1.30 Before making any proposed rules, the PRA is required by FSMA to comply with several legal obligations, including to have regard to any representations made to it, and to publish an account, in general terms, of those representations and its feedback to them.^{footnote [10]} In this PS, the ‘Summary of responses’ section above contains a general account of the representations made in response to CP19/23 and the ‘Feedback to responses’ section within each chapter contains the detailed responses and the PRA’s feedback for each policy area.

1.31 When making rules, the PRA is also required to consider responses to consultation and publish an explanation of the PRA’s reasons for believing that making the proposed rules is compatible with its objectives and with its duty to have regard to the regulatory principles.^{footnote [11]} In CP19/23, the PRA set out details of the applicable accountability framework in Chapter 1 (Overview). The PRA also provided an assessment of relevant considerations for the proposed reforms against its objectives separately in each chapter. Where the PRA has made changes to the draft policy proposed in CP19/23, it considers that generally this analysis provided against the PRA’s primary and secondary objectives continues to apply, unless otherwise explained in the relevant chapters of this PS.

1.32 The PRA considers that the final rules, near-final rules and final policy have taken into account the wide range of points raised by respondents, and continue to support the ‘have regards’ analysis as set out in CP19/23. In most cases, the PRA considers that the responses to the consultation, and associated changes to the final rules, near-final rules and final policy, did not significantly alter its consideration of the matters to which it must have regard in implementing the final policy. Where the consideration of the matters to which the PRA must have regard changed in relation to any proposed reforms, further explanation is provided in Chapter 11 (Have regards analysis) of this PS.

Structure of the PS

1.33 The PRA’s feedback to responses received to the proposed reforms in CP19/23 and an explanation of the final policy, near-final rules and final rules are structured into the following chapters, which correspond with the chapters in the CP. Chapter 12 has been added to cover general responses received.

Chapter 2 – Investment flexibility
Chapter 3 – Liability eligibility
Chapter 4 – Credit ratings under the MA
Chapter 5 – MA permissions, breaches and consequential rule changes
Chapter 6 – Matching adjustment attestation
Chapter 7 – Assumptions underlying the MA
Chapter 8 – Matching adjustment asset and liability information
Chapter 9 – Notching
Chapter 10 – Cost benefit analysis
Chapter 11 – Have regards analysis
Chapter 12 – General points raised by respondents

1.34 Appendices 1 and 2 of this PS set out amended final and near-final rules for the reforms set out in Chapters 2 to 9 of this PS. Final versions of the relevant SSs and SoPs arising from the reforms in this PS are included in the other appendices, as explained in the individual chapters.

1.35 Within this PS, to increase the readability of each chapter, a number of abbreviations have been defined in full upon first usage.

Implementation

1.36 The implementation date for final rules and policy materials reflecting policy changes set out in this PS is 30 June 2024, unless otherwise stated. The PRA considers that implementation of the MA reforms on 30 June 2024 will mean that firms that use the MA will be able to take advantage of these MA reforms in advance of the remainder of the Solvency II reforms, which will be implemented on 31 December 2024.

1.37 In summary, this means that from 30 June 2024:

firms can apply for permission to include assets and liabilities in MA portfolios based on the expanded asset and liability eligibility criteria;
firms can remove the SIG MA cap;
the PRA expectations on internal credit assessments come into effect;
the changes to the MA permissions and breaches come into effect;
firms can choose to implement notching; and
firms may at their own discretion start to apply voluntary FS additions, although the PRA notes that firms are likely to need to take account of their attestation processes, which are not required to be in place until the first MA attestation date. Furthermore, given the short time between the publication of this PS and the 30 June implementation date, the PRA understands that firms are unlikely to be in a position to apply any voluntary FS additions to the 30 June 2024 balance sheet.

1.38 Subsequently:

the rules and expectations relating to notching in the calculation of the TPs will come into effect on 31 December 2024 as the transitional arrangements in place expire;
the first required attestation for any firm with MA permission prior to 31 December 2024 will be in respect of its first financial year end from 31 December 2024 (or material change in risk profile with an attestation reference date after 31 December 2024, if sooner);
voluntary FS additions should be considered as part of the first MA attestation, subject to the last bullet point of the previous paragraph; and
the MALIR rules come into effect from 31 December 2024 and each firm will have 130 business days from its respective financial year end (noting this may be 31 December) to complete and submit its first formalised MALIR to the PRA.

1.39 Based on the responses received on CP19/23 on the practical implementation of the MA reforms and further points raised by the insurance industry, the PRA has made some updates to SS7/18 and its new MA SoP, in Appendices 3 and 8 respectively. These clarify and expand on practical implementation details. This includes changes that are not necessarily specific to any of the proposals in CP19/23 and are summarised below with further details given in Chapter 5 of this PS (if relevant):

the PRA has reduced the extent of documentary evidence that firms will generally be expected to submit with an MA application, noting in particular that for all MA applications firms are only expected to provide confirmation of their compliance with eligibility conditions in respect of the rateability of assets (Regulation 4(4) of the IRPR Regulations) and the PPP (Rule 2.2(6) of the Matching Adjustment Part of the PRA Rulebook). If other information is needed in respect these eligibility conditions, then the PRA would request this during its review;
the PRA has made changes to SS7/18 to remove references to information expected to be provided as part of an MA application. The information expected to be provided by firms as part of their MA applications is instead set out in the MA SoP;
the PRA expects firms making an MA application to use the relevant supplementary information forms published on its website. The PRA considers that these forms should help firms in making MA applications. The supplementary information form would be submitted alongside the section 138BA application form and the firm’s suite of documentary evidence (the latter with track changes as appropriate); and
the PRA has developed an Application Readiness Assessment Process (ARAP) to support engagement between the firm and the PRA in advance of a formal MA application submission in the new MA regime. This will assist the PRA in assessing the firm’s readiness to submit a formal MA application and help the PRA determine resourcing needs.

1.40 The PRA published reporting templates effective from 31 December 2024 in PS3/24. These templates did not reflect any changes in this PS. Template IRR.22.03 R0050 reports the increase of the FS for SIG assets. This adjustment is no longer required by MA rules and R0050 should be reported as zero from 31 December 2024. The instructions for template IRR.22.03 R0050 will be updated to reflect this change before 31 December 2024 as part of the ‘Review of Solvency II: Restatement of assimilated law’ PS.

1.41 In addition to the proposals set out in CP19/23, references related to the UK’s membership of the EU in SS7/18, SS8/18, SS3/17 and SS1/20 covered by the policy in this PS have been updated as part of this PS to reflect the UK’s withdrawal from the EU. Unless otherwise stated, any remaining references to EU or EU-derived legislation refer to the version of that legislation which forms part of assimilated law.^{footnote [12]}

1.42 The PRA has also made a number of minor changes to SS11/16, SS3/17, SS7/18, SS8/18, SS1/20 and the new MA SoP to conform with PRA house style and enhance readability; these do not change the policy or the meaning of the text.

2: Investment flexibility

2.1 This chapter provides feedback to responses relating to the proposals in Chapter 2 (Investment Flexibility) of CP19/23. It also contains the PRA’s final policy, as follows:

amendments to the Glossary Part of the PRA Rulebook (Appendix 1);
introduce particular elements of the new Matching Adjustment Part of the PRA Rulebook (Appendix 1);
introduce particular elements in the SoP - Solvency II: Matching Adjustment Permissions (Appendix 8);
updates to SS3/17 – Solvency II: Illiquid unrated assets (Appendix 5);
updates to SS7/18 – Solvency II: Matching adjustment (Appendix 3);
updates to SS8/18 – Solvency II: Internal models – modelling of the matching adjustment (Appendix 4); and
updates to SS1/20 – Solvency II: Prudent Person Principle (Appendix 6).

2.2 In chapter 2 of CP19/23 the PRA proposed to introduce new rules and expectations in relation to:

criteria for the inclusion of a wider range of assets in firms’ MA portfolios beyond those currently eligible;
controls on the quality of matching to account for the additional sources of cash flow uncertainty introduced by the extension to asset eligibility;
approaches for the determination of FS additions for the additional retained risks arising from non-fixed cash flows on these additional assets; and
approaches for the determination of best estimate cash flows for these additional assets.

2.3 The PRA received 21 responses to Chapter 2 of CP19/23. The responses were generally supportive of the proposals to expand the asset eligibility criteria, the design of the additional matching tests and the design of the standard approaches to FS additions. The most common responses related to the matters below. These are addressed in detail later in this chapter:

that the PRA should have proposed wider criteria for asset eligibility;
that the proposed changes to SS7/18 in relation to assets with certain types of issuer optionality conflicted with existing expectations and were likely to cause affected firms to need to reclassify a material proportion of their existing fixed cash flow assets as assets with HP cash flows;
that the proposed controls framework for assets with HP cash flows could be modified, for example by reassessing the thresholds for the proposed matching tests; and
that the proposed standard approaches for additions to the FS should be modified.

Respondents made several other observations and requests for clarification that are also set out later in this chapter.

Changes to draft policy

2.4 After considering the responses, the PRA has made changes to the draft policy. A summary of the changes is set out below:

paragraphs 2,12A, 2.12E^{footnote [13]}, 2.21, 2.21A, 2.21B, 2.22 and 2.23A of SS7/18 have been amended or deleted to clarify the PRA’s policy intent for assets with ‘fixed’ cash flows;
paragraphs 2.55B and 2.57A of SS7/18 have been amended to clarify the PRA’s policy intent for assets that have been internally restructured;
paragraphs 4.10A, 4.10B, 8.4 and Appendix 1 of SS7/18 have been amended to address responses on the controls framework for assets with HP cash flows, including an increase in the thresholds for Matching Test 4 and 5 from 3% to 5%;
paragraphs 4.1D, 5.20, 5.22A^{footnote [14]}, 5.23, 5.24, 5.24A, 5.24B, 5.24C^{footnote [15]} and 5.30 of SS7/18 have been amended to address responses on the determination of FS additions for assets with HP cash flows; and
paragraph 4.1M of SS7/18 has been amended to address responses on how best estimate cash flows may be determined for assets with HP cash flows.

These changes are explained in more detail in the ‘Feedback to responses’ section of this chapter.

2.5 In considering specific responses, the PRA has proposed a recalibration of the proposed additional matching tests to reflect a change in its baseline assumptions in relation to the current level of matching within firms’ MA portfolios (see paragraph 2.51). This, along with any other significant changes to the CBA or ‘have regards’ analysis arising from changes to the draft policy proposed in CP19/23, are considered in Chapter 10 (Cost Benefit Analysis) and Chapter 11 (Have regards analysis) of this PS.

Feedback to responses

2.6 The PRA has considered the responses to Chapter 2 of CP19/23. This section sets out the PRA’s feedback to those responses and its final decisions.

2.7 The sections below have been structured along the same lines as Chapter 2 of the CP, with some areas rearranged to facilitate the responses to related issues. The responses have been grouped as follows:

criteria for inclusion of assets with HP cash flows;
controls on the quality of matching;
fundamental spread additions for assets with HP cash flows;
determination of best estimate cash flows for assets with HP cash flows; and
modelling assets with HP cash flows under stress.

Criteria for inclusion of assets with HP cash flows

2.8 In CP19/23 the PRA proposed criteria that assets must meet in order to be considered as having HP cash flows. The PRA also proposed an additional MA eligibility condition, relevant to all assets, that firms must demonstrate that the assets can be managed in line with the PPP (see Chapter 5 of this PS). Finally, the PRA proposed changes to the sections of SS7/18 covering internal restructurings to incorporate the inclusion in MA portfolios of assets with HP cash flows, and to set out additional or updated expectations for internally restructured assets.

2.9 One respondent considered that the PRA should not permit assets with HP cash flows to be included in MA portfolios, noting the undermining of the fixity principle, the uncertainty of the FS addition and the inclusion of such assets being a barrier to effective competition.

2.10 The IRPR Regulations have permitted the inclusion of assets without fixed cash flows in MA portfolios. The PRA considers that the proposals in CP19/23 for assets with HP cash flows advances its statutory objectives within the framework set by this legislation, and that it also gives firms greater certainty that their new investments will meet the PRA’s expectations. Having considered the response, the PRA has decided not to change the draft policy.

2.11 Two respondents asked the PRA to clarify the meaning of paragraph 2.73 in CP19/23, which stated ‘If firms invest in increasing volumes of assets with HP cash flows, the PRA will keep under review whether it may be appropriate to set out further expectations based on firms’ observed practices’.

2.12 The PRA clarifies that this is a statement of general PRA policy to allow for potential future market and industry developments. Having considered the responses, the PRA has decided not to change the draft policy.

2.13 Five respondents asked the PRA to clarify what contractual bounding means, or for examples of assets that would previously not have been eligible and are now eligible for inclusion in MA portfolios.

2.14 The PRA considers that where asset cash flows are not fixed, contractual bounding is achieved where the legal documentation underlying a bond or loan sets out a finite range for the cash flow timings and amounts, for example:

the cash flow profile (with the payment dates and amounts, or how the cash flow amounts are to be calculated);
the situations where the cash flow profile may or must be varied by the issuer; and
where the cash flows can be varied, the amount and timing of the varied cash flows.

2.15 The PRA’s implementation of the MA already permits certain investments, such as callable bonds, which can be included in MA portfolios by the use of specific treatments or assumptions to determine the portion of cash flows that are fixed. Such assets may now be permitted in MA portfolios where they meet the HP criteria, without the application of these prudent assumptions. Other assets, such as those with entirely unfettered issuer or third-party optionality, will also be permitted in MA portfolios where they meet the MA eligibility criteria. The PRA has amended paragraph 2.12A of SS7/18 to clarify its expectations on contractual bounding.

2.16 Eight respondents considered the contractual bounding requirement too strict, proposing it should instead be ‘bounded’ rather than ‘contractually bounded’, or that junior notes from internal securitisations, or unrestructured portfolios of assets, should be allowed in MA portfolios subject to notional restructuring or cash flow haircuts.

2.17 The PRA considers that contractual bounding of cash flows is necessary to be confident that the MA can be earned in practice. Without contractual bounding, the PRA considers that the cash flow profile could be changed to a pattern unsuitable for the recognition upfront, as loss-absorbing capital resources, of the value of a proportion of the credit spread hoped to be earned over the lifetime of the investments. The PRA considered that this would compromise the PRA’s statutory objectives. The PRA further considers that contractual bounding, and the ability to assess the variability of cash flow payments, is necessary to allow the credit quality of the assets to be assessed through a credit rating or the undertaking’s internal credit assessment of a comparable standard. Having considered the responses, the PRA has decided not to change the draft policy.

Clarification of expectations in relation to fixed assets in MA portfolios

2.18 SS7/18 sets out a number of expectations around firms’ classification of certain assets as having fixed cash flows on account of specific treatments or assumptions, specifically in the following sections:

partial recognition of an asset’s cash flows;
redemption or termination clauses;
extension on default clauses;
cash flows with uncertain but bounded timing; and
sufficient compensation.

2.19 In paragraph 2.16 of CP19/23, the PRA proposed to further clarify the circumstances in which particular types of assets were likely to satisfy the PRA’s expectations of assets with fixed cash flows. These changes were reflected in SS7/18 in the addition of paragraphs 2.21A, 2.21B and 2.23A, a change to the heading above paragraph 2.23 and an additional condition in paragraph 2.21.

2.20 13 respondents said that the proposed changes in SS7/18 would lead to a re-evaluation of certain assets that are currently considered to have fixed cash flows. This might then lead to a material volume of existing MA portfolio assets needing to be reclassified as assets with HP cash flows, which would mean a significant reduction in the capacity that firms had to invest in assets with HP cash flows (noting the limit of 10% that applies to the MA benefit arising from assets with HP cash flows). One respondent added that any feedback from the PRA to the concerns raised about the potential reclassification of existing assets that involved exemptions for existing assets from the new expectations would present practical implementation challenges. Five respondents considered that, as a more general point, any changes to SS7/18 should not require firms to reassess the eligibility of existing MA portfolio assets.

2.21 On consideration of these responses, the PRA has decided not to carry forward the proposed changes listed above and has therefore deleted the relevant paragraphs from SS7/18 (2.21A, 2.21B and 2.23A), reversed the proposed change to the heading above paragraph 2.23 and removed the additional condition from paragraph 2.21. The PRA agrees with respondents that the draft policy should not act as a trigger for firms to reclassify assets previously considered to be ‘fixed’ assets.

2.22 Specifically, the PRA’s reasons for not carrying forward the proposed changes in SS7/18 listed in paragraph 2.19 above are:

the primary purpose of the proposed changes was to support the extension of MA eligibility criteria to include assets with HP cash flows (in particular to provide examples with regard to assets with issuer optionality);
the intent of the proposals was not to make changes that would reclassify assets currently in MA portfolios as being assets with HP cash flows. The PRA has considered that the proposals will result in assets in existing MA portfolios being in the ‘fixed’ component following implementation of these proposals as set out in paragraph 2.17 of CP19/23; and
the PRA recognises that the wording of paragraphs 2.21A and 2.23A potentially would have conflicted with the existing wording of paragraphs 2.20 and 2.21, and hence could have led firms to reclassify assets that satisfy the expectations in paragraphs 2.20 and 2.21 as assets with HP cash flows. This was not the PRA’s intention and was not reflected in the estimated costs of the PRA’s proposals in the CBA outlined in the CP. As noted above, responses to the consultation make it clear that a significant proportion of the 10% limit for MA benefit arising from HP cash flows (or any additional voluntary exposure limit which the firm has elected to apply) could be used up under such a reclassification.

2.23 Notwithstanding the points above, the PRA notes that, in their CP responses, and in bilateral engagement before and after CP publication, firms suggested there were a variety of interpretations of the aforementioned sections of SS7/18. Accordingly, the PRA would like to better understand firms’ approaches in this area to ensure that the PRA's expectations in relation to assets with fixed cash flows are being interpreted consistently, in particular to:

ensure consistent interpretations of the paragraphs in each of the relevant sections in SS7/18 (listed in paragraph 2.18 above) and that there is a level playing field between firms;
ensure firms are making appropriate assumptions in their treatment of assets, particularly in relation to the timing of cash flows and the rating of the assets; and
understand the extent to which there may be a build-up of concentrated exposures in individual firms, or systemic risk across the industry from changes in cash flows arising due to low probability events causing issuers to redeem or terminate assets without sufficient compensation being paid to replace foregone cash flows, or to defer cash flows.

2.24 However, the PRA considers that these matters should be addressed initially through bilateral supervisory engagement with firms, for example using case studies or considering how firms assess compliance with paragraph 2.22 of SS7/18, rather than as part of the implementing of the proposals that were subject to this consultation.

Perimeter between fixed assets and assets with HP cash flows

2.25 Paragraph 2.12D of SS7/18 (published as 2.12E following renumbering of SS7/18) noted that some assets could either be considered to have HP cash flows or could be considered to have fixed cash flows where firms apply the relevant expectations in Chapter 2 of SS7/18. Where those assets are considered to have HP cash flows, their inclusion in the MA portfolio would be contingent on having specific permission to do so (eg regarding best estimate cash flows, appropriate level of FS additions etc). In those cases, firms may have an option to treat those assets either as having HP cash flows, following their specific permission, or as having fixed cash flows, following the expectations of Chapter 2 of SS7/18.

2.26 Two respondents sought clarity, where there was such optionality, as to whether the PRA would place any restrictions on firms in terms of movements of MA portfolio assets between a fixed and HP cash flow treatment.

2.27 As noted above, treating assets as having HP cash flows would require having a specific permission to do so. In light of the responses received to clarify this issue, the PRA has made changes to paragraph 2.12D of SS7/18 (published as 2.12E following renumbering of SS7/18) to reflect the fact that the PRA considers that firms that have such a permission should be able to switch between a fixed and HP cash flow treatment on MA portfolio assets.

2.28 More specifically, the PRA is open to movements of assets between fixed and HP cash flow treatments where:

the firm has permission to apply the new treatment for a particular asset;
the firm manages the assets in line with the MA permission, including applying an FS addition (where moving from ‘fixed’ to HP) and considering the implications for the attestation;
changes in treatment are subject to the firm’s policies on managing the MA portfolio such that they are subject to an appropriate level of governance and oversight;
frequent changes in treatment for individual assets are subject to justification; and
firms carefully consider the operational implications from applying different treatments to holdings of the same asset.

Asset restructuring

2.29 The PRA proposed changes in paragraphs 2.52 to 2.61A of SS7/18 to both incorporate the inclusion in MA portfolios of assets with HP cash flows, and to set out additional or updated expectations for assets that firms have internally restructured.

2.30 One respondent considered the PRA’s expectation for extension clauses on internal restructures in paragraph 2.57A of SS7/18 to be arbitrary and unnecessary and suggested it should be removed.

2.31 The PRA has set out above in paragraph 2.21 that the proposed paragraph 2.23A of SS7/18 has been deleted. The PRA considers it appropriate that the expectations for such clauses are consistent for both internally restructured assets and assets that have not been internally restructured. The PRA has updated paragraph 2.57A of SS7/18 to reflect this.

2.32 Six respondents asked the PRA to clarify that the ability to restructure assets will not be curtailed by the proposed reforms. Two respondents disagreed with the PRA’s new expectation in paragraph 2.58 of SS7/18 that Special Purpose Vehicles (SPVs) set up for internal restructures should have sufficient assets to meet future funding commitments to complete an investment that will be used to secure cash flows on the notes issued. These respondents considered that other approaches such as ring-fencing would achieve the same outcome.

2.33 The PRA expects that firms will generally include MA eligible assets, whether with fixed or HP cash flows, in MA portfolios without restructuring. Where a firm restructures MA eligible assets, and then makes an application to include eligible note(s) from such a restructure in an MA portfolio, the PRA expects that the firm will additionally explain the reasons for the restructure. The PRA has updated paragraph 2.55B of SS7/18 to reflect this.

2.34 The PRA considers that to ensure the security of cash flows necessary to justify up-front recognition of a proportion of the spread hoped to be earned over the lifetime of the investments, SPVs should hold sufficient assets to meet the commitments that give rise to that MA. Such SPVs should be operated independently of the firm and on an arm’s-length basis. Accordingly, the PRA does not consider it appropriate that they rely on non-contractual funding arrangements. The PRA has therefore not made a change to this expectation.

2.35 Five respondents disagreed with the PRA’s proposed new expectation in paragraph 2.55B of SS7/18 that the aggregate value of a restructuring arrangement, including the MA benefit from the notes issued by the subsidiary company and the value of any residual interest in the company, would not exceed the value that would result from including the assets directly in the MA portfolio. One respondent asked the PRA to confirm if the new expectation should be considered as a limit akin to the effective value test (EVT) for equity release mortgages (ERMs).

2.36 The PRA considers that restructuring arrangements, for example a liquidity facility, may improve the certainty of cash flows and hence the credit rating. The PRA expects that activities to improve the security of a SPV and the rating of the notes would need to be on an arm’s-length basis, including the SPV paying an economic rate for the use of such facilities. Where firms consider that value has been created by restructuring, they should be able explain how this has arisen and why this is a reasonable outcome (and not, for example, from the use of a liquidity facility for which the SPV is paying below a market rate). The PRA confirms that the proposed new expectation is not intended to be considered as a limit, but as an indicator that value has been created by restructuring. Paragraph 2.55B of SS7/18 has been amended to reflect this.

Controls on the quality of matching

2.37 The PRA proposed an extension to the MA controls framework to allow for the inclusion of assets with HP cash flows and the introduction of two specific HP cash flow matching tests along with minor amendments to the existing tests.

2.38 Nine respondents considered that the proposed controls framework for assets with HP cash flows was too onerous and that individual controls could be removed or weakened without detriment to the PRA’s objectives. Two of these respondents and one further respondent considered that the costs associated with implementing and managing these controls are disproportionate for smaller firms and act as a barrier to entry for new firms. One respondent also asked the PRA to confirm that the additional matching tests do not apply to assets with fixed cash flows.

2.39 The PRA considers the proposed controls framework appropriate and sufficient to meet the additional risks introduced by the inclusion of assets with HP cash flows in MA portfolios. Furthermore, the PRA considers that each individual safeguard serves a particular purpose. The specific challenges to individual safeguards raised by respondents are covered below, including where the PRA has made changes to the operation or calibration of the safeguards.

2.40 The PRA has considered the responses regarding disproportionate costs for smaller and new firms. The PRA considers that weakening the controls framework for such firms would not advance its primary objectives, and the additional costs from the controls framework are likely to be proportionate given the additional cash flow variability risks posed by investing in assets with HP cash flows. The PRA confirms that, as set out in their specification, the additional matching tests do not apply to portfolios of assets with only fixed cash flows. Accordingly, the PRA has decided not to change the draft policy as a result of these responses.

Total exposures of additional risks

2.41 The PRA proposed making a rule to limit the amount of MA that can be derived from assets with HP cash flows; specifically, that a maximum of 10% of the MA benefit on an MA portfolio may be generated by assets with HP cash flows. Six respondents considered that the overall limit on HP cash flow exposure should be more risk sensitive, with one of these respondents considering that a buffer would be required below this limit to allow for changing market conditions.

2.42 Six respondents suggested that the MA benefit limit should be higher than 10%, or removed entirely, with one further respondent considering that the 10% limit provided insufficient capacity for smaller MA portfolios to invest in assets with HP cash flows. This respondent proposed that firms should be permitted to set their own exposure limit.

2.43 Having considered these responses, and possible alternative formulations of this rule, the PRA considers that the proposed formulation is an appropriate outcomes-based measure that ensures the vast majority of assets in MA portfolios have fixed cash flows and is straightforward for firms to implement and monitor. The PRA has therefore decided not to change the draft policy. The PRA considers that the disadvantages of more complex formulations of this rule, including in design, calibration and implementation, would outweigh the benefits of a more risk-sensitive measure. A more complex rule, for example one that considers only the MA contribution from cash flows that are uncertain, would be significantly more onerous for firms to implement as the MA would need to be considered at a cash flow level.

2.44 Further, having considered the responses, the PRA has decided to maintain the level of the proposed limit on MA benefit generated from assets with HP cash flows at 10% and no change has been made to the draft policy. The PRA sees the 10% limit as key to: ensuring that the vast majority of assets in MA portfolios continue to have fixed cash flows, in line with the Government’s November 2022 statement; limiting the overall risks to the quality of matching; and advancing the PRA’s objectives. A firm-specific limit would require advance knowledge of the types and volumes of optionality that are intended to be written by firms, requiring greater PRA scrutiny in assessing the quality of matching for individual firms, and potentially creating a barrier to new or smaller MA portfolios. Similarly, tailored rules for smaller MA portfolios would lead to differing levels of PRA scrutiny and would be difficult to implement without increasing the risks to the PRA’s objectives.

2.45 One respondent, while supportive of the 10% limit, considered that it should be reviewed periodically. The PRA has now published PS4/24 – PRA statement on the review of rules, setting out the PRA’s final statement on the review of rules. Any future review will consider the PRA’s objectives, which may be influenced by the types and pace of investment in assets with HP cash flows, the types of optionality that have been written and the risks to the quality of cash flow matching.

2.46 One respondent expressed support for the 10% limit, with the provision that assets considered fixed by treatment or assumption would not contribute towards the measure. In paragraph 2.21 of this PS the PRA has set out changes to the proposed content of SS7/18 and the PRA confirms that existing approved assets that meet the eligibility requirements will not be counted towards the 10% MA benefit limit while they follow a ‘fixed’ treatment. Having considered the responses, the PRA has decided not to change the draft policy.

2.47 Two respondents requested clarity on whether the 10% limit would be met where firms chose to recognise a lower amount of MA from assets with HP cash flows, for example if such assets generated more than 10% of the MA benefit but the firm applied a reduction to the MA benefit to meet the 10% limit. These respondents also asked the PRA to set out the actions available to firms in breach of this limit.

2.48 The PRA considers the 10% limit to apply to the amount of MA benefit being taken credit for, rather than that which could be generated by those assets, and that there are a range of options available to firms to manage the MA portfolio to meet this limit. The PRA also considers that the availability of such actions means that, even as market conditions change, firms will be able to use the whole 10% limit if they wish to. The PRA has considered the responses received and considers it would support its objectives to set out possible approaches for firms to manage MA portfolios against the 10% limit. Accordingly, the PRA has set out possible approaches firms may take in paragraph 8.4 of SS7/18. These include:

the reallocation, where firms have appropriate permissions, of assets within the components of the MA portfolio; or
the reallocation, where firms have appropriate permissions, of assets between the MA portfolio and non-MA portfolio; or
reducing the MA contribution from assets with HP cash flows by the application of further FS additions to assets with HP cash flows. However, the PRA considers that routinely applying further FS additions to remain within the 10% limit may be evidence of a failure of the firm’s risk management framework.

Additional matching tests

2.49 In CP19/23 the PRA proposed two additional matching tests and including an assessment of the risks from assets with HP cash flows in firms’ liquidity plans.

2.50 Five respondents considered that the thresholds on the additional matching tests for firms holding assets with HP cash flows were too low, with one of these respondents and one further respondent suggesting the widening of safeguards other than the 10% MA benefit limit, in order to provide more headroom for investment in assets with HP cash flows. Four of these five respondents proposed alternative test threshold calibrations of 5% or more, but none set out how these proposals were determined or how they advanced the PRA’s objectives. One respondent proposed alternative methodologies for carrying out a recalibration of the additional matching tests.

2.51 Having considered these responses, the PRA has decided to revisit the calibration of the additional matching tests. In the CBA baseline, the PRA considered that firms would be closely cash flow matched, and the proposed thresholds gave sufficient capacity to facilitate investment in assets with HP cash flows. Following the responses noted above, the PRA has recognised that firms are typically not as well cash flow matched as assumed in the baseline. Accordingly, the PRA agrees that the proposed calibrations of Matching Tests 4 and 5 may be too low. In particular, firms close to the threshold for Matching Test 1 might have little additional capacity to invest in assets with HP cash flows, without rebalancing their investment in assets with fixed cash flows. The PRA has therefore increased the thresholds for Matching Tests 4 and 5 from 3% to 5%. Notwithstanding that the types of additional cash flow variability that firms will accept from assets with HP cash flows is uncertain, the PRA considers that a 5% level for those tests (which was supported by some respondents) would advance the PRA’s objectives and will provide adequate headroom for investment in assets with HP cash flows, up to the 10% overall limit noted above.

2.52 Two respondents considered that the need to assume a ‘worst-case’ scenario for every asset in Matching Tests 4 and 5 was unduly onerous and, in not allowing for diversification between assets or groups of assets, may lead to implausible scenarios being considered. One further respondent requested clarification regarding the determination of a ‘worst-case’ scenario and suggested that low probability / high impact events should be excluded from the definition of ‘worst-case’. A further respondent considered that the contractual cash flow profile used in Matching Test 4 may not give rise to a scenario with the greatest reinvestment risk.

2.53 Allowing for diversification in the additional matching tests would be significantly more challenging for the PRA to calibrate and for firms to implement, and the PRA considers this would introduce a barrier to investment in assets with HP cash flows. As the proposed design recognised that diversification was not allowed for, the PRA considers that an alternative design would likely have a lower threshold, consistent with the increased risk sensitivity, in order to advance the PRA’s objectives. The PRA notes that there are limitations with the additional matching tests, but considers that introducing an allowance for diversification would lead to a more complex suite of controls that would not advance its objectives, and therefore has not changed the design of Matching Tests 4 and 5 in SS7/18.

2.54 The PRA recognises that under Matching Test 4, which assesses the lowest MA benefit of a specific asset using a cash flow profile permitted under the contract, the contractual cash flow that results in the minimised MA benefit may not be the cash flow profile that results in the greatest level of reinvestment risk. The PRA considers that in such cases firms should consider whether a further assessment of the quality of matching is required. The PRA has accordingly updated paragraph 4.10B of SS7/18.

2.55 One respondent suggested that component B assets should be included in Matching Test 5, as these may be used to meet policyholder liabilities.

2.56 The PRA notes that Matching Tests 1 and 4 have been constructed to consider component A assets only, as it is these assets that replicate the expected liability cash flows after being adjusted for the component of the FS that corresponds to the probability of default (PD). The PRA considers that Matching Test 5 should be based upon a similar design, as under stress it is not certain that component B assets will provide a close match for the resulting cash flow mismatches. For instance, component B assets may be longer dated than the mismatch observed under stress, and hence portfolio rebalancing may be required. The PRA does not consider that component B assets should be included in Matching Test 5 and has decided not to change the design of the additional matching test.

2.57 One respondent requested further clarity around the use of different reinvestment assumptions between Matching Tests 1, 4 and 5.

2.58 The PRA considers that the use of different reinvestment assumptions reflects the different objective of each matching test.

Matching Test 1 is an accumulated cash flow shortfall test with no explicit reinvestment assumption, though implicitly reinvestment is assumed to be at the prevailing risk-free rate. The PRA considers this is appropriate as the matching test is intended to assess the quality of matching in the current MA portfolio, and including an explicit reinvestment assumption in the matching test would be inconsistent with the MA benefit being earned by the MA portfolio held at a specific point in time.
Matching Test 4 is a scenario where assets with HP cash flows are received in a manner that minimises the MA benefit, consistent with the contractual terms. As this is a scenario, rather than an assessment of the quality of matching within the current MA portfolio, the PRA considers it is appropriate that firms may, optionally, apply a reinvestment assumption that is different to the implicit assumption of the risk-free rate. The PRA has clarified its expectations for determining the reinvestment assumption in paragraph 4.10A and Appendix 1 of SS7/18.
Matching Test 5 is a scenario showing the extent to which firms may be forced sellers of assets. In such a scenario, a reinvestment assumption should not be required as the firm is disposing of, rather than acquiring, assets.

FS additions for assets with HP cash flows

2.59 In paragraph 2.45 of CP19/23 the PRA proposed making rules that firms should identify all sources of uncertainty regarding the timing and amount of cash flows from any asset in the relevant portfolio with cash flows that are HP, and add to the FS an amount that reflects the risks arising from these uncertainties. The PRA also proposed expectations in paragraphs 5.17 – 5.29 of SS7/18 on how firms can demonstrate they are making an adequate allowance for these risks, including the use of standard approaches, for example where data on optionality risks is scarce.

2.60 The PRA received responses on the rationale for requiring additions to the FS for the additional risks posed by assets with HP cash flows. One respondent said that the PRA needed to provide a clearer link between the Government’s objectives and the proposals in the CP for additions to the FS for assets with HP cash flows.

2.61 The IRPR Regulations gives the PRA power to make rules to require firms to make an addition to the FS for assets without fixed cash flows. Paragraph 2.41 of CP19/23 set out that the FS should reflect all of the risks retained by firms, and assets with HP cash flows have additional risks beyond those on assets with fixed cash flows arising from the potential for cash flows to vary, including reinvestment and liquidity risks. The PRA’s overarching policy objective for the FS addition is for there to be an appropriate provision for the additional risks to matching from potential variations to cash flows. Any increase in credit spread that is compensation for cash flow variability should therefore not result in an increased MA benefit for firms. The PRA considers that the proposals set out in CP19/23 advance its statutory objectives and hence has decided not to change the draft policy.

2.62 Another respondent questioned whether the proposed FS addition is an appropriate tool for these additional non-credit risks, given that the FS (as currently calibrated) is defined for default and downgrade risks only.

2.63 The PRA notes that the existing FS design and calibration is focused on default and downgrade risks because the original (existing) MA specification requires fixed cash flows. Hence the wider risks on cash flow timing and amount are not routinely present and are not included in the design of the existing FS. The PRA considers that FS additions are a key element of the controls package for managing the additional risks from assets with HP cash flows in conjunction with the other proposed controls. These include the limit on the proportion of the MA resulting from assets with HP cash flows and the additional matching tests. Having considered the responses, the PRA has therefore decided not to change the draft policy.

2.64 One respondent commented that proposed Rule 8.1 of the Matching Adjustment Part of the PRA Rulebook requiring firms to identify ‘all sources of uncertainty regarding the timing and amount of cash flows’ should be amended to ‘all material sources of uncertainty’ to be consistent with the proportionality principle of the Solvency II regime.

2.65 Having considered this response, the PRA has decided not to change the rule as it considers that it is important that firms identify all sources of uncertainty in respect of timing and amount of cash flows that are relevant to the calibration of the FS addition. The materiality of these sources of uncertainty can be addressed in the calibration of the FS addition, which is a more appropriate approach for considering the proportionality principle.

2.66 One respondent suggested that an addition to the FS should not be needed where the credit spread on an asset does not include compensation for additional risks.

2.67 In the CP (paragraphs 2.43 and 2.44), the PRA noted that the increased uncertainty arising from HP rather than fixed cash flows may be compensated for in the asset pricing and hence the credit spread on the asset, and that it would not be appropriate for this to be used to increase Tier 1 capital. The requirements, however, for an MA portfolio differ from other investors in that greater certainty over the timing and amount of cash flows is required to match policyholder liabilities. Hence even in cases where the credit spread may not include compensation for these additional risks, the PRA considers that additions to the FS remain appropriate and has decided not to change the draft policy.

2.68 One respondent sought clarification on whether a more sophisticated approach to modelling the FS addition could be adopted from the outset. Another respondent considered the requirement to model a term structure for FS additions to be unduly onerous.

2.69 The PRA can confirm it is possible for firms to adopt a more sophisticated approach to modelling FS additions from the outset. However, where a firm uses a sophisticated approach for determining the FS addition for an asset, the PRA would not expect the firm’s MA application in respect of that asset to go through the streamlined MA application process. The PRA notes that a term structure is not required for the standard approach. For a sophisticated approach, the PRA set out in paragraph 5.21 of SS7/18 that it would expect firms to consider the appropriateness of applying a term structure even if the conclusion is that a term structure is not needed. The PRA has therefore decided not to change its draft policy.

2.70 One respondent proposed that FS additions should only be applied to cash flows at risk of variation.

2.71 In the proposed standard approach, firms would be expected to apply the additional FS to all cash flows. Having considered this response, the PRA is not proposing to change this expectation, noting that crystallisation of the variability could occur at any point in the asset’s lifetime. For example, where a ‘yield to worst’ approach assumes that the asset would be repaid at the earliest call date, the cash flow pattern of the asset is still at risk of changing if there is a sufficient change to the underlying economics. Hence there is a risk that the asset may need to be traded even during the non-call period to rebalance the MA portfolio and match the liability cash flows. In paragraphs 5.26 and 5.27 of SS7/18 the PRA set out considerations on how firms might determine an appropriate methodology for modelling the FS addition using a sophisticated approach. Having considered this response, the PRA has decided not to change the draft policy.

2.72 One respondent asked whether the PRA’s scrutiny of applications to include assets with HP cash flows would also cover the firm’s proposed derivation of the addition to the FS. If so, they questioned whether SS11/16 should be amended to reflect that the auditor does not need to assess the appropriateness of the methodology for that derivation but would instead need to assess the application of the approved methodology.

2.73 FS additions are required by Rule 8.2 of the Matching Adjustment Part of the PRA Rulebook for assets with HP cash flows and the PRA has published standard (mechanistic) methodologies that firms may apply to determine the FS addition. Alternatively, a firm may apply to use its own sophisticated approach to determine the FS addition. The PRA considers that the calculation of the MA, including the FS addition methodology, is not part of the scope of the MA eligibility conditions. Accordingly, the PRA considers that paragraphs 4.2H and 4.2I of SS11/16 that set out the scope of the PRA’s review of the MA should apply. Auditors are expected to form their own view on the calculation of the MA as part of their work to give an opinion as to whether the relevant elements of the Solvency and Financial Condition Report (SFCR) are prepared, in all material respects, in accordance with the PRA rules and Solvency II Regulations on which it is based. Having considered this response, the PRA has decided not to change the draft policy.

Standard Approach for FS additions

2.74 There was general support from respondents to the proposal for standard approaches for the FS additions, with two respondents noting that this should help avoid delays to investment. Four respondents welcomed the proposals for the standard approaches, considering them a practical, albeit prudent, basis on which to build, but requested worked examples for applying the different methods. In light of this, the PRA has set out worked examples in paragraph 5.24C of SS7/18.

2.75 One respondent queried whether it should be necessary to seek regulatory permission for an asset in the MA portfolio where the firm intends to use the standard approach for the FS addition.

2.76 The PRA notes that for new assets with HP cash flows, regulatory permission will be needed because the asset includes new features that would be the primary focus of the PRA review. Having considered this response, the PRA has decided not to change the draft policy.

2.77 Some respondents sought clarification on the allocation of an asset between the proposed ‘economic’ and ‘event’ risk standard approaches, in particular in cases where the risks could be considered both economic and event risk driven. Four respondents commented that firms should have a degree of flexibility in allocation to a standard FS addition approach. For example, a Residential Mortgage-Backed Security asset with underlying pooled asset exposures might technically fall under the definition for economic optionality, but in practice may have strong evidence of cash flow predictability more suited to using the standard approach for event risks. One respondent noted general support for the standard FS addition approaches, but also that the inclusion of ‘statistically predictable variability’ in the proposed expectation in paragraph 4.1D III of SS7/18 was an example of economic risk exposure that would lead to firms needing to apply an economic risk approach for the standard FS addition rather than the approach for event risks.

2.78 The PRA acknowledges that for such assets the risk profile in terms of changes to the expected cash flows may be less variable given it is driven by the repayment behaviour on the underlying mortgages and not the unfettered discretion of the issuer. Paragraph 5.22A of SS7/18 has therefore been added to note that for such assets where there is sufficient evidence of predictability, firms may propose to apply the standard approach to the FS addition for event risks. The PRA has also amended the examples of economic exposure set out in paragraph 4.1D of SS7/18. Firms can also use a sophisticated approach to modelling the FS addition, provided this allows for an appropriate range of stresses to the changes in repayment behaviour on the underlying assets.

2.79 Three respondents questioned whether an asset with both economic and event risk exposures would need to apply the sum of both standard approaches for the FS addition.

2.80 The PRA would not expect the sum of both approaches to be applied and has added the expectation in paragraph 5.22A of SS7/18 to state that in such cases a firm should follow the approach for the dominant risks.

2.81 The proposed standard approach for the FS addition specified a minimum 10bps allowance in normal market conditions for the potential costs of reinvestment or rebalancing of the MA portfolio resulting from changes to cash flows. Five respondents asked the PRA to clarify whether the 10bps is a floor, or an increase to the FS addition when calculated based on the asset cash flow variability risks.

2.82 The PRA expects the 10bps to act as a floor rather than a specific increase to the FS addition. The expectation in paragraph 5.20 of SS7/18 has been updated to clarify this.

2.83 The PRA received a number of varying responses on the calibration of the 10bps allowance for reinvestment or rebalancing costs. These included two responses that the 10bps is too onerous for longer duration assets or that the credit spread uplift for cash flow variability might be less than 10bps. Other general concerns raised about the calibration included:

it being too prudent where trading of assets would not be required, for example assets with upwards-only rent reviews (one response); and
the 10bps being too prudent as the respondent believed the PRA was trying to capture the costs of a forced sale whereas in practice a firm could elect not to trade.

2.84 Two respondents commented that the 10bps allowance was too prudent and not needed as the matching tests assume that assets are held to maturity. Two further respondents said that it was unclear what the 10bps allowance represents or how it should be interpreted. They sought further detail on the PRA’s analysis for the calibration to avoid the risk of firms double counting potential trading costs in the FS addition. One respondent also requested PRA guidance for the volume of trading data needed to depart from the 10bps allowance in the standard approach or when using a sophisticated approach to modelling the FS addition.

2.85 The PRA considers that for assets with cash flows that are fixed by the use of a specific treatment or assumption, for example where cash flows can only be revised upwards, no FS addition would be required. The PRA considers that for assets with HP cash flows, which require an addition to the FS to reflect the risks arising from the additional uncertainties, it is appropriate that rebalancing or reinvestments costs are provisioned for on the basis that the MA portfolio would need to trade (ie assets are not held to maturity). The risk profile of these assets differs from the management of a portfolio of assets with fixed cash flows where the MA portfolio is not exposed to the risk of cash flow variability and hence the need to trade, and therefore the PRA has decided not to change the draft policy as a result of these responses. The cash flow matching tests set tolerances on how compliance with the requirement to replicate liability cash flows can be demonstrated, whereas the FS addition is a provision for additional risks (from cash flow variability) which should not be included in the MA benefit and hence the PRA considers that the 10bps allowance for rebalancing or reinvestment costs is appropriate.

2.86 The PRA notes that the standard approach for FS additions has been designed to make sufficient allowance for the costs of trading or rebalancing the portfolio in the event of asset cash flows changing. In paragraph 2.48 of CP19/23 the PRA set out how the calibration has been determined. Having considered the responses, the PRA has decided not to change the draft policy. The standard approach is designed to be simple and represents an average; hence in some cases the 10bps may be too conservative and in others, such as for less liquid assets, it may be insufficient. In such cases, as paragraph 5.20 of SS7/18 notes, where there is sufficient evidence of relevant trading costs, the PRA is open to firms justifying a lower allowance. The PRA does not, however, consider it practical to give further guidance on the volume of trading data required as it expects this to vary according to the profile and management of individual MA portfolios.

2.87 One respondent commented that while many of the assets discussed at the Investment Flexibility SEG (IFSEG) could be included in an MA portfolio post-reform as assets with HP cash flows, they would need to consider whether they would be viable investments given the proposed approach for FS additions. The respondent also commented that if assets with extension clauses are reclassified as assets with HP cash flows, then the 10bps allowance should not be applied as otherwise there is a risk that firms will reassess these investments.

2.88 The PRA considers that the FS addition is to provision for additional risks to the MA portfolio. The standard approach is necessarily simple, and the calibration is designed to be an average. Hence it will not be fully reflective of cash flow variability risks for each asset individually. Where there is sufficient data, firms have the option to apply to use an adjusted standard approach or a sophisticated approach to model the FS addition. Having considered this response, the PRA has decided not to change the draft policy. The PRA considers that the changes to SS7/18 outlined in paragraph 2.21 earlier in this chapter will ensure that existing approved assets with extension clauses that meet the eligibility requirement shall be considered as having fixed cash flows and will not need to be reclassified as assets with HP cash flows.

2.89 One respondent proposed that attestation can be used as a safeguard for potential under-calibration of the proposed standard approach. The PRA agrees that attestation is a key safeguard; it is, however, important for firms to note that the calibration of the standard approach should, where appropriate, be adjusted to reflect the risks to the quality of cash flow matching. For example, for assets with event risk exposure, the standard approach expectation is that the proportion of the difference between worst and best estimate MA taken as a provision would not generally be expected to be less than one quarter. Where a higher proportion is more appropriate, this should be included in FS addition under Rule 8.2 of the Matching Adjustment Part of the PRA Rulebook rather than as a voluntary addition from attestation. The case for adjustment to the main FS addition would similarly apply to the allowance for reinvestment or rebalancing costs where firms expect to incur higher trading costs than the 10bps set out in the standard approach.

2.90 The proposed standard approach for assets with economic risk exposure set out in paragraph 5.23 of SS7/18 is for the firm to project cash flows on a ‘yield to worst’ basis. Four respondents stated that ‘yield to worst’ may not be appropriate for some assets, as this would imply the shortest term and hence no MA. One respondent asked the PRA to clarify how ‘yield to worst’ should be applied in conjunction with partial recognition of cash flows.

2.91 The PRA considers that where a ‘yield to worst’ cash flow projection results in a reduction or elimination of MA benefit, this is indicating that the MA benefit cannot be sufficiently relied on and therefore the capitalised value should not be included as Tier 1 capital. The PRA can confirm that where a ‘yield to worst’ cash flow projection is used, no further reduction, ie partial recognition of cash flows, needs to be applied on top of the FS addition. Similarly, where a firm recognises only the part of an asset’s cash flows that can be considered certain or fixed, there is no need in addition to apply a ‘yield to worst’ cash flow projection or FS addition as this would not be considered to be an asset with HP cash flows.

2.92 Two respondents asked the PRA to consider allowing firms to apply alternatives to a ‘yield to worst’ approach for the standard approach, for example a ‘spread to worst’ approach to projecting cash flows.

2.93 Having considered this possible alternative, the PRA has decided to keep the ‘yield to worst’ approach as the standard methodology. The PRA notes that should the features or contractual terms of an asset mean that an alternative approach would be more appropriate, firms may ask the PRA to consider this on a case-by-case basis, provided that the approach retains the assumption that the issuer will behave in an economically rational manner. The PRA has updated the expectation in paragraph 5.23 of SS7/18 to reflect this.

2.94 For assets with HP cash flows with event (non-economic) risk exposure, the PRA proposed in paragraph 5.24 of SS7/18 that the FS addition should provision for a proportion, not generally expected to be less than one quarter, of the difference between the worst-case MA benefit (typically subject to a minimum of zero) and the MA benefit resulting from cash flows projected on a best estimate basis. One respondent noted general support for the proposed approach for determining the FS addition on a spectrum between best estimate and worst-case outcomes.

2.95 There were several responses on the appropriateness of the one quarter (25%) proportion of the difference between worst and best estimate MA for more remote risks. Two respondents commented that the 25% calibration was too high when combined with a high bar of moving to a sophisticated approach for the FS addition. Four respondents considered that for assets with event risks, firms should be allowed to deviate from the standard approach of provisioning 25% of the difference in MA between the best estimate and worst-case cash flow profiles. This includes events that firms consider to be tail risks, where four responses proposed that a separate category in the standard approach should be created with a flat FS addition of a low number of bps, or even that no FS addition should be required and any residual risk from a change in cash flow profile should be covered by the SCR. Two respondents asked for firms to be able to include in their MA portfolio assets with tail event risk exposure without seeking prior permission, including for the FS addition.

2.96 Having considered the responses, the PRA has decided not to change the draft policy to create a separate category for remote event risks in the standard approach, or to have a category of assets with HP cash flows for which no FS addition is required as a standard approach, or for which prior PRA permission is not required. The proposed calibration of a provision of 25% of the additional MA above the lower bound was predicated on a mix of more and less variability within the assets, and paragraph 2.53 of the CP acknowledged that a standard approach, which is designed to be simple, is unlikely to be suitable for all risks. Paragraph 2.53 of the CP did, however, acknowledge that there may be some variation in the proportion of the additional MA benefit retained by firms given the range of sources of cash flow variability, and that where a firm has credible data, it may be able to justify retaining a higher proportion for particularly remote risks. The expectation on FS additions in paragraph 5.24 of SS7/18 has been updated to reflect this.

2.97 Two respondents proposed that when calculating the ‘worst MA’ outcome required for the standard approach for event risks, and where this results from earlier repayment compared to the best estimate cash flow pattern, firms should be allowed to assume a prudent reinvestment spread over the risk-free rate, similar to the approach for Matching Test 4.

2.98 Having considered these responses, the PRA accepts that there is a case for aligning the assumed reinvestment rates applied in the FS addition with those used for Matching Test 4 and has added paragraph 5.24A to SS7/18 to accommodate this. In practice, the PRA’s view is that for many assets this will only result in a small improvement to the ‘worst MA’ outcome given that if a credit spread over the risk-free rate is assumed then firms would also need to deduct the relevant FS from this (prudent) spread.

2.99 One respondent commented that a methodology for FS additions based on a proportion of the difference between worst and best estimate would introduce volatility and be pro-cyclical.

2.100 The PRA acknowledges that this matter was discussed in the IFSEG. However, the expectation in paragraph 5.24 of SS7/18 notes that firms can express the FS addition as a number of bps at the point of investment while ensuring that they have a framework for assessing whether the fixed allowance remains adequate as conditions change. The PRA also considers that, as a maximum 10% of the MA benefit is derived from assets with HP cash flows, it is unlikely that this standard approach for the FS addition would introduce material volatility and hence pro-cyclicality, and therefore it has decided to make no change to the draft policy.

2.101 Two respondents commented that the PRA had given limited justification for the proposed 85^th percentile target of the MA loss distribution for firms using a sophisticated approach for event risks, and that the 75^th percentile, in line with the International Financial Reporting Standards 17 (IFRS 17) median assumption for risk adjustment, would be more appropriate and strike a better balance between the PRA’s primary and secondary objectives. One respondent added that the 25% in the standard approach calibration is prudent, but that this prudence should not be carried across into the sophisticated approach.

2.102 Having considered these responses, the PRA has decided not to change the draft policy, and to retain the expectation of an FS addition calibration at the 85^th percentile. The PRA considers that 25% of the difference in the standard approach and the 85^th percentile in the sophisticated approach are both consistent with a risk appetite that firms should be able to earn the MA with a high degree of confidence (HDC). The IFRS 17 median represents a different risk appetite and the same target percentile should not necessarily be applied. The PRA also notes that a 75^th percentile in a heavier-tailed distribution could in practice mean a firm retaining up to 90% of the additional spread over worst MA outcome even in cases where there is little or no data, which the PRA does not consider to be consistent with advancing its primary objectives.

2.103 One respondent noted the proposed expectation for FS additions to be held in component B, commenting that component A would be more technically correct, while accepting the PRA’s proposed approach as practical.

2.104 The PRA agrees that including the FS addition in component A may be more technically correct and is open to firms adopting an approach of holding the FS addition in either component A or component B. Paragraph 5.30 of SS7/18 has been updated accordingly.

Determination of best estimate cash flows for assets with HP cash flows

2.105 The PRA set out rules and expectations for the determination of best estimate cash flows for assets with HP cash flows in Rule 5.4 of the Matching Adjustment Part of the PRA Rulebook of CP19/23 and in paragraphs 4.1A to 4.1M of SS7/18.

2.106 Nine respondents supported the PRA’s proposal in paragraph 4.1I of SS7/18 which stated that firms may adopt a deterministic approach to determining best estimate cash flows. Two respondents asked for clarity over whether firms will have sole discretion over which best estimate cash flow methodologies to apply. One respondent said that a certainty equivalent approach should also be a permitted deterministic approach (in addition to the median best estimate outcome stated in paragraph 4.1I of SS7/18).

2.107 The PRA is not requiring a particular methodology and hence firms should consider carefully, and be able to justify, the methodology chosen. The PRA may review and challenge the methodology used. The PRA is open to considering alternative approaches to that set out in paragraph 4.1I but would expect such approaches to be justified. Having considered the responses, the PRA has decided not to change the draft policy.

2.108 The PRA expects the cash flow profile of best estimate cash flows to be consistent with that used for fair valuation under IFRS 17 as set out in paragraph 4.1M of SS7/18. Two respondents raised concerns over what would happen if this expectation was met but the PRA challenged a firm’s best estimate cash flow profile.

2.109 The PRA recognises this concern and has amended the wording in paragraph 4.1M of SS7/18 such that the expectation is that in most circumstances there would be consistency, with scope for firms to justify deviations from this.

Modelling assets with HP cash flows under stress

2.110 In paragraph 2.70 of CP19/23 the PRA proposed that firms should carefully consider how the FS addition may change under stress and that the FS addition may need to be updated, potentially materially, to reflect changes in likelihood of early repayment and potential loss of future MA benefit.

2.111 Three respondents commented that the PRA appeared to be setting an expectation for a significant impact (increase) to FS additions being needed under stress. Another respondent requested guidance on the process for reviewing the SCR.

2.112 The PRA confirms that it is not setting an expectation for the FS addition to automatically materially increase in stress. Paragraph 2.69 of CP19/23 noted that for statistically modelled best estimate cash flows, the projection may need to be updated to reflect stressed conditions. Where a deterministic approach has been adopted, the PRA’s expectations are set out in paragraphs 4.40-4.42 of SS8/18, that firms will need to carefully consider whether the FS addition needs to be changed in stress and how the stress affects the risks of cash flow variation. Having considered these responses, the PRA has decided not to change the draft policy.

2.113 Having reviewed these responses, the PRA confirms that there is no change in the process for assessing ongoing adequacy of the internal model, and the PRA has decided not to change the draft policy. Firms will need to consider the materiality of the additional risks arising from assets with HP cash flows and whether these are adequately allowed for in the internal model and develop plans to update the model as needed.

3: Liability eligibility

3.1 This chapter provides feedback to responses on Chapter 3 (Liability eligibility) of CP19/23. It also contains the PRA’s final policy, as follows:

amendments to the Glossary Part of the PRA Rulebook (Appendix 1);
amendments to the Matching Adjustment Part of the PRA Rulebook (Appendix 1); and
updates to SS7/18 – Solvency II: Matching adjustment (Appendix 3).

3.2 In Chapter 3 of CP19/23 the PRA proposed to:

restate the existing MA liability eligibility conditions from The Solvency 2 Regulations 2015 into the PRA Rulebook (with certain changes as described in the CP);
expand the underwriting risks permitted in MA portfolios to include recovery time risk; and
permit the inclusion of the guaranteed benefits component of with-profits annuities in MA portfolios. These components should be legally established and identifiable as guaranteed within an insurance contract and otherwise meet the MA eligibility requirements. Under this proposal, the non-guaranteed element would be retained outside the MA portfolio.

3.3 The PRA received 15 responses to Chapter 3 of CP19/23. Responses generally welcomed the expansion to MA liability eligibility criteria and noted that the newly eligible liabilities were suitable for inclusion in MA portfolios. Several respondents commented that the size of the newly eligible pool of liabilities was unlikely to be significant relative to the overall size of existing MA portfolios across the industry. Many also commented that the proposed PRA rules would lead to only individual in-payment income protection policies being eligible for inclusion into MA portfolios, leaving group policies ineligible. Some respondents also sought clarification relating to aspects of the expansion of the eligibility criteria and suggested they should be further extended to include additional liabilities.

Changes to draft policy

3.4 Following consideration of the respondents’ comments, the PRA has made changes to the PRA Rulebook to allow the inclusion of in-payment GDAs in MA portfolios, and to ensure that the policy intent of allowing in-payment individual and group income protection liabilities in MA portfolios was achieved. A summary of these changes is set out below:

amending Rule 2.3 of the Matching Adjustment Part of the PRA Rulebook to create an exception to MA liability eligibility conditions permitting the inclusion of both in-payment income protection liabilities and GDAs in MA portfolios; and
adding a new Rule 2.5 of the Matching Adjustment Part of the PRA Rulebook permitting the inclusion of both in-payment income protection liabilities and GDAs in MA portfolios.

3.5 Additionally, the PRA has made the following changes to its policy materials:

updating the wording of paragraph 3.5B of SS7/18 to refer to PRA expectations in respect of the updated liability eligibility conditions relating to income protection policies; and
adding a new section (paragraph 3.5C) to SS7/18 to refer to PRA expectations in respect of the updated liability eligibility conditions relating to GDAs.

3.6 Any significant changes to the CBA or ‘have regards’ analysis arising from changes to the draft policy proposed in CP19/23 are considered in Chapter 10 (Cost benefit analysis) and Chapter 11 (Have regards analysis) of this PS.

Feedback to responses

3.7 The PRA has considered the responses received to Chapter 3 of CP19/23. This section sets out the PRA’s feedback to those responses and its final decisions.

3.8 The sections below have been structured along the same lines as the sections of Chapter 3 of CP19/23. The responses have been grouped together as follows:

restatement of existing MA liability conditions; and
expanded liability eligibility conditions.

Restatement of existing MA liability conditions

3.9 The PRA proposed that the existing MA liability conditions that are not included directly in the IRPR Regulations would be restated in the PRA Rulebook, with amendments where necessary (including to update references to the Solvency II Directive). These restatements covered mortality risk limits, an exclusion from MA portfolios of liabilities with surrender values in excess of the value of assets covering the liabilities, and an exclusion from MA portfolios of liabilities that assume future premium payments.

3.10 Nine respondents commented on the retention of the existing exclusion of liabilities that assume future premiums. The respondents noted that it is increasingly becoming market practice for ‘Bulk Purchase Annuity’ (BPA) negotiations to consider the potential deferral of a portion of the premium payment, given the potential need of pension trustees to liquidate holdings in illiquid assets to fund the premium payment. Some of the respondents suggested that the PRA’s proposed exception for guaranteed with-profit annuities from this condition could be further extended (although the suggestions varied with regard to the extent to which liabilities with future premiums should be permitted in MA portfolios).

3.11 Having reviewed the responses, the PRA has decided not to change the draft policy. As noted in CP19/23, the PRA considers liabilities that assume future premium payments to be unsuitable for inclusion in an MA portfolio, as the MA should only be available where the portfolio already holds sufficient assets to meet the liability cash flows. Notwithstanding this exclusion, the PRA notes that in some cases an agreed partial deferral of the payment of a premium may not constitute a future premium on a liability. Depending on the contractual terms agreed, as well as a firm’s risk control framework relating to the partial premium deferral and the risk to the MA portfolio (eg if deferred premium payment is not received), a partial deferral of a premium payment may not necessarily cause a liability to be ineligible for inclusion in MA portfolios.

3.12 Five respondents expressed concern that the proposed restatement, in the PRA Rulebook, of the MA liability eligibility condition prohibiting the splitting of liabilities into different parts between the MA portfolio and the non-MA portfolio, would inappropriately limit the liabilities that could be included in MA portfolios. The respondents’ reasons for this concern varied but included the views that the PRA’s proposals would be:

allowing policies with ineligible components to be notionally (rather than legally) split from liabilities included in MA portfolios;
permitting liabilities with future premiums to be held in MA portfolios (where the portion of liability relating to the future premium would be notionally split apart from the liability in the MA portfolio); and
providing a less complex approach where liabilities in an MA portfolio have been partly reinsured.

3.13 Having considered the responses, the PRA has decided not to change the draft policy to remove the general prohibition on the splitting of liabilities into different parts between the MA portfolio and the non-MA portfolio. The PRA considers that removing this prohibition in MA portfolios would present a risk to the separate management of the MA portfolio, as required by Regulation 4(6) of the IRPR Regulations.

3.14 Four respondents suggested that the MA liability eligibility criteria should be further extended to also permit liabilities with less predictable cash flows into MA portfolios. Examples cited included periodic payment orders (PPOs), additional pension member benefits that may be considered under BPA contracts and also a general mirroring of the proposed additional flexibility relating to assets in MA portfolios (ie assets with HP cash flows, as discussed in Chapter 2 of this PS).

3.15 The PRA has considered the responses, and decided not to change the draft policy, beyond the changes described elsewhere in this chapter. The PRA considers that in other cases the liability eligibility conditions remain appropriate for MA portfolios, noting the justification provided in CP19/23 for the restatement of liability eligibility criteria. As described in the MA SoP, the PRA will consider proposals for the inclusion of liabilities in MA portfolios on a case-by-case basis. This may include where appropriate (ie where adequate justification is provided when considering the liability eligibility criteria), the examples cited above by these respondents, for example PPOs.

3.16 One respondent requested additional clarity relating to the PRA’s approach to surrender options on liabilities in MA portfolios.

3.17 Having considered the responses, the PRA has decided not to change the draft policy, as the PRA’s approach to surrender options on liabilities in MA portfolios is already described in existing policy materials. Paragraphs 3.8 to 3.13 of SS7/18 describe PRA expectations regarding surrender options.

Expanded liability eligibility conditions

3.18 In CP19/23, the PRA set out amendments to the MA liability eligibility conditions to widen the scope of eligible liabilities to those that are capable of being sufficiently well-matched to support the assumptions underlying the MA (described in Chapter 7). In particular, the PRA proposed that the permitted underwriting risks should be extended to include recovery time risk, in order to permit in-payment income protection claims to be included in MA portfolios. Amendments were also proposed to permit the guaranteed element of a with-profits annuity, where the fixed elements may be clearly identified and are not dependent on future premiums, to be included in MA portfolios.

3.19 Nine respondents welcomed the proposals to extend the types of liabilities that could be included within MA portfolios. Seven respondents commented that the stated PRA policy intent, of permitting in-payment income protection liabilities in MA portfolios by extending the list of permitted underwriting risks, could be partially impacted by the proposed restatement of other liability eligibility criteria. Respondents expressed concern that group income protection policies would be excluded, as a single group income protection policy could cover a number of employees working for an employer. Where claims were in payment for a subset of the covered employees under a policy, premiums would continue to be paid under this insurance contract for the other employees that were not in claim. This feature of group income protection business would not comply with the restated liability eligibility criteria that prohibit future premiums and the splitting of MA liabilities into different parts. Two respondents further noted that some income protection policies may provide specific benefits that would not satisfy the proposed MA liability eligibility conditions (for example a lump sum payment upon the diagnosis of a specified illness).

3.20 Having considered the responses, the PRA has updated Rule 2.3(1)(b) of the Matching Adjustment Part of the PRA Rulebook to permit the splitting of income protection policies into different parts, such that only parts relating to in-payment claims are included in MA portfolios. This update achieves the previously proposed policy intent to permit the inclusion of in-payment income protection liabilities in MA portfolios. Further, the existence of future premiums on a group income protection policy would no longer exclude the in-payment part of that policy from an MA portfolio, under new Rule 2.5 of the Matching Adjustment Part of the PRA Rulebook. The PRA has also updated paragraph 3.5B of SS7/18 to set out relevant expectations that such parts are separately identifiable and should not require payment of future premiums while in claim.

3.21 Two respondents suggested that in-payment ‘group death in service dependants annuities’, or GDAs, would be suitable liabilities for MA portfolios. These respondents noted that, once a claim is in payment, it would have features similar to other annuities that are MA eligible. Both respondents noted that these liabilities would not be permitted in MA portfolios under the proposed rules due to restriction on future premiums, and the inability of a firm to notionally separate the in-payment claims from the remainder of the contract. Where a dependant annuity is in-payment, it may form part of a group policy, for which premiums would still be payable.

3.22 The PRA agrees that GDAs, which are in-payment, can have features similar to other in-payment annuities that are eligible for inclusion in MA portfolios. Therefore, these liabilities could be suitable for inclusion in MA portfolios. To enable inclusion of in-payment GDA liabilities in MA portfolios, the PRA has updated Rule 2.3 (1)(b) of the Matching Adjustment Part of the PRA Rulebook to permit an exception from the requirement that liabilities within an MA portfolio must not be split into different parts. Further, the existence of future premiums for a GDA policy would no longer exclude the in-payment part of that policy from an MA portfolio, under new Rule 2.5 of the Matching Adjustment Part of the PRA Rulebook. Paragraph 3.5C of SS7/18 has also been updated to set out PRA expectations that where in-payment GDAs are included in MA portfolios, the in-payment GDA liabilities should be separately identifiable and future premiums should not be payable in respect of those liabilities while in claim.

3.23 Three respondents sought clarification on aspects of the PRA policy materials relating to the extension of the MA liability eligibility conditions to permit the inclusion of guaranteed elements of with-profit annuities in MA portfolios. Two respondents sought clarification regarding the allocation of future bonus declarations on with-profits annuities between the MA and non-MA portfolios. One respondent sought clarification on whether the non-guaranteed portion of with-profit annuities would need to be held outside of MA portfolios.

3.24 Having considered the responses, the PRA has decided not to change the draft policy. In paragraph 3.5A of SS7/18, the PRA presents an expectation that firms set out a clear policy regarding the addition of future attaching bonuses on with-profits annuities in the MA portfolio or elsewhere. This expectation does not refer to a specific split of a future attaching bonus between an MA portfolio and a non-MA portfolio, as this would depend on the circumstances of a firm. The same paragraph of SS7/18 also notes an expectation that firms demonstrate that only contractually guaranteed elements of with-profits annuities are included in MA portfolios. To achieve this, firms also need to demonstrate that the non-guaranteed portion of with-profits annuities are held outside of MA portfolios.

4: Credit ratings under the MA

4.1 This chapter provides feedback to responses relating to the proposals in Chapter 4 (Credit ratings under the MA) of CP19/23. It also contains the PRA’s final policy, as follows:

amendments to the Technical Provisions and Matching Adjustment Parts of the PRA Rulebook (Appendix 1);
updates to SS7/18 – Solvency II: Matching adjustment (Appendix 3);
updates to SS8/18 – Solvency II: Internal models – modelling of the matching adjustment (Appendix 4); and
updates to SS3/17 – Solvency II: Illiquid unrated assets (Appendix 5).

4.2 In Chapter 4 of CP19/23 the PRA proposed to:

remove the requirement in Rule 7.2(3) of the Technical Provisions Part of the PRA Rulebook for firms to apply the SIG MA cap;
remove expectations in respect of how the SIG MA cap is reflected in the calculation of the MA and in firms’ internal model methodologies;
introduce new expectations in relation to the prudent management of assets backing policyholder liabilities, specifically investment in SIG assets in the MA portfolio;
introduce new expectations around the assessment of the appropriateness of firms’ internal models for modelling the MA in respect of SIG assets;
with slight modification, substitute existing PRA expectations with requirements that internal credit assessments of assets in the assigned portfolio, as referred to in the IRPR Regulations, would have to satisfy; and
update expectations on the use of internal credit assessments to reflect the substituted requirements above and the PRA’s current supervisory approach.

4.3 The PRA received 19 responses to Chapter 4 of CP19/23. On the SIG MA cap, respondents generally welcomed the proposals to remove the cap from PRA policy and rules, citing benefits including increased opportunities to support productive projects and promotion of investment in ‘green’ assets. However, some respondents questioned whether all the new expectations were necessary and requested further clarification on points specific to internal models. On internal credit assessments, responses were also generally supportive of the proposals. However, respondents also made a number of observations and requests for further clarification or additional guidance that are addressed in this chapter.

Changes to draft policy

4.4 Following consideration of the respondents’ comments, the PRA has made the following changes to its policy:

updates to paragraph 7.13A of SS7/18 to reflect a wider range of metrics that firms could consider when determining the contribution/importance of SIG assets to their MA portfolios;
updates to paragraph 4.29 of SS8/18 to be clear that firms should ensure the MA on SIG assets is appropriate post-stress, including for those assets that downgrade to SIG as a result of the stress; and
updates to paragraph 2.8J of SS3/17 to clarify the PRA’s expectations for the individual with responsibility for the internal credit assessment function.

4.5 There are no significant changes to the CBA or ‘have regards’ analysis arising from the changes to the draft policy proposed in CP19/23 in respect of this PS chapter.

Feedback to responses

4.6 The PRA has considered the responses received to Chapter 4 of CP19/23. This section sets out the PRA’s feedback to those responses and its final decisions.

4.7 The sections below have been structured broadly along the same lines as Chapter 4 of CP19/23, with some areas rearranged to better respond to related issues. The responses have been grouped as follows:

investment in SIG assets;
modelling of the FS for SIG assets in internal models;
proportionality of requirements and expectations for internal credit assessments;
assurance on internal credit assessments;
responsibility for the internal credit assessment function; and
other responses.

Investment in SIG assets

4.8 The PRA noted in CP19/23 that the proposed changes to the PRA Rulebook to remove the SIG MA cap would remove a disincentive for firms to invest in SIG assets, and as a result firms may choose to invest more in these assets, or assets close to SIG, within their MA portfolios. The PRA therefore proposed to introduce several expectations including that any investment in SIG assets should be kept at prudent levels.

4.9 Many respondents were supportive of the PRA’s proposed new expectations with specific acknowledgement that SIG assets would need to be managed according to their potentially higher risk profile relative to assets rated investment grade. However, two respondents considered the PRA's proposed expectation that investment in SIG assets be kept to ‘prudent levels’ was unnecessary, given existing PPP requirements. One of the respondents considered all the proposed expectations around investment in SIG assets were unnecessary given they are already dealt with in firms' existing risk management, investment and modelling frameworks.

4.10 Having considered these responses, the PRA has decided to retain these expectations, which are intended to make points that are currently within the existing PRA rules more explicit in the context of SIG assets. The intention of this is to improve transparency, making the PRA’s expectations clear to all firms and improving consistency of approach. There is no intention to change existing policy, increase requirements on firms, diminish the importance of firms’ own risk management or interfere with existing processes that firms have in place.

4.11 One respondent suggested that while average default rates have historically been materially higher for SIG assets (compared to investment grade) this did have some benefits by reducing the sampling risk associated with default rates. The respondent asked that the PRA acknowledges this when supervising firms, as it may justify a modestly higher allocation to SIG assets than might initially appear to be reasonable.

4.12 The PRA does not consider it necessary to specifically refer to this point in SS7/18 and has therefore decided not to change the draft policy. The PRA acknowledges the statistical argument that is being made and recognises that it is one of many factors that firms may consider when setting their risk appetite and limits for investment in SIG assets. As set out in paragraph 7.13A of SS7/18, the PRA expects firms to justify their level of investment in SIG assets and has cautioned against high levels of investment given that these assets do exhibit higher expected default rates and may introduce material uncertainty around adequate cash flow matching.

4.13 One respondent proposed that it would be beneficial if the PRA were to highlight explicitly in SS7/18 and/or SS8/18 that when monitoring the SIG assets in their MA portfolios, firms should consider the contributions made by such assets to:

PD-adjusted cash flow;
total monetary value of MA; and
total monetary value of FS.

4.14 The respondent suggested each of these measures is likely to be more informative in an MA portfolio context than measures based on percentage of asset market value (MV).

4.15 The PRA considers that there are merits to the metrics suggested by the respondent and that they can give further insights to the extent to which SIG assets are contributing to the MA portfolio. Paragraph 7.13A of SS7/18 has therefore been updated beyond the changes set out in CP19/23 to include these alternative metrics as points firms may wish to consider in this context.

Modelling the FS for SIG assets in internal models

4.16 In CP19/23, the PRA acknowledged that historically some firms may have placed less focus on the calibration of internal model stresses for SIG assets due to the existence of the SIG MA cap in stress. To help ensure that internal models capture all material quantifiable risks to which firms are exposed, the PRA proposed to introduce an expectation that regardless of whether a firm includes a SIG MA cap in its model, the firm should be able to demonstrate that the internal model adequately reflects the risk profile for SIG assets. In addition, the PRA proposed specific factors that should be considered in this context including concentrations of SIG exposure.

4.17 Respondents generally appreciated the rationale for the PRA’s proposed expectations for the modelling of SIG assets in internal models. One respondent noted that firms should ensure their modelling of the FS under stress for SIG assets is robust and another respondent stated that they supported the PRA’s proposals to ensure internal models adequately reflect the risk profile of SIG assets and are calibrated accordingly.

4.18 One respondent noted that paragraph 4.29 of SS8/18 considers the impact of the SIG MA cap on SIG assets in stress. The respondent considered the SIG MA cap could be a material component of the stress on assets rated BBB (and to a lesser extent, assets rated A) where these assets downgrade to SIG in stress and pick up the higher FS, increased to allow for the cap where appropriate. The respondent suggested the PRA may wish to clarify that paragraph 4.29 relates to assets rated SIG post-stress and not assets that are rated SIG in base.

4.19 The PRA has updated paragraph 4.29 of SS8/18 to be clear that firms should ensure the MA on SIG assets is appropriate post-stress, including for those assets that downgrade to SIG as a result of the stress. The key point the PRA would emphasise is that internal models should result in appropriate capital being held given the firm’s risk profile and that this should remain the case if the SIG MA cap is removed.

4.20 In support of the concerns outlined by the PRA around the appropriateness of internal models for SIG assets, one respondent stated that it was vital that in removing the SIG MA cap in base, an uneven playing field was not unintentionally created in stress via mechanistic removal of the SIG MA cap from internal models. Another respondent also noted the PRA’s proposals but requested the PRA engage with them bilaterally on expectations regarding the calibration of SIG assets.

4.21 The PRA notes the points made and considers they are aligned to the points in CP19/23 requiring internal models to be appropriately calibrated for SIG assets before any removal of the cap should take place. This will be particularly pertinent for firms where the SIG MA cap has a material impact on the SCR. The PRA considers that these points are already adequately reflected in the draft policy and that no further changes are required. The PRA would welcome further discussions with firms about the calibration of internal models for SIG assets and encourages firms to reach out to their usual supervisory contact as appropriate.

4.22 One respondent stated that a firm’s internal model may not meet the internal model calibration standards following the removal of the SIG MA cap and agreed that artificially reincorporating some, or all, of the impact of the SIG MA cap may be a reasonable temporary measure. However, they considered that any such adjustment should be regarded as a Model Limitation Adjustment (MLA) within the meaning of CP12/23^{footnote [16]} and subject to a remediation plan. Another respondent argued on similar lines that the temporary retention of the SIG MA cap should be treated as a capital add-on.

4.23 The PRA has decided not to update the draft policy and rules to set out stronger or more prescriptive requirements and/or expectations in this area. The PRA considers it is for firms to determine if they should retain the SIG MA cap in their internal models, including whether any such retention should be temporary or permanent and the way in which the cap is incorporated into their internal models. The PRA would expect such considerations to take account of the firm’s investment strategy and the extent to which the MA portfolio could see downgrades to SIG in stress. The PRA considers that a firm’s decision to retain the SIG MA cap could either be temporary, while the firm enhances its modelling, or could be more permanent due to data issues and uncertainty surrounding SIG assets as well as proportionality and materiality considerations. The PRA considers retention of the SIG MA cap to be one approach by which firms may be able to demonstrate compliance with the internal model calibration standards, but notes there are also other ways that firms may seek to do this. If a firm’s retention of the SIG MA cap is not intended to be permanent, the PRA considers it is for the firm to determine the best way to retain the cap as a temporary measure (eg as an adjustment to the SCR or as a more integral part of the model).

4.24 One respondent expressed a concern that the PRA should not place unnecessary obstacles in the way of removal of the SIG MA cap in stress, and that the stressed FS should be allowed to reduce for SIG assets if this is consistent with the risk profile of the firm and/or other aspects of the internal model.

4.25 The PRA notes the points made but has decided not to change the draft policy, as it already acknowledges that firms may seek to remove the SIG MA cap from their internal models. The PRA considers its expectations help ensure that internal models continue to meet the internal model requirements and calibration standards and does not consider the expectations to introduce unnecessary obstacles for removal of the SIG MA cap in stress. For avoidance of doubt, the PRA is open to firms potentially reducing the stressed FS for SIG assets, relative to the FS that would have been used were the SIG MA cap to be applied, provided this is justified appropriately as set out in paragraph 4.29 of SS8/18.

Proportionality of requirements and expectations for internal credit assessments

4.26 The PRA proposed to substitute (with slight modification) existing PRA expectations with requirements that internal credit assessments of assets would have to satisfy. The PRA also proposed to update the expectations to reflect the substituted requirements and the PRA’s current supervisory approach.

4.27 Ten respondents commented on proportionality, either in terms of the proposals on internal credit assessments as a whole or regarding specific aspects. Most respondents requested that the PRA adopt a proportionate approach, particularly towards those firms with smaller MA portfolios or limited exposure to internally rated assets, given the increase in governance costs. Two of the respondents went further and suggested that certain aspects of the policy should not apply to these firms. Another two respondents highlighted that not all internal credit rating assessments pose the same level of risk to the PRA’s objectives. One of the respondents recommended that the PRA take a pragmatic approach, commenting that it is feasible that an internal credit rating subject to suitable governance may be more appropriate than an external Credit Rating Agency (CRA) assessment, particularly for emerging or innovative assets.

4.28 After considering the responses, the PRA has decided not to change the draft policy as ratings are and will continue to be, a key driver of the FS. Nevertheless, the PRA is mindful of the impact of its proposals on smaller firms and firms with smaller MA portfolios or limited exposure to internally rated assets. The PRA’s supervision of these requirements will be risk-based and proportionate, consistent with its usual supervisory approach, and it will continue to engage with firms accordingly.

Assurance on internal credit assessments

4.29 The PRA proposed that internal credit assessments must be subject to proportionate independent external assurance to ensure that the internal credit assessment outcomes lie within a plausible range of issue ratings that could have resulted from a CRA.

4.30 Eight respondents requested further guidance or clarification on the PRA’s expectations for assurance on internal credit assessments covering a range of topics. Five of these commented on, or specifically sought clarification on, the possible providers of external assurance and whether these had to be CRAs.

4.31 Three respondents noted that in some cases CRAs did not have a clear methodology available to rate specific assets, with one respondent highlighting that firms may need to rely on their internal ratings frameworks being as robust as the CRA frameworks. Another commented that they need to allow for novel risks without having to pay a CRA to develop a methodology and that some internal ratings can be better than external ones. They added that firms should be allowed to deviate from CRA criteria, provided that they have gone through a robust internal governance and testing process. They also said that alternative methods of validation should be considered and that the costs and benefits of CRA validation against alternatives should be borne in mind. The final respondent commented that the policy should be updated to reflect that the lack of a comparable CRA methodology should not be a barrier to investment.

4.32 One respondent questioned the level of assurance to be provided and if the same level was required in all cases. They also asked if the PRA expected the proposed assurance to assess whether the credit assessment for a particular asset is within a plausible range, as opposed to whether the firm has an appropriate rating methodology. One respondent suggested that the PRA set an expectation for firms to categorise internal ratings by their broad degree of prudential risk, with ratings in the lowest category of risk, such as those where the rating could be inferred from an externally rated pari passu (equal ranking) bond from the same issuer, being exempt from relevant expectations such as external assurance. Another respondent was concerned that all internally rated assets would be treated equally, despite having different levels of risk of mis-statement. A further respondent stated the need for more guidance on the coverage and frequency of the external assurance.

4.33 Finally, one respondent requested clarification on whether the internal credit assessment remained in scope of the external audit, noting that this may result in duplication of work.

4.34 After considering the responses the PRA has decided not to change the draft policy consulted on in CP19/23.

The PRA clarifies that the proposed policy is not specific on the possible providers of independent external assurance on the internal credit assessment outcomes. While in most cases CRAs will be best placed to provide assurance on outcomes, assurance may also be obtained from other providers, but firms will need to consider their specific expertise and the level of assurance provided. Firms may also consider using other providers to support the validation of their internal credit assessment processes.
As set out in paragraph 2.7A of SS3/17, a firm should select the validation frequency and coverage sample size according to the complexity and materiality of its internally rated assets. For example, where ratings can be reliably inferred from a pari passu (equal ranking) bond that is externally rated from the same issuer, external assurance should not be necessary.
The PRA will continue to engage with firms and assess their current and forward-looking validation plans and approaches, as part of its ongoing supervisory activities, particularly where CRAs’ methodologies are lacking, or other limitations exist.
The PRA also clarifies that in its proposed policy there is no intention to change the scope of the audit work on internal credit assessments. Auditors will need to continue to determine what reliance to place on the assurance obtained by a firm as part of their own assessment.

Responsibility for the internal credit assessment function

4.35 Given the critical role of the internal credit assessment function in meeting the proposed requirements, the PRA proposed certain expected criteria for the individual with responsibility for the function.

4.36 Five respondents commented on the proposals. Two respondents noted that they would be concerned if the PRA was to impose or enforce an organisational design on firms. Another respondent suggested that the PRA set out an explicit view of the proposed structure of the credit rating function. The same respondent also commented that the expectation to designate an individual responsible for the internal credit assessment, who has been appointed by the management body, is disproportionate for firms with a small volume of private credit assets. Another respondent said that it was not clear that introducing another discrete responsibility into the process will lead to better risk management outcomes and requested further clarification. The remaining respondent wanted greater focus on the independence and integrity of the parties assigning the ratings, rather than the split between internal and external ratings.

4.37 Having considered the responses, the PRA has revised the wording proposed in paragraph 2.8J of SS3/17 to clarify its expectations, which consider the nature, scale and complexity of assets held by the firm. The PRA recognises that firms have different organisational structures and there is no policy intent to impose a specific organisational design on firms. Nevertheless, firms must ensure the independence of the internal credit assessment function and that effective controls are in place to manage any potential conflicts of interest.

Other responses

4.38 One respondent commented that in finalising its policy, the PRA should explain why it is appropriate that insurance regulations, particularly in relation to the FS, should include such reliance on credit ratings. The PRA considers that credit ratings provide a relevant assessment of risk for usage in prudential regulation. As noted in Chapter 1, the proposals in CP19/23 have been developed to be consistent with the legislative framework for the MA and FS. As part of the proposed attestation process, firms will have to consider the extent to which the FS is sufficient compensation for all retained risks, including potential risks not captured in ratings.

4.39 One respondent commented specifically on paragraph 2.7D of SS3/17, noting that the words ‘any potential weakness or issue’ set a very low bar and do not consider the materiality of any risk arising. The PRA clarifies that materiality should be taken into account by a firm when considering the appropriateness of its FS and MA as part of the attestation process and the adequacy of its SCR. Although individual weaknesses or issues may not be material on their own, in some cases they could be when considered together. The PRA has therefore decided not to change the draft policy.

4.40 One respondent flagged that the hyperlink to Commission Implementing Regulation (EU) 2016/1800 included in paragraph 4.32 of CP19/23 did not include all Financial Conduct Authority (FCA) authorised CRAs. The PRA clarifies that the hyperlink was incorrect as it pointed to the original rather than the updated version of the regulations. The latest available version can be found on the UK legislation website.

5: MA permissions, breaches, and consequential rule changes

Introduction

5.1 This chapter provides feedback to responses relating to the proposals in Chapter 5 (MA Permissions, Breaches, and Consequential Rule Changes) of CP19/23. It also contains the PRA’s final policy, as follows:

amendments to the Matching Adjustment Part of the PRA Rulebook (Appendix 1);
updates to SS7/18 – Solvency II: Matching adjustment (Appendix 3); and
updates to SoP – Solvency II: Matching Adjustment Permissions (Appendix 8).

5.2 In Chapter 5 of CP19/23 the PRA proposed to:

restate various existing regulations relating to MA eligibility conditions and applications into PRA rules and also into a new MA SoP;
create an additional MA eligibility condition that firms must demonstrate that the portfolio of assets and each individual asset can be managed in line with the PPP;
create a streamlined approach for granting permission to apply the MA for certain applications;
amend the consequences of breaching MA eligibility conditions; and
clarify PRA expectations around the use of delegated authority to submit MA applications.

5.3 The PRA received 15 responses to Chapter 5 of CP19/23. Respondents generally welcomed the PRA’s proposals but requested additional changes and clarifications. In particular, several respondents sought for the PRA to go further with its proposed changes to the application process. Many requested that the documentation requirements be reduced more and some suggested that firms should be allowed to include a limited proportion of assets in an MA portfolio without, or prior to, receiving the relevant MA permission. Details of the responses, and the PRA's feedback and final decisions, are set out below.

Changes to draft policy

5.4 After considering the respondents’ comments, the PRA has made changes to the draft policy. A summary of the changes is set out below.

Triggers for variations of MA permissions – the PRA has updated paragraphs 7.25, 9.1A, 9.2, 9.4, 9.5 and 9.6 of SS7/18 and paragraphs 1.2, 2.4, 2.20, 2.33, 3.9 and 3.11 of the MA SoP (the latter two having originally been numbered 3.11 and 3.13 in the consultation draft of the MA SoP) to remove references to ‘new risks’ triggering variations of MA permissions. The only exception is where new combinations of features give rise to material risks resulting from interactions and/or dependencies not considered as part of the existing permission.
Timeframes for MA application reviews – the PRA has updated the MA SoP to include a new chapter (Chapter 4: Monitoring of the MA permissions framework) to reflect the PRA’s approach to public reporting on the MA review framework.
Use of MA permissions – the PRA has updated paragraph 2.20 of the MA SoP to clarify the PRA approach where aspects of a firm’s MA permission are not in use.
Concurrent applications – the PRA has amended paragraph 2.27 of the MA SoP to clarify that the PRA considers that firms are not required to await the determination of one MA application before the submission of another.
Reasons for refusal – the PRA has added a new paragraph 2.34 to the MA SoP to clarify its intention to provide reasons where it decides to refuse an application.
Application documentation – the PRA has amended paragraphs 3.10, 4.3, 4.8, 4.10, 4.12, 5.1, 5.5, 5.8, 6.1, 6.2, 6.5, 7.2, 7.3, 7.6 – 7.9, 7.13, 7.14, 7.16, 7.19, 7.21 and 7.22 of SS7/18 and paragraphs 2.3, 2.4, 2.5, 2.7, 2.9, 2.11, 2.14, 2.16, 2.18, 2.19, 2.22 and 3.3 of the MA SoP, and deleted paragraphs 2.7, 3.3, 4.1, 4.9, 5.2 – 5.4 and 7.24 of SS7/18, to clarify its expectations for firms submitting MA applications, including a reduced scope of required documentation.
SCRs for firms in breach of MA eligibility conditions – the PRA has updated SS7/18 with a new paragraph 8.1G to clarify that firms that have applied a reduction to the MA as a result of a breach will not be expected to recalculate the SCR to reflect this reduction.

5.5 The PRA considers that its amendments address the responses received on the draft policy contained in Chapter 5 of CP19/23. Any significant changes to the CBA or ‘have regards’ analysis arising from changes to the draft policy proposed in CP19/23 are considered in Chapter 10 (Cost benefit analysis) and Chapter 11 (Have Regards analysis) of this PS.

Feedback to responses

5.6 The sections below have been structured along the same lines as Chapter 5 of the CP. The responses have been grouped as follows:

restatement of existing regulations into PRA rules and new MA SoP;
PPP;
streamlined MA application approach;
breaches of MA eligibility conditions; and
delegated authority to submit MA applications.

Restatement of existing regulations into PRA Rules and new MA SoP

5.7 In CP19/23 the PRA proposed that certain provisions within the Solvency 2 Regulations 2015 that have not been restated in the IRPR Regulations or within the Technical Provisions Part of the PRA Rulebook, and would otherwise have been deleted, should be established without change in the Matching Adjustment Part of the PRA Rulebook. The PRA also proposed to restate, explicitly or implicitly, the existing provisions of the on-shored Commission Implementing Regulation (EU) 2015/500 (the ‘MA ITS’) that continue to be relevant and appropriate in the new regime, in a new MA SoP. In order to reduce the burden on firms, the PRA proposed not to restate certain provisions of the MA ITS relating to the submission of documentation that the PRA did not consider necessary for making a decision on an application. Finally, the PRA proposed to set an expectation that firms with an MA permission should consolidate all information provided in support of this permission into one suite of documents.

5.8 Responses to these proposals were generally supportive. Several respondents asked for further commitments from the PRA to improve the speed of MA applications.

One respondent asked for the PRA to highlight other items that could be requested on an ad hoc basis rather than being required as part of an MA application.
Four respondents requested that the PRA make clearer commitments to timeframes for reviewing MA applications. One of the respondents suggested that the PRA set an expected timeframe of two months for streamlined applications, while another asked that the PRA regularly report on the length of time taken to respond to MA applications going forward. One respondent requested that the pre-application process should also be subject to a set timeframe.

5.9 Having considered the responses, the PRA has decided to make changes to the draft policy.

The PRA has further evaluated the documentary evidence it has previously expected firms to provide to enable the PRA to assess an application against the MA eligibility conditions. As a result, the PRA has reduced the extent of documentary evidence that firms will generally be expected to submit (noting that there may be circumstances where the PRA considers it necessary to request additional information). In particular the PRA does not generally expect that firms should provide documentary evidence in relation to:
- Rule 2.2(6) of the Matching Adjustment Part of the PRA Rulebook, relating to compliance with the PPP, or Regulation 4(4) of the IRPR Regulations (the requirement that the credit quality of MA assets be capable of being assessed through a credit rating or an internal credit assessment of a comparable standard);
- comprehensive assessment of the risks associated with the exercise of surrender options;
- some quantitative information that the PRA previously expected to be included in MA applications, including:
  - a breakdown of the number and value of each type of insurance contract;
  - assessments of market risks arising from any cash flow mismatch;
  - details regarding the exclusion of excess assets (ie assets not required to demonstrate matching) from the MA calculation;
  - results of matching tests (Appendix 1 of SS7/18) and asset and liability cashflow data; and
- the firm’s proposed methodology for determining best estimate cash flows (other than where the firm is including assets with HP cash flows and not proposing to use the standard approach to calculate any FS addition).
Nevertheless, the PRA will expect the above information to be available upon request where it is considered necessary to support its review of the application. The PRA has also published updated supplementary information forms on its website, which firms are strongly encouraged to submit alongside the required section 138BA application form. These will help firms to ensure that applications contain appropriate evidence to enable PRA review. The MA SoP has been updated to reflect the documentary evidence that the PRA expects firms to submit.
The PRA will publish regular reports on the MA framework alongside the PRA Annual Report, covering application review timelines and decision rates, with the first report to be published in 2025. As noted in CP19/23 and the MA SoP, the PRA has set a target of six months for decisions on MA applications. The PRA also intends to develop a specific target timeline for the completion of streamlined reviews. This will be determined after implementation of the current reforms, reflecting the experience of the new regime. The MA SoP has been updated with a new chapter to reflect these changes (Chapter 4: Monitoring of the MA permissions framework) where appropriate.
The PRA does not consider it appropriate to determine a set timeframe for the pre-application process. However, the PRA has developed an Application Readiness Assessment Process (ARAP) to support a more consistent and efficient approach to engagement with firms in advance of submission of an MA application. Firms are strongly encouraged to engage in the ARAP and submit the related ARAP form, which is available on the PRA website. This will assist the PRA in assessing the firm’s readiness to submit a formal application and help the PRA determine resourcing needs and the likely level of review required.

5.10 Two respondents considered that having to make applications to include assets with HP cash flows in MA portfolios, where the FS addition follows a standard approach, would be a barrier to investment and could slow the pace of investment into such assets. One of these respondents further considered that MA applications where a sophisticated approach to the FS addition was applied would be a barrier to investment or could slow the pace of investments into such assets.

5.11 The PRA considers that the MA application process is an important tool for advancing the PRA’s statutory objectives and that it would not be appropriate for assets with HP cash flows to be included in MA portfolios without appropriate scrutiny. Accordingly, no change has been made to the draft policy in relation to this point.

5.12 One respondent expressed concern that the PRA had committed to a minimum six-month decision-making period for non-streamlined reviews, which the respondent considered would impede swift investment.

5.13 Having considered this response, the PRA has decided not to change the draft policy in this area. The MA SoP states that the PRA expects to reach a decision on an application ‘no later than six months from the date of receipt’. As such, the six-month timeframe represents an expected maximum, rather than a minimum.

5.14 Three respondents requested that the PRA provide more information on the meaning of 'asset types/classes/features', as used in PRA policy material. Two of these respondents requested that the PRA provide a list of asset classes that have been approved for inclusion in MA portfolios, and also queried the application of ‘asset types’ in relation to some of the graphics included in CP19/23. One respondent requested that the PRA publish a list of assets that the PRA deemed acceptable to be treated as having HP cash flows, without specific MA applications being required.

5.15 Having considered the responses, the PRA has decided not to change the draft policy in this area. In particular, the PRA does not consider it feasible to provide a comprehensive list of asset features that satisfy MA eligibility, though the PRA may from time to time publish expectations that clarify where it considers certain features to render an asset ineligible. In order to provide a list of acceptable asset classes/types, the PRA would have to review a list of assets without applications from firms, which the PRA considers to be a disproportionate use of its resources. Furthermore, any list of approved or ‘pre-approved’ asset features would provide no guarantee of a successful application, as the MA eligibility conditions also cover the management of assets. The PRA confirms that year end 2020 MALIR data were used to produce Chart 2 in CP19/23, and the MALIR instructions (Appendix 8 to CP19/23) provide definitions for the asset types used.

5.16 One respondent expressed concern that the expectation that firms should have credible plans to invest in a proposed new asset within 12 months was unduly limiting, and would be contrary to good risk management if a firm were obliged to enter into a derivatives contract or take on insurance obligations within a specific timeframe in order to retain an MA permission.

5.17 Having considered this response, the PRA has updated paragraph 2.20 of the MA SoP to provide greater clarity. The PRA does not consider that speculative applications from firms would be appropriate, given the need to appropriately manage PRA resources. Nevertheless, the PRA’s approach to considering applications relating to future assets or liabilities will be proportionate and reflect the specific circumstances of a firm’s application.

5.18 One respondent asked the PRA to commit to providing justification for the reasons for a rejection of an application, even if it is not required to under section 138BA.

5.19 Having considered this response, the PRA has added a new paragraph 2.34 to the MA SoP to clarify that where it decides to refuse an application for permission to apply an MA, or an application to vary a permission, it will communicate the reasons for this decision to the firm along with notification of its decision. This is in line with the requirements of Regulation 4 of the Financial Services and Markets Act 2000 (Disapplication or Modification of Financial Regulator Rules in Individual Cases) Regulations 2024.

5.20 Nine respondents raised concerns about the PRA’s proposed language in SS7/18 to describe circumstances where an application to vary an MA permission would be required. Four of these respondents commented on the addition of ‘new risks’ as such a circumstance, suggesting that it was not necessary and would lead to many more additional applications than under the current MA regime. Five of these respondents raised similar concerns over the proposed expectation that firms submit new applications for new combinations of existing approved asset features, with two of these respondents challenging whether this policy would be unfairly onerous for firms with less extensive existing MA permissions. Two of these respondents expressed the view that the expectation that firms submit an application for a variation of permission for new ‘features’ was a new and burdensome expectation, which would lead to a large increase in applications.

5.21 Having considered the responses, the PRA has updated paragraphs 7.25, 9.1A, 9.2, 9.4, 9.5 and 9.6 of SS7/18 as well as paragraphs 1.2, 2.4, 2.20, 2.33, 3.9 and 3.11 of the MA SoP (the latter two having originally been numbered 3.11 and 3.13 in the consultation draft of the MA SoP). The PRA’s intention with regard to this proposal in CP19/23 was to restate existing provisions from the MA ITS, in clear language, in the MA SoP and SS7/18, and provide clarity relating to existing practice. The PRA did not intend that the proposed wording would extend the scope of circumstances in which firms would be expected to apply to vary an MA permission. As such, while the PRA considers it appropriate to retain references to ‘new features’, which should be interpreted in the context of the MA eligibility conditions, it has removed references to ‘new risks’ in the MA SoP and SS7/18, except in the context of new combinations of existing features. The PRA considers that where assets possess features covered by an existing permission, but in a new combination, this should be covered in a new application where that combination gives rise to material risks associated with interactions and/or dependencies not considered as part of the existing permission.

5.22 A few respondents asked for minor clarifications.

One respondent noted the deletion of paragraph 9.7 in SS7/18 and asked for clarification over whether the PRA would consider the submission of concurrent applications to be appropriate.
One respondent noted that the IRPR Regulations did not retain the requirement for assets not to be used to cover losses arising outside the MA portfolio and asked that paragraph 7.1 of SS7/18 be updated to reflect a more permissive expectation regarding the mixing of collateral between the MA portfolio and non-MA portfolio.
One respondent sought clarification on whether it would be possible to partially revert to an approach of not applying the MA to a portion of the portfolio, by removing that portion from the portfolio.
One respondent queried whether a formal pre-application process would be required in all circumstances.

5.23 Having considered the responses, the PRA has made a minor change to paragraph 2.27 of the MA SoP to clarify that firms are not necessarily required to await the decision on one MA application before submitting a further application. Paragraph 2.27 of the MA SoP already sets out the approach firms may wish to take where concurrent applications are submitted.

5.24 The PRA has decided not to make any other changes to the draft policy materials in relation to these requests for clarification. This is because:

CP19/23 explicitly proposed the retention of the restriction on the use of assets within an MA portfolio to cover losses arising outside the MA portfolio, in Rule 2.2(5) of the Matching Adjustment Part of the PRA Rulebook. The PRA continues to consider this an appropriate condition of an MA permission, and as such it has not made changes to paragraph 7.1 of SS7/18;
as stated in paragraph 2.21 of the MA SoP, firms wishing to remove liabilities from an MA portfolio are expected to obtain approval for a variation of the MA permission; and
a formal pre-application process is not a pre-requisite of an application to vary an MA permission, but as noted in paragraph 2.3 of the MA SoP, the PRA ‘welcomes early engagement with a firm that is considering submitting an MA application’. As noted above, the PRA also strongly encourages firms to engage in the ARAP, which the PRA has developed to support a more consistent and efficient approach to engagement with firms in advance of submission of an MA application.

PPP

5.25 In CP19/23 the PRA proposed an additional MA eligibility condition requiring that the relevant portfolio of assets and each individual asset contained within it meet the requirements of the PPP contained in Chapters 2 and 3 of the Investments Part of the PRA Rulebook (Rule 2.2(6) of the Matching Adjustment Part of the PRA Rulebook).

5.26 Of the responses received, three were supportive of the proposal, while one respondent objected to a specific part of the proposal, noting that some parts of the PPP wording (including within Investments Chapter 5) can only reasonably be applied to the portfolio as a whole, and not to individual assets.

5.27 Having considered the responses, the PRA has decided not to change the rule as drafted. The text of the rule proposed in CP19/23 refers only to Chapters 2 and 3 of the Investments Part of the PRA Rulebook, and not the chapter mentioned by the respondent. The PRA considers that the Rulebook chapters referenced in the proposed rule do not contain rules that are inapplicable to individual assets.

Streamlined MA application approach

5.28 In CP19/23, the PRA proposed the introduction of a streamlined MA application approach for certain types of applications. Applications reviewed under this approach would only be assessed against the MA eligibility conditions prior to permission being granted. The PRA proposed that assessment of other factors relating to the ongoing application of the MA (eg ratings or valuations) may be deferred until after MA permission has been granted, and conducted as part of the PRA’s ongoing supervision of the firm.

5.29 While many respondents were broadly supportive of these proposals, several argued for additional changes to further streamline the application process.

Two respondents requested that the PRA streamline all applications to vary an MA permission, rather than limiting it to those proposed.
Three respondents proposed allowing firms to make minor changes to an MA permission (eg changes to safeguards or exposure limits) without formal application to the PRA, relying instead on internal processes and external audit.

5.30 Having reviewed the responses, the PRA has decided not to implement these suggestions. The PRA considers that for the purposes of policyholder protection it is necessary to review all applications to vary an MA permission before such a variation may be implemented. In some particularly complex cases this will mean reviewing aspects such as valuation and internal ratings in order to ensure an appropriate impact on the firm’s MA benefit and SCR. Nevertheless, the PRA notes the comments received indicate a desire for efficient MA application review processes, and having reviewed the proposed process it has identified that further efficiency gains are possible. These include (additional examples of policy changes to enable further efficiency gains are noted in paragraph 5.9):

the ARAP to support engagement between the firm and the PRA in advance of a formal application;
a triage process, through which an initial assessment of the scope and completeness of applications will be made, before appropriate applications are allocated to a streamlined review path;
a reduction in the documentary evidence that the PRA expects a firm to submit within its application (with greater reliance placed on the firm’s written confirmation that the assets meet the requirement that the credit quality is capable of being assessed through a credit rating or an internal credit assessment of a comparable standard, and the requirements of the PPP); and
updated supplementary information forms published on the PRA website to ensure that applications contain appropriate evidence to enable PRA review.

5.31 Five respondents raised additional suggestions for the PRA’s consideration that were not directly related to the proposals in CP19/23. The suggestions received fell broadly into two design variants, which would permit a limited proportion of the MA portfolio to be used for, and MA benefit derived from:

assets that a firm considers meet the MA eligibility conditions, but for which either the streamlined application process would be too slow to facilitate swift investment, or the costs of the application process would be prohibitive, given the size of the firm’s exposure; and
assets that explicitly do not meet the MA eligibility conditions, but which the firm nevertheless considers suitable to back annuity liabilities, with the suggestion that they be treated as possessing HP cash flows with prudent assumptions until the long-term treatment is agreed with the PRA.

5.32 One respondent provided an unprompted objection to the second suggestion above, arguing that it would threaten the prudential nature of the MA.

5.33 Having considered these responses, the PRA considers that it would not be appropriate to implement either of these proposals at this time. The PRA did not consult on introducing processes of this nature, and at this time has not determined whether these proposals would advance its objectives. In addition, attempting to introduce proposals of this nature would require changes to PRA rules and/or policy that would need to be developed through the formal process of public consultation. Nevertheless, the PRA is open to considering how the MA application process may be developed to ensure that life insurers are able to make appropriate long-term investments that advance the PRA’s statutory objectives. To that end, the PRA has commenced engagement with the life insurance industry to obtain an appropriate range of ideas and views in this area, and will monitor the effectiveness of the new regime.

Breaches of MA eligibility conditions

5.34 In CP19/23, the PRA proposed to retain the existing two-month period provided for firms to restore compliance with MA eligibility conditions in light of a breach, but where compliance is not restored within the two-month window, firms would automatically be required to reduce the amount of MA in a staggered fashion, rather than be subject to immediate loss of the MA permission for at least 24 months.

5.35 The PRA proposed that this reduction would be at least 10% of the unadjusted MA, increasing by an additional 10% for each further month after the two-month window that a firm is not in compliance with MA eligibility conditions. If the MA has been reduced to zero, the PRA would expect to revoke the permission to apply the MA.

5.36 The PRA received several comments in support of the proposals, but some respondents requested minor clarifications. Five respondents asked the PRA to clarify how the reduction to a firm’s MA in response to an extended breach would interact with the firm’s calculation of its SCR. Two respondents requested clarification on whether management actions to restore compliance could continue to be allowed for in internal models, if in fact a firm were in breach.

5.37 Having considered the responses, the PRA has updated SS7/18 (new paragraph 8.1G) to provide clarification. The PRA considers that while there may be circumstances (potentially related to the nature of the breach) where it may be appropriate for a firm to reflect the breach in its SCR calculation, this will not necessarily always be the case. As a result, the PRA does not expect that firms should have to recalculate the SCR to recognise the reduction in MA. The PRA considers it appropriate for firms in breach of the MA eligibility conditions to continue to allow for management actions to restore compliance within internal models, where those management actions may be undertaken within a 12-month window. The PRA notes that expectations regarding the allowance for management actions relating to the MA portfolio in the SCR calculation are set out in SS8/18.

5.38 One respondent asked the PRA to confirm that the expectations set in paragraph 9.2 of SS7/18 relating to ‘variations of MA permissions’ would be applied to each firm consistently, and to clarify whether failure to comply with those expectations would be considered a breach of MA compliance.

5.39 Having considered this response, the PRA has decided not to change to the draft policy in this area. The PRA considers that failure to comply with PRA expectations would only incur the consequences set out in Chapter 13 of the Matching Adjustment Part of the PRA Rulebook where this includes a breach of the MA eligibility conditions. The PRA’s overall approach to supervision, including consideration of consistency in the treatment of different firms, is set out on the PRA website.^{footnote [17]}

5.40 One respondent suggested that the two-month window should be extended to three or four months to recognise the time it takes to adjust the MA portfolio. Another respondent asked for confirmation that the two-month window for rectification of breaches would start when the breach is identified, rather than when it occurs, and also asked for confirmation that the 10% reduction to the MA could be varied as deemed necessary by the PRA.

5.41 The PRA has reviewed the responses and decided not to change the draft policy materials in this area. Taking the response requesting an extension to the two-month window into consideration alongside other comments in support of the proposal, the PRA has not changed the draft policy, as it considers the proposed framework will provide a proportionate balance between making allowance for the additional risks during non-compliance, and allowing firms sufficient time to undertake the necessary management actions, with a reduced risk of balance sheet instability in response to minor breaches. Paragraph 8.3 of SS7/18 sets out PRA expectations on the appropriate ‘start date’ for the rectification of a breach. Paragraph 8.1B of SS7/18 sets out PRA expectations that the 10% reduction in MA may be varied where appropriate.

5.42 Some respondents asked for minor clarifications.

One respondent queried whether a breach penalty would be removed if the breach were remedied.
One respondent queried how the limit on the MA benefit generated by assets with HP cash flows would be affected by the reduction in MA following a breach.

5.43 Having considered these responses, the PRA has decided not to change the draft policy in this area. As stated in paragraph 8.1B of SS7/18, ‘A reduction of the MA will cease to apply once the firm restores compliance with MA eligibility conditions. The PRA expects that the firm will discuss with its usual supervisory contact whether a particular breach has been satisfactorily resolved before removing a reduction to its MA’. As stated in Rule 5.5 of the Matching Adjustment Part of the PRA Rulebook, the calculation of MA benefit in the context of the exposure limit for assets with HP cash flows should ignore any reduction in MA resulting from a breach of MA eligibility conditions.

Delegated authority to submit MA applications

5.44 Four respondents noted their support for the PRA's proposal to clarify its expectations, in paragraph 9.6A of SS7/18, that the board of a firm could delegate authority for submission of new MA applications and applications to modify the scope of existing MA permissions, to a suitable sub-committee of the board or to approved senior managers. The PRA has therefore decided not to change the draft policy materials in this area.

6: Matching adjustment attestation

Introduction

6.1 This chapter provides feedback to responses relating to the proposals in Chapter 6 (Matching adjustment attestation) of CP19/23. It also contains the PRA’s final policy, as follows:

chapters 9 to 12 of the new Matching Adjustment Part of the PRA Rulebook (Appendix 1);
updates to SS7/18 ‘Solvency II: Matching Adjustment’ (Appendix 3);
updates to SS3/17 ‘Solvency II: Illiquid unrated assets’ (Appendix 5); and
update to SS11/16 ‘Solvency II: External audit of, and responsibilities of the governing body in relation to, the public disclosure requirement’ (Appendix 7).

6.2 In Chapter 6 of CP19/23 the PRA proposed to introduce the following requirements:

an MA attestation must be made to the PRA using standardised wording that is set out in the PRA Rulebook;
an attestation must be given, for each MA portfolio within the firm, annually and additionally upon any material change in the firm’s risk profile;
the PRA senior management function holder (SMF) who holds the prescribed responsibility for the production and integrity of the firm’s financial information and its regulatory reporting (PR Q),^{footnote [18]} must be responsible for the attestation;
a policy on providing the attestation must be put in place and maintained by firms, as well as appropriate internal processes, systems and controls to allow a firm to analyse and justify its use of the FS in accordance with the attestation; and
an attestation document must be provided to the PRA, setting out the attestation itself alongside a supporting attestation report.

6.3 The PRA received 21 responses to Chapter 6 of CP19/23. Respondents were broadly supportive of many elements of the proposal, including the purpose of increasing firms’ accountability for the level of MA benefit being taken, the removal of liquidity premium from the attestation wording, the proportionate approach to reviewing and the list of factors to consider for the FS, and not subjecting the attestation material to external audit or disclosure. However, respondents also made a number of observations and requests for changes and clarification, which are set out later in this chapter.

Changes to draft policy

6.4 Following consideration of the respondents’ comments, the PRA has amended:

its expectations in paragraph 5.35 of SS7/18 for firms’ analysis of corporate bond portfolios where these broadly reflect the calibration data and have up-to-date accurate credit ratings; in such circumstances firms are generally expected to rely on the basic FS;
the wording of paragraph 5.36 of SS7/18 to clarify that, to facilitate a top-down initial approach, firms are expected to be able to group assets into homogenous risk groups when determining whether FS additions are needed. Where necessary, this should be followed up by examination of specific assets;
the wording of paragraph 5.37 of SS7/18 to clarify the PRA’s expectations of firms in establishing the degree of confidence in the ability to earn the MA on all assets by reference to the degree of confidence achieved for corporate bonds by applying the basic FS;
SS3/17 to include a new paragraph 3.25A to clarify that where the EVT is used for ERMs for the purposes of attestation, firms are expected to use their own assumptions that they judge to be appropriate for the attestation requirements and these should not fall below the PRA’s published minimum parameters;
the wording of paragraphs 5.35 and 5.36A of SS7/18 to incorporate some of the suggestions from respondents on the expectations for FS additions, including rebalancing costs and an example of a characteristic of a corporate bond that may not be reflected in the calibration data;
the final rules to specify the attestation reference date for firms’ out-of-cycle attestations, which will be a date no later than three months after the date of the material change in risk profile;
the wording of paragraph 5.33 of SS7/18 to clarify the purpose of a firm entering into a discussion with the PRA in the situation of an out-of-cycle attestation;
the wording of paragraph 5.41 of SS7/18 and added a new paragraph 4.2FA in SS11/16 to clarify that there is no change to the scope of external audit arising from either the attestation requirement or the option for firms to apply voluntary FS additions;
the wording of paragraph 5.41 of SS7/18 to clarify that there is no requirement, as opposed to no expectation, from the PRA for a firm to publicly disclose the content of its attestation report; and
the final rules to include a six-month transitional period such that firms may, but will not be required to, provide any annual or out-of-cycle attestations with an attestation reference date that is before 31 December 2024.

6.5 Any significant changes to the CBA or ‘have regards’ analysis arising from changes to the draft policy proposed in CP19/23 are considered in Chapter 10 (Cost benefit analysis) and Chapter 11 (Have regards analysis) of this PS.

Feedback to responses

6.6 The PRA has considered the responses to Chapter 6 of CP19/23. Feedback to the responses has been grouped into the following broad headings:

FS and MA attestation wording;
FS additions guidance;
attestation frequency and responsibility;
public disclosure requirement and external audit; and
implementation.

FS and MA attestation wording

6.7 The PRA proposed in Rule 9.1(1) of the Matching Adjustment Part of the PRA Rulebook, a standardised wording for the attestation where all assets are included in the review, considered on an asset-by-asset basis, with no offsets between different assets and taking proportionality into account.

Treatment of corporate bonds

6.8 Twelve respondents commented on the treatment of corporate bonds overall. Seven of these respondents stated that the analysis expectations for corporate bonds were overly onerous and disproportionate, with three suggesting that no ‘burden of proof’ should be required and two suggesting that only outliers should require further consideration. A few of these respondents highlighted that FS additions for corporate bonds would be contrary to the Government’s decision to leave the basic FS unchanged, where this FS is already stated to meet the requirements and assumptions of transparency, prudence, reliability and objectivity. One respondent suggested that whether the basic FS results in a high degree of confidence (HDC) in the MA for corporate bonds should be a matter for policy design.

6.9 Three respondents suggested that the PRA could help firms meet its expectations by publishing the source data, the list of risks, key judgments used in the determination of the basic FS and any limitations, as well as worked examples for different asset classes. One other respondent suggested that an industry-wide review of the basic FS’s appropriateness would save costs to individual firms. Two other respondents agreed that corporate bonds would generally require less focus and that focus on outlier assets would result in a more effective review process.

6.10 Having considered these responses, the PRA has amended its expectations in paragraph 5.35 of SS7/18 for firms’ analysis of corporate bond portfolios that broadly reflect the portfolio used to calibrate the FS and have up-to-date accurate credit ratings, where reliance can generally be placed on the basic FS. Under this simplified analysis process, firms will not be expected to consider risks that may not have been adequately captured by the historical credit performance data for corporate bonds used to calibrate the basic FS. Firms will still be expected to consider the need for any FS additions for other reasons that are stated in the same paragraph in SS7/18. The PRA considers that this amendment better reflects the proportionate approach that firms are expected to take and leads to a more focused review of particular assets within this asset class (which should be the minority). The PRA also considers that the retained expectations from the proposal would implicitly capture at least some of the risks not reflected in the calibration data, hence would not have a material adverse impact on policyholder protection and firms’ safety and soundness.

6.11 With regard to the basic FS being unchanged, the PRA reminds firms that they still need to satisfy themselves that applying it to the assets they hold, including corporate bonds, meets the attestation requirements and they have the option to apply increases to the FS where they conclude that it does not. This is clear from Regulation 6(9) of the IRPR Regulations where the basic FS, while unchanged, is now a minimum.

Analysis level and offsets

6.12 15 respondents commented on the level at which attestation analysis should be carried out and / or the level at which FS additions should be applied.

6.13 Six of these respondents stated that analysis should be performed at the homogenous risk group (HRG) level for practicality reasons, with three of these respondents also recognising that idiosyncratic risks could still exist within these groups.

6.14 Having considered these responses, the PRA has amended the wording of paragraph 5.36 in SS7/18 to clarify that firms are expected to be able to group assets into HRGs when determining whether FS additions are needed as part of the attestation. The PRA considers that HRGs should be granular enough to ensure that both the type and the level of risks are sufficiently similar within each group. As such, HRGs are expected to be defined by a minimum set of factors which the PRA has listed in paragraph 5.36 of SS7/18. In addition, the ability to group assets into HRGs should not preclude firms from identifying idiosyncratic and other downgrade risks and material MA contributors that affect specific assets within each HRG. That is, the grouping of assets into HRGs should merely facilitate a top-down initial approach, followed by examinations of specific assets where necessary, and not result in fewer risks being identified compared to an asset-by-asset analysis.

6.15 In relation to the level at which FS additions should be applied, 14 respondents considered that offsets against other assets should be taken into account when determining the size of any FS additions that were judged to be needed for specific assets. Respondents argued that the FS calibration is based on average statistics such as spread indices and transition matrices, where there will be ‘overs and unders’ across different assets. They also stated that the MA calculation is performed, and the attestation should therefore also be made, at the portfolio level, where diversification of risks should be allowed. These respondents argued that not allowing offsets would be overly prudent. Four of these respondents stated that this did not reflect the portfolio’s underlying risks and disincentivised good risk management and investment behaviour, with one of these respondents suggesting that it could ultimately affect annuitants’ income. Six of these respondents gave specific examples where in their view offsetting between assets should be taken into account when determining FS additions.

6.16 Having considered these responses, the PRA has decided not to change draft policy that firms should not assume that prudence for one asset can be offset against an insufficient FS for another. This is because:

the PRA considers that despite the FS calibration being based on averages, it is still applied at a cash flow level for each asset and allowing offsets would effectively mean setting negative FS additions for some assets, which would undermine the basic FS as per Regulation 6 of the IRPR Regulations;
consistent with the expectation for firms to place reliance on, and therefore accept, the basic FS for corporate bond portfolios that reflect the calibration data and have up-to-date and accurate ratings (which should be the majority of firms' corporate bonds), the PRA does not expect any prudence to be identified in the basic FS for this asset class;
for other asset classes, credible calibration data that satisfies the transparency, prudence, reliability and objectivity criteria in Regulation 6(1) of the IRPR Regulations does not exist and the PRA considers firms should focus their analysis on where the basic FS may not be enough rather than where it may be excessive, given the uncertainty involved; and
the PRA considers that firms will still be able to benefit from some diversification, with implicit offsets between different assets purely due to random fluctuation in the actual experience of each asset within each HRG. This is because adverse credit experience could still result from assets that have not been identified as requiring any FS additions and voluntary FS additions on specific assets might prove to be inadequate relative to the actual credit experience.

High degree of confidence that the MA will be earned

6.17 Four respondents said that an HDC requirement for the MA was not appropriate, because the MA is part of firms' TPs, which are only required to be ‘adequate’ rather than providing a capital buffer. In their view, HDC as a metric would be more appropriately applied when the SCR is also allowed for in addition to the TPs. One of these respondents considered that the HDC requirement was unnecessary given cash flow matching has already been established to be acceptable. Other respondents welcomed the change to the MA attestation requirement from the ‘liquidity premium only’ wording in the November 2022 statement.

6.18 After considering these responses, the PRA has decided not to change the draft policy in this area. The PRA considers that TP adequacy requires a suitable adjusted risk-free discount rate, and the risk-free nature of the discount rate necessitates the HDC requirement for the MA, which is applied to the risk-free rate used in the discounting. In addition, the MA further reduces the credit part of the SCR and the same HDC requirement for the MA should hold under different conditions. Therefore, the PRA does not consider that the HDC requirement for the MA should be replaced or weakened by the presence of the SCR itself. The PRA also considers that the MA cash flow matching and the HDC requirement serve different purposes – the former is designed to ensure that firms would not be forced sellers of their MA assets to meet liabilities, while the latter governs how much MA benefit could be safely claimed up-front by firms as a result.

6.19 13 respondents commented on the percentile that firms should target in meeting the HDC requirement.

Many recognised the significant degree of qualitative judgment involved and considered the measure relative to corporate bonds useful. One of these respondents sought clarification for the expected treatment of credit risk premium (CRP) in meeting the retained risks and HDC requirements, considering the possible need to deviate from the buy-and-hold strategy, and the level of CRP that may be present in the MA of corporate bonds when the basic FS is applied.
Other respondents considered that a target percentile should be prescribed by the PRA to achieve consistency across firms. Six respondents gave their views on an appropriate target percentile, which ranged from ’not significantly above best estimate’ to ’certainty’, with two of these respondents agreeing that it should be materially more than the 50^th percentile. Two of these respondents interpreted the qualitative wording as a means to allow firms to have different views on the reasonable level of risk to accept. One of these respondents noted that differences in interpretation and attestation standards should converge over time.
Two other respondents commented on the ability for firms to make use of their internal models to satisfy the HDC requirement, with one of these respondents expressing concerns over the potential conceptual and practical issues that could arise from this requirement and making further suggestions on potential simplifications.

6.20 After considering these responses, the PRA has decided not to change the draft policy in this area.

The PRA considers that the qualitative interpretation of HDC is clearly set out in paragraph 5.37 of SS7/18. The standard should be commensurate with the MA being an addition to the risk-free discount rate and contributing to Tier 1 own funds. Firms may find the MA on corporate bond portfolios with fixed cash flows helpful as a benchmark, taking into account the review process expected of corporate bonds as outlined in paragraph 5.35 of SS7/18. The PRA has amended the wording of paragraph 5.37 of SS7/18 to clarify this point.
The PRA considers that while considerable judgement lies with the attesting SMF in quantifying the risks retained, which introduces uncertainty into the actual percentile achieved (which is also why a quantitative requirement has not been set), this does not mean that the level of standard should be set according to the firm’s risk appetite and the PRA agrees that the level of standard is likely to converge over time. On the other hand, paragraph 6.19 of CP19/23 explained why the PRA does not consider requiring the MA to be earned free of any risk to be practical.
Finally, the PRA expects that in making the attestation firms will leverage their existing tools and processes with qualitative reasoning applied, as opposed to solely relying on their internal models, given attestation of the MA requires it to be earned over the lifetime of the asset portfolio where expert judgement plays an important role.

Suggestions on FS and MA attestation wordings

6.21 Five respondents suggested that the attestation wording should be further softened or relaxed.

Three of these respondents considered the use of the words ‘all risks’ to be absolute, including those that are unforeseeable as well as unquantifiable, which would make attestation more challenging.
One of these respondents suggested a negative rather than a positive assurance type of wording and around the expectation rather than confidence that the MA will be earned. Another respondent suggested that ‘compensation’ for retained risks cannot be objectively derived and instead the FS should simply refer to an adequate allowance for all retained risks.
One respondent expressed the view that the basic FS does not reflect compensation for all retained risks and the FS attestation wording would need to be changed to avoid a generic increase in the FS for corporate bonds.

6.22 Separately, three respondents suggested that separate analysis of the FS and the MA was not practical, with two of these respondents highlighting that the residual spread itself could vary depending on the asset valuation which could be subjective, hence the focus should be on the robust derivation of the FS instead. In contrast, six other respondents considered the proposed attestation wording to be reasonable, with one of these respondents also recognising the rationale for attesting to both the appropriateness of the FS and the ability to earn the MA.

6.23 Having considered these responses, the PRA has decided not to change the proposed attestation wording.

The PRA notes that the ‘all retained risks’ wording is already aligned to Regulation 7(f) of the IRPR Regulations and the MA attestation wording has already been broadened from the MA being 'liquidity premium only' as per the November 2022 statement, to having a 'high degree of confidence' of being earned. In respect of unforeseeable risks, the PRA considers that although it would be unreasonable to judge a firm’s attestation with the benefit of hindsight, it notes the expectation for firms to consider whether high residual spread could be indicative of any additional but unidentified risks or greater uncertainty which would reduce the confidence that the MA could be earned as set out in paragraph 5.39 of SS7/18.
The PRA also considers that a negative assurance type of attestation would significantly reduce the strength of the measure and that an expectation for the MA to be earned would be clearly short of the HDC that is required. The reference to ‘compensation’ for retained risks indicates that the allowance should be a reflection of what a third party or the market would demand, which is more objective than any allowance that an owner of the asset would set, albeit both will still require judgement.
In response to the respondent whose view was that the basic FS does not reflect compensation for all retained risks for corporate bonds, the PRA considers that firms have the option to apply increases to the FS in such a case, as per paragraph 6.11 above.
Finally, the PRA considers that having first determined the FS, the justification of the resultant MA by firms is important given the significant judgement involved; paragraph 5.35 of SS7/18 also explains that this will add rigour to the process with the MA acting as a market-based check on the level of FS. While asset values are subject to external audit, if there are any uncertainties in the asset valuation, or the attestor thinks that there is a possibility that an asset value is overstated such that there may be an insufficient allowance for risk, a voluntary FS addition could be applied.

FS additions guidance

6.24 The PRA proposed some guidance for firms in determining the need for any FS additions in an example review process in SS7/18, including suggested metrics for material MA contributors that should receive more focus in the review.

Spread volatility, credit rating and the MA

6.25 Five respondents noted that the basic FS design is based on a long-term view of credit risk that is reflected in credit ratings and not market spreads, which means that the resulting MA acts as a valuable countercyclical tool by absorbing changes in spread that are not accompanied by credit rating changes. Two of these respondents said that assets with high absolute MA-to-spread ratios, as indicated in Step 3 of the example attestation process provided in CP19/23, should not be candidates for FS additions. One of these respondents wanted the PRA to state explicitly that there was no intention to apply capital add-ons relating to the current level of asset spreads, as they argued that this would run counter to the Government’s decision to leave the basic design and calibration of the FS unchanged.

6.26 Having considered these responses, the PRA has decided not to change the draft policy in this area, which it considers to be consistent with the legislative MA framework and the November 2022 statement. The PRA does not expect firms to link current asset spreads and potential FS additions in a mechanistic way, whether in a relative sense (ie excess spread compared to average) or an absolute sense. Instead, firms are expected to apply judgement in considering whether information in market spreads, alongside other potential indicators, might contain some ‘signal’ suggesting increased long-term retained risks as opposed to ‘noise’ due to short-term market dynamics. The PRA considers that this approach will ensure the regime is sufficiently risk-sensitive to the assets held in firms’ MA portfolios and will avoid procyclicality, while allowing firms to take into account any long-term structural (as opposed to cyclical) changes in the credit environment, which may be signalled by a sustained increase in the general level of spreads. The PRA also considers that the hold-to-maturity assumption underlying the MA may not always be borne out in practice given the requirement to restore any cash flow mismatch, which means that some exposure to spread volatility may remain depending on firms’ rebalancing strategies. Therefore, although the PRA agrees that it is important that the MA should not incentivise procyclical behaviour, the PRA does not consider the MA to be an explicit countercyclical tool designed to insulate firms from all spread movements regardless of the potential reason for a change in spreads; instead the MA is designed to ensure firms should only take credit for future returns that are not due to retained risks.

6.27 Two respondents commented on the merits of using credit ratings as the key indicator of risk within the analysis of the FS. One respondent stated that many assets have been structured to have the same characteristics as corporate bonds, where structuring could also reduce risks with the credit enhancement reflected in the senior note’s rating. They added that the effect of rating lags may also already be reflected in transition and default statistics that underpin the basic FS. Another respondent noted that some ratings only consider the PD and not the expected loss taking recovery rates into account and that data limitations are typically taken into account in rating frameworks. The same respondent also stated that CRA consistent ratings could lead to a mapped FS that is too low for paired assets that are exposed to either counterparty failing, but reflect the worst rating of the two assets; and conversely these ratings could lead to a mapped FS that is too high for wrapped assets that are only exposed to both counterparties failing, but reflect the best rating of the underlying assets and the insurer. The same respondent sought clarity on what up-to-date ratings for liquid corporate bonds meant and also what the wording ‘all the risks are fully captured in the asset’s rating’ meant given the existing requirement to form an internal view of the credit rating of assets in the MA portfolio.^{footnote [19]}

6.28 Having considered these responses, the PRA has decided not to change the draft policy, noting that it is the responsibility of the attestor to determine whether the basic FS mapped from an asset’s rating captures all retained risks, based on the latest market and / or sector outlook and, for assets other than corporate bonds, the attestor’s knowledge of the rating methodology. The PRA notes that assets with different types of ratings at the same grade level may behave differently, have different risk characteristics or are subject to different levels of uncertainty; for example, some CRAs distinguish ratings for structured finance assets from fundamental ratings given on non-financial corporates, financials and public sector entities. The PRA also notes that where rating lags are already reflected in the underlying calibration data, the impact could be significantly masked by the fact that the FS is itself based on average statistics over a long period of time, whereas an up-to-date rating is required for the correct basic FS.

EVT for ERMs in attestation

6.29 Two respondents asked whether the current EVT for ERMs would be sufficient to meet the HDC MA attestation requirement, with one of these respondents suggesting that any additional expectations for ERMs would be unnecessarily onerous. In contrast, another respondent pointed out that the current EVT was only a diagnostic tool with minimum parameters set by the PRA and suggested an additional expectation for firms to use a version of the EVT with sufficiently prudent parameters for attestation purposes, with similar expectations for the EVT in stress to determine the need for any FS additions in stress.

6.30 Having considered these responses, the PRA has included a new paragraph 3.25A in SS3/17 to provide clarity around the use of the EVT. The PRA reminds firms that the current EVT with minimum parameters published by the PRA constitutes a test of plausibility enabling the PRA to direct its supervisory focus to the areas that are of most relevance to its objectives.^{footnote [20]} For attestation purposes, particularly given the focus on firms’ accountability for the level of MA benefit taken, the PRA expects firms to use their own assumptions that are judged to be appropriate for the HDC requirement of the MA, which should not fall below the PRA’s published minimum parameters where the EVT is used. Firms should additionally consider any retained risks other than the No Negative Equity Guarantee that are not assessed by the EVT. The PRA expects firms to take a similar approach in determining how any FS additions set in base would change in their SCR calculations.

FS additions in stress

6.31 Four respondents commented on the proposed expectation that changes to the FS in stress conditions should include any changes made to the FS used to calculate the TPs, including those made as part of the attestation process. Three respondents considered this expectation to be unduly burdensome and disproportionate to the risk management benefits and that it would result in increased complexity for credit risk in internal models when any FS additions were likely to be based on expert judgement. One of these respondents suggested that FS additions made in relation to attestation in base should remain unchanged in stress unless there were compelling reasons for these to vary. Another respondent suggested that FS additions made in base should not materially increase in the SCR in base conditions and not increase at all in the SCR in stress conditions. Another respondent sought clarification that a non-mechanistic, forward-looking view for the FS in stress was still expected, in line with attestation requirements, despite the basic FS being expressly based on a long-term backwards-looking calibration that must be consistent over time according to legislation and the PRA Rulebook. Finally, one other respondent commented on the implementation of any internal model changes potentially taking longer than the time available before the first attestation date.

6.32 Having considered these responses, the PRA has decided not to change the draft policy, that is, firms are still expected to consider how any FS additions would change in stress depending on what is already included in the SCR. The PRA does not have any prior expectation that base FS additions would necessarily increase in stress. The PRA recognises that firms might need to take a proportionate approach to assessing this issue, particularly until internal model changes (where required) have been made. The PRA confirms that it would still expect a non-mechanistic, forward-looking view of the FS in stress to be adopted by firms, independent of the new rules and expectations that have been proposed.

Principles-based guidance versus benchmarking

6.33 Three respondents supported the principles-based approach to attestation and supervisory guidance which provided the space for firms’ accountability for the level of MA benefit and which they preferred over a more prescriptive approach or more specific expectations on quantitative analysis that should be performed. One of these respondents commented that there was the potential for any regulatory thematic review to turn into a benchmarking exercise, which would disenfranchise the attestor; another respondent said that potential FS additions due to future regulatory influence would introduce uncertainty into the economic returns and hence asset sourcing and investment process, which could work against the PRA’s other reform proposals that would increase efficiency.

6.34 The PRA has noted the viewpoints from these respondents and considers that balance needs to be struck between applying regulatory scrutiny where required to ensure consistent interpretation of the supervisory guidance and a level playing field, versus allowing firms full flexibility to make their own judgments and assume accountability. Therefore, the PRA has decided not to change the draft policy in this area.

Miscellaneous suggestions

6.35 Five respondents commented on the expectation for firms to consider whether exposure concentration could lead to FS additions being expected as per paragraph 5.36A of SS7/18 (which was originally numbered 5.36 in the consultation draft of SS7/18).

Two of these respondents said that FS additions might not be the best mitigant against exposure concentration risk, which was already controlled via risk limits and already allowed for in the capital setting and rating methodology.
One of these respondents asked whether it was possible to have enhanced levels of diversification due to a broader investment base than that which was represented in the dataset underlying the basic FS.
Another respondent suggested that firms should be required to quantify the impact of a credit event affecting the most material MA contributors on a deterministic basis as opposed to all assets on a probabilistic basis, with an appropriate threshold applied to the difference between the two in a new matching test.
One respondent agreed that for concentrated corporate bond holdings, or within asset classes that were all exposed to similar risks, some consideration of risk diversification was necessary to ensure the calculation of the portfolio FS was sufficiently prudent.

6.36 Having considered the responses, the PRA has decided not to change the draft policy in this area. The PRA considers that limits on exposures are important mitigants against concentration risk, but any exposure that is more concentrated than the underlying calibration data still represents an increase in retained risks due to reduced diversification, which should be taken into account in considering the need for any FS additions. The PRA also reminds firms that the FS and the SCR serve different roles and, in particular, have different time horizons. The PRA considers that while it is possible for a portfolio to have a greater degree of diversification than the underlying calibration data, firms are not expected to reduce the basic FS, nor use this to offset against FS increases that are considered to be required for other assets. Finally, firms may use any reasonable approach to assess exposure to concentration risk, but the PRA considers this to be separate from matching tests which help demonstrate compliance with the relevant MA eligibility conditions. In particular, if a credit event did adversely impact a large proportion of the MA portfolio, a firm would be more protected if FS additions had been applied as a result of exposure concentration, but in any case, would still have the opportunity to restore matching.

6.37 One respondent commented on the FS being highly dependent on rebalancing cost assumptions which are firm-specific and may hence deviate from those assumed in the basic FS, even for corporate bonds. This respondent suggested that rebalancing costs should be explicitly captured as a retained risk within the FS, where FS additions may be required to allow for any transition / spread risk interactions. Another respondent sought clarification on how rating transitions and downgrade losses could be different from that assumed within the basic FS.

6.38 The PRA agrees that rebalancing costs that arise because of other retained risks are an important consideration for firms in determining the need for any FS additions and has incorporated these into paragraph 5.36A of SS7/18. Downgrade losses within the basic FS are derived based on deterministic spreads and the assumption that downgraded assets are immediately replaced to restore the original rating while preserving the original cash flow pattern, which may differ in practice depending on firms’ own rebalancing strategies and the market conditions under which rebalancing takes place.

6.39 One respondent suggested that an example of idiosyncratic risk for corporate bonds might be an asset with a maturity term that exceeds 30 years. Another respondent suggested that where ‘residual spread’ is used in paragraphs 5.37 and 5.38 of SS7/18, these should be replaced with ‘MA’, consistent with the rest of the SS. Another respondent suggested that rather than creating bespoke FS additions, it might be operationally simpler for firms to reduce their internal ratings, which would increase the basic FS through the usual mapping process. Finally, one respondent has identified a typographical error in paragraph 6.11 of CP19/23.

6.40 Having considered these responses, the PRA has incorporated the example of corporate bond idiosyncratic risk into paragraph 5.35 of SS7/18, but noting that the risks mentioned in the review process are not exhaustive. The PRA considers that in the context of paragraphs 5.37 and 5.38 of SS7/18, the residual spread is the quantity that firms should explain having gone through steps 1 and 2 of the example review process, with any further FS additions applied as necessary within this final step, in order for the resulting MA at the end of this process to be earned with a HDC. While internal rating methodologies may be amended if there were judged to be risks not included, or where their impact has been identified to be underestimated through the attestation process, the PRA considers that FS additions should be explicitly set out and documented in the attestation report in all other cases, rather than implicitly incorporated by artificially reducing internal ratings that map to a higher basic FS. Finally, the PRA confirms that in paragraph 6.11 of CP19/23, the reference to Chapter 17 of the Matching Adjustment Part of the PRA Rulebook should point to Rule 4.17 of the Matching Adjustment Part of the PRA Rulebook instead.

Attestation frequency and responsibility

6.41 The PRA proposed that attestations would be required from firms annually and where there has been a material change in the risk profile of the firm. The responsibility for the attestation would rest with the senior manager with the prescribed responsibility for the production and integrity of the firm’s financial information and its regulatory reporting (PR Q).

Out-of-cycle attestations

6.42 One respondent said that out-of-cycle attestations should rarely be required. The same respondent said that since the attestation policy would already set out the circumstances whereby an out-of-cycle attestation would be required, there would be no need to discuss with the PRA before concluding whether or not there had been a material change in risk profile that would trigger an out-of-cycle attestation.

6.43 Having considered these comments, the PRA has decided not to change the draft policy in this area. In a merger or acquisition, for example, there might be different attestors with different views of risks and approaches to supporting analysis. It is the relevant SMF with the attestation responsibility who has the accountability for the MA and who cannot rely on someone else’s attestation of the same MA portfolio.

6.44 However, the PRA accepts that there could be circumstances, despite a material change in risk profile, where an out-of-cycle attestation might be onerous or untimely. The PRA has amended the wording of paragraph 5.33 of SS7/18 to clarify the purpose of a firm entering into discussion with the PRA in the situation of an out-of-cycle attestation, which in addition to concluding whether or not there has been a material change in risk profile, now includes the agreement of the most appropriate date for the attestation reference date and the timeframe for the completion of the out-of-cycle attestation. The PRA has also amended the final rules to specify that for out-of-cycle attestations, the attestation reference date will be a date that is no later than three months after the material risk profile change. Where a firm considers that it would be appropriate to defer an out-of-cycle attestation until the next annual attestation, for example because the out-of-cycle attestation would otherwise be very close in time to the annual attestation, the firm should discuss with the PRA whether it would be appropriate for the firm to apply for a section 138A waiver.

Attestation responsibility

6.45 One respondent queried the need for SMF attestation since the Chief Actuary (SMF20) will effectively have to attest to the MA as part of their sign-off of the overall TPs.

6.46 Having considered this comment, the PRA has decided not to change the draft policy in this area. The PRA agrees that the work of the Chief Actuary in overseeing TPs is a component of a firm’s assurance on its financial soundness. However, TPs are driven by other demographic assumptions as much as the relevant risk-free rate (including the MA) that is used to discount the liabilities; whereas attestation purely focuses on the MA part of the relevant risk-free discount rate, earned from the assets side of the balance sheet. Therefore, the PRA considers that SMF MA attestation is a key wider step, in addition to the remit of the Chief Actuary’s role in the TP sign-off, in the firm taking accountability for its MA and FS decisions, as it affects the quality of own funds / capital.

Public disclosure requirement and external audit

6.47 The PRA proposed that disclosure would be required in the SFCR as to whether a firm has provided an attestation to the PRA, but with no expectation for the disclosure of the content of its attestation report nor the underlying evidence, both of which would also remain outside the scope of external audit.

Public disclosure requirement

6.48 Two respondents suggested that the SFCR statement on attestation should be in a section that is not subject to audit, such as the ’director’s responsibility’ or the ’system of governance’ section. One respondent asked if PRA had a view about a firm voluntarily disclosing the evidence underlying its attestation. Another respondent expected that firms might enhance their disclosures where appropriate to explain their approach to the attestation and setting FS additions for non-fixity.

6.49 The PRA considers that firms are free to disclose the statement on attestation in any suitable part of the SFCR, as well as the reasoning behind any attestation, or voluntary or mandatory FS additions, to the extent they wish to do so. While the PRA notes these views do not require a change to the draft policy, amendments have been made to the wording of paragraph 5.41 of SS7/18 to clarify that there is no requirement, as opposed to no expectation, from the PRA for a firm to publicly disclose the content of its attestation report.

External audit

6.50 Five respondents sought clarification over the scope of external audit in relation to the attestation material, including the rationale for this. One respondent asked whether the auditor could continue to rely on the published FS for any given rating and mapped credit quality step (CQS) and exclude the assessment of the need for, and quantum of, an FS addition from the audit scope. Two respondents noted the potentially highly significant judgements involved in firms’ assessment of the appropriateness of the basic FS and determination of any FS additions and the material incremental audit work that would be required in reviewing these compared to the current approach. One respondent commented that clarity would be required by early Q2 2024 to ensure a robust year end 2024 audit, if additional judgements involved in the MA calculations were within the audit scope, to allow time to plan, prepare and undertake this work.

6.51 The PRA considers that neither the attestation report nor the underlying evidence are within the scope of an external audit, as the attestation is directed to the PRA. In addition, consistent with the MA legislative framework, additions made to the FS to ensure it covers all risks retained, other than those arising from the uncertainty regarding the timing and amount of cash flows from assets with HP cash flows, remain at the discretion of firms and hence these need not be covered by an external audit. Therefore, the PRA considers that there is no change to the scope of external audit arising from either the attestation requirement,^{footnote [21]} or the option for firms to apply voluntary FS additions.^{footnote [22]} This is consistent with the use of the attestation measure and any related voluntary FS additions as a supervisory tool by the PRA, and the attestation material being excluded from any public disclosure requirement. While the PRA considers that no change to the draft policy is required in this area, the PRA has amended the wording of paragraph 5.41 of SS7/18 and added a new paragraph 4.2FA in SS11/16 to clarify these points.

Implementation

6.52 Eight respondents sought clarification as to the date of the first attestation, with some of these respondents querying whether there are any formal requirements before this date, including where a firm’s risk profile changes materially. Some of these respondents noted the significant time and effort required to make an attestation. Two of these respondents encouraged the PRA to allow a pragmatic initial approach and / or flexibility to agree feasible timelines for the first attestation report and one other respondent suggested that the attestation process could be streamlined and made proportionate as part of firms’ regular reporting cycle. One respondent suggested that the SMF making the attestation may want to have reviewed the MALIR data, which is not due until June 2025, as part of the attestation. Another respondent noted that ongoing engagement with the PRA would be valuable in all areas of reform as the PRA and the industry move towards implementation.

6.53 Having considered these responses, the PRA has decided to introduce a transitional period such that a firm may, but will not be required to, provide any annual or out-of-cycle attestations as set out in Rule 9.1(2) of the Matching Adjustment Part of the PRA Rulebook before 31 December 2024 – this has been incorporated into Rule 9.2 of the Matching Adjustment Part of the PRA Rulebook. This means that the first required attestation for any firm with MA permission prior to 31 December 2024 will be in respect of its first financial year end from 31 December 2024 (or material change in risk profile with an attestation reference date after 31 December 2024, if sooner). The introduction of the transitional period recognises the time that firms will need to put in place a policy, embed their processes and fully document an attestation. The PRA now expects that firms will make appropriate use of the transitional period to prepare for their first attestation following its expiry.

6.54 For the first attestation after the transitional period, the PRA also expects that firms will leverage existing analysis and processes and perform a focused and thorough review of at least the material MA contributors. As per Rules 10.3 and 12.3 of the Matching Adjustment Part of the PRA Rulebook, the PRA requires firms to put in place an attestation policy, which firms are expected to follow during its attestation process and which must be submitted to the PRA together with the first required attestation (and then must be resubmitted to the PRA with any subsequent attestation where changes have been made to the policy). Beyond the first attestation, the PRA recognises that the quality of the evidence used for attestation may improve over time, potentially with refinements to existing FS additions or new FS additions.

6.55 Firms may at their own discretion start to apply voluntary FS additions with effect from 30 June 2024, as these are not necessarily only for the purposes of attestation. The PRA notes the importance of ensuring that these voluntary FS additions have been determined using a rigorous approach, particularly in the context of firms’ attestations. Consequently, the PRA recognises that a firm’s processes in respect of such voluntary FS additions are likely to need to take account of the firm’s attestation processes, which are not required to be fully embedded until the first year-end from 31 December 2024 (aligned with the firm’s SFCR, which for most firms will be 31 December 2024). Furthermore, given the short time between the publication of this PS and the 30 June implementation date, the PRA also understands that firms are unlikely to be in a position to apply any voluntary FS additions to a 30 June 2024 balance sheet.

6.56 One respondent sought clarification as to whether there might be any potential conflict between the new MA rules and a firm’s existing MA permission, for example where the use of notched ratings and the removal of the SIG MA cap might require modification to the MA permission and how any conflict should be dealt with for attestation purposes.

6.57 Having considered this response, the PRA considers that firms will not require a modification to the MA permission for the removal of the SIG MA cap nor the introduction of notched ratings, unless these are coupled with other changes to the MA portfolio that would require an MA application. Hence there should be no conflict between the new MA rules and firms’ existing MA permissions that could affect attestation.

7: Assumptions underlying the MA

7.1 This chapter provides feedback to responses relating to the proposals in Chapter 7 (Assumptions underlying the MA) of CP19/23. It also contains the PRA’s final policy, set out in chapter 1A of SS7/18 – Solvency II: Matching adjustment (Appendix 3).

7.2 In CP19/23, the PRA proposed to insert a new chapter in SS7/18 that sets out the key conceptual and technical assumptions underlying the MA (and FS) and how firms would be expected to use and consider them.

7.3 The PRA also proposed that these assumptions be considered when determining if the risk profile of a firm deviates significantly from the assumptions underlying the MA, consistent with the existing Solvency II practice in respect of capital add-ons for the MA. The PRA explained that it would consult in due course on reflecting this proposal in its Statement of Policy: Solvency II: Capital add-ons^{footnote [23]} (‘Capital Add-On SoP’). The PRA is consulting on this proposal as part of CP5/24.

7.4 The PRA received 15 responses to Chapter 7 of CP19/23. Respondents generally welcomed the inclusion of these assumptions in the PRA’s policy materials, with one respondent specifically noting the overall objective of improving consistency and transparency.

7.5 Additional comments related to the consistency between the conceptual and technical assumptions, the application of the assumptions in practice and the detailed drafting. These, together with the PRA’s response, are set out later in this chapter.

Changes to draft policy

7.6 Following consideration of the respondents’ comments, the PRA has made amendments to Chapter 1A of SS7/18. A summary of the changes is set out below:

updates to refer (where appropriate) to the relevant regulation of the IRPR Regulations as well as the relevant PRA rules;
revisions to paragraph 1A.1 to give more background to the operation of the MA;
small amendments to paragraph 1A.4 to further clarify the intended meaning of the term ‘objective’ in the context of credit ratings as well as to the description, and intended use, of the assumptions underlying the PD and cost of downgrade (CoD) calculations; and
small drafting changes across the chapter to address other minor comments made by respondents.

7.7 There are no significant changes to the CBA or ‘have regards’ analysis arising from the changes to the draft policy proposed in CP19/23 in respect of this PS chapter.

Feedback to responses

7.8 The sections below have been structured along the same lines as Chapter 7 of the CP. The responses have been grouped as follows:

the assumptions underlying the MA;
use of the assumptions underlying the MA in practice; and
capital add-ons in respect of the MA.

The assumptions underlying the MA

7.9 The PRA proposed setting out the assumptions underlying the MA in one place to provide clarity to firms on what the PRA expects them to consider when reviewing consistency with such assumptions for example as part of the attestation process. The PRA also proposed splitting the assumptions into conceptual and technical categories.

7.10 A number of respondents considered it helpful to have the assumptions set out in one location, with one specifically commenting that they found the distinction between conceptual and technical categories useful. Some respondents also welcomed the PRA’s increased clarity in its interpretation and expectations around the use of the MA assumptions. However, there were a range of views on whether the PRA needed to say more around the assumptions as well as the operation of, and background to, the MA itself.

7.11 One respondent expressed a view that the MA is an artificial mechanism with a limited grounding in the principles of balance sheet accounting. The respondent considered that the MA was a mechanism to insulate insurers from certain risks (eg forced sale of assets) and, in light of this, considered it unclear as to why it should result in such a large balance sheet benefit. The respondent also questioned why the PRA was comfortable that all of the spread in excess of the FS can be assumed to be earned with full confidence. It was suggested that the PRA further justify the current form of the MA (including the level of benefit provided), giving due consideration to alternatives. The same respondent commented on the perceived incentives that they considered to be created by the MA, notably the relative attractiveness of investment in complex illiquid assets. In their view, large firms with permission to apply the MA were better placed to take advantage of these incentives relative to smaller firms with smaller MA portfolios. The respondent questioned whether this was consistent with the PRA’s secondary competition objective.

7.12 The PRA has considered this feedback and has decided not to change the draft policy. In CP19/23, the PRA’s articulation of the assumptions underlying the MA were designed to be consistent with the anticipated legislation (at the time CP19/23 was published) as set out in the November 2022 statement. The Government has since made and then laid the SI containing its IRPR Regulations. The PRA has reviewed this in detail and considers no change to the draft policy on MA assumptions, as set out in CP19/23, is needed. However, to aid use of the PRA’s policy materials, small changes have been made to SS7/18 to refer directly to the relevant legislation for each assumption (if appropriate), in addition to relevant PRA rules.

7.13 One respondent suggested that the description of the conceptual assumptions underlying the MA be expanded to better describe the fact that the MA framework does not encourage pro-cyclical investment behaviour. In a similar vein, two other respondents suggested that some of the recital text in the Omnibus II Directive (recitals 30 to 32 in particular) be restated in PRA policy.

7.14 Recitals 30 and 32 of the Omnibus II Directive cover the relevant risk-free rate and the VA respectively. The PRA therefore does not consider it appropriate for this text to be included in its policy on the MA (but notes that updates to PRA policy relevant to the risk-free rate and VA have been proposed in CP5/24). Additionally, in CP19/23 the proposed changes to SS7/18 relating to the conceptual assumptions underlying the MA already made direct reference to recital 31 of the Omnibus II Directive (paragraph 1A.3 – first bullet – see footnote 7). However, having considered the points raised, the PRA has updated paragraph 1A.1 of SS7/18 to provide further comment on the impact of the MA on insurers.

7.15 Two respondents highlighted that there may be inconsistencies between the conceptual assumptions underlying the MA and the technical assumptions, which underpin the calibration of the basic FS. One of these respondents noted a specific inconsistency they had considered, which may have implications for the attestation process. Their view was that the expected return for an investor intending to hold an asset to maturity would include the CRP. While they interpreted the conceptual assumptions to imply that this is a retained risk (and therefore should be included in the FS), the technical assumptions do not explicitly include a CRP for the purposes of the basic FS calibration. The respondent acknowledged that the basic FS may exceed expected credit losses, but in their view, there was no theoretical or empirical reason to suggest that this excess was a good estimate of the market compensation for the retained risks. The respondent suggested that the proposed wording in paragraph 1A.6(iv) of SS7/18 be amended accordingly and/or that paragraph 1A.3 of SS7/18 be deleted in its entirety.

7.16 The assumptions underlying the MA included in CP19/23 reflect the MA framework set out in the IRPR Regulations (as replicated in PRA rules), including the requirements for the calculation of the FS which the technical assumptions are consistent with. The PRA has considered the points raised and has decided not to change the draft policy other than the change set out above, to update its policy to refer directly to the relevant legislation where appropriate, in addition to relevant PRA rules. As described in Chapter 6, a possible three-step process to completing an MA attestation has also been provided in SS7/18. This process has been designed to take account of the conceptual and technical assumptions underlying the MA.

7.17 Three respondents commented on the proposed language in paragraph 1A.3 of SS7/18 that credit spreads can ‘be decomposed into two components’ (ie the MA and the FS). Two respondents made reference to the degree of judgement that would be required in decomposing the spread. Concern was also expressed that this language may imply that a particular approach is expected for completing the analysis needed to justify the level of MA. A request was also made for clarification on the granularity of the decomposition that the PRA was expecting given that a more granular decomposition could be justified, especially for individual assets or smaller groups of assets.

7.18 The PRA considered these responses and has decided not to change the draft policy. The specific language in SS7/18 does not indicate a particular analysis that the PRA expects firms to complete for the attestation or any other purpose. Instead, it refers to a firm’s ability to estimate the size of the component parts of the spread on assets it is holding in its MA portfolio.

7.19 Two respondents challenged the PRA’s proposed text in paragraph 1A.4 of SS7/18. This text described credit ratings as an ‘objective’ measure of risk. The respondents noted that rating methodologies involve the application of subjective judgements and were therefore concerned by the use of the word ‘objective’ in this context.

7.20 The PRA notes that a rating process can be subjective and requires the application of judgement. However, the MA framework in legislation assumes that the outcome of a credit rating process provides an objective measure of risk. The PRA has updated paragraph 1A.4 of SS7/18 to clarify this point.

7.21 Three respondents queried the proposed reference to ‘market value’ in paragraph 1A.4 of SS7/18, where the PRA had said that 30% of an asset’s market value can be considered recoverable on default. Two suggested that 'nominal value' would be more accurate in the context of this paragraph.

7.22 The use of ‘market value’ is consistent with the PRA’s calculation of the credit spread corresponding to both the PD and expected loss resulting from downgrading of an asset (ie the PD and CoD in bps). The calculation that the PRA is required to undertake is set out in Regulation 6(6)(a) of the IRPR Regulations.^{footnote [24]} To clarify further, the PRA has updated SS7/18 to explain the use of ‘market value’ in this context.

7.23 Two respondents queried the description in paragraph 1A.4 of the determination of expected downgrade losses in the CoD calculation. One respondent queried whether the rebalancing costs within the CoD calculation are based on full CQS differences in ratings or on notch-level differences. Another respondent queried the intended interaction/linkage between the CoD calibration assumptions and firms’ own investment strategies.

7.24 Having considered these responses, the PRA has updated paragraph 1A.4 of SS7/18 to improve clarity in respect of the CoD calculation. In particular, the text has been updated to make clear that the calculation of expected downgrade losses is based on downgrades and replacement of at least one CQS. Additionally, the PRA can confirm that the assumptions used to calibrate the CoD relate only to the calculation of the FS, rather than firms’ own investment strategies. However, as set out in paragraph 1A.6 of SS7/18, firms should consider any differences between the risk profile of their MA portfolio(s) and the assumptions underlying the MA, in order to identify any appropriate action.

7.25 One respondent sought clarity on the data underpinning the FS calibration, and another respondent on the meaning of the term ‘duration’ in the context of the assumptions underlying the MA.

7.26 The PRA has decided not to change the draft policy to reflect the above points. The PRA notes that the technical documentation issued by the European Insurance and Occupational Pensions Authority (EIOPA) contains the details of the data underpinning its calibration of FS values that it publishes for use by insurance undertakings in the EU.^{footnote [25]} The PRA considers EIOPA’s approach to calibration to be consistent with the technical assumptions underpinning the MA and so has continued to use this approach as the starting point for the Solvency II TI that the PRA itself publishes for use by insurance undertakings in the UK. In its TI SoP (see Appendix 9),^{footnote [26]} the PRA has set out the details of any differences between its approach and the methodology used by EIOPA. The PRA’s use of ‘duration’ in SS7/18 reflects the manner in which the FS is derived (ie using zero-coupon bonds – for which the duration and tenor will be identical) rather than the application of the FS to MA asset cash flows. The PRA’s use of this term is also consistent with the wording of the IRPR Regulations.

7.27 Additional wording changes were proposed by one respondent, relating to the description of the impact of the MA on an insurer’s balance sheet, the exposure of MA firms to certain risks and the separation of the credit spread into the MA and the FS.

7.28 Having considered the suggestions raised, where appropriate the PRA has made minor clarifications to paragraph 1A.3 of SS7/18. Where the PRA judged that the changes suggested by respondents would not achieve greater clarity, or would risk misunderstanding, they have not been reflected in SS7/18.

Use of the assumptions underlying the MA in practice

7.29 As set out in CP19/23, the PRA proposed clarifying that it expects firms to consider the conceptual and technical assumptions underlying the MA when considering how they comply with specific areas of the Solvency II requirements, noting specific circumstances as examples.

7.30 One respondent expressed concern that assets with HP cash flows would be automatically assumed to deviate from the assumptions underlying the MA and therefore would attract an FS addition. The respondent suggested that an FS addition should not be assumed to be always necessary for such assets and that firms should instead consider whether or not the cash flow variability would impact on the rating assessment to determine if an FS addition would be required for such assets.

7.31 Having considered this response, the PRA has decided not to change the draft policy. The PRA can confirm that even if an asset or assets were to have cash flows that met the definition of ‘highly predictable’ (as set out in Chapter 5 of the Matching Adjustment Part of the PRA Rulebook), the PRA rules and expectations for such assets would not apply where firms modify their treatment of these assets to allow them to be treated as akin to assets with fixed cash flows. Where an asset has cash flows that meet the definition of ‘highly predictable’ and is also treated as such, then the firm would be required to apply an FS addition in line with Chapter 8 of the Matching Adjustment Part of the PRA Rulebook. In both cases, the PRA would expect the firm to consider whether any further FS additions are needed (in line with Rule 4.17 of the Matching Adjustment Part of the PRA Rulebook) on a case-by-case basis.

7.32 One respondent commented on areas where, in their view, there was an element of prudence in the FS calculation.

7.33 The PRA has decided not to change the draft policy. Policy relating to the MA attestation (Chapter 6) provides further comment on the appropriate treatment of assets where a firm considers that the FS may be higher than needed to reflect the retained risks on an asset.

7.34 Three respondents identified potential inconsistencies between the assumptions underlying the MA and firms’ own approaches to managing MA portfolios and queried the impact of such differences in practice. Particular comments were made on the conceptual assumption relating to a ‘hold-to-maturity’ investment strategy.

7.35 The PRA has decided not to change the draft policy. As noted in the final bullet of paragraph 1A.3 of SS7/18, firms are expected to act in accordance with effective risk management practices. The assumptions underlying the MA are not intended to constrain firms' ability to act in this way, but instead provide a basis for firms to determine any differences between the risk profile of their MA portfolios and the MA assumptions.

7.36 Four respondents raised concerns relating to the applicability of the assumptions to all MA assets, including assets where lack of data may present challenges regarding the assessment for the potential need for an FS addition to be applied. Two respondents raised a specific comment relating to the applicability of the 30% recovery rate (ie the portion of an asset’s value that can be considered recoverable on default) to all MA assets.

7.37 The PRA has decided not to change the draft policy. The assumptions underlying the MA are likely to be most reflective of the risks associated with assets considered in the original derivation of the MA framework (ie corporate bonds). Further consideration may therefore be required for other assets. The PRA notes that, as set out in Chapter 2, assets categorised as having HP cash flows will be subject to an FS addition to allow for their additional risks. The assumptions underlying the MA are intended to be useful for firms in assessing whether or not differences between the risk profile of their MA portfolio holdings and the MA assumptions necessitate a particular action (including those noted in paragraph 1A.7 of SS7/18). For example, firms may wish to consider practicalities around recoveries and whether alternative recovery rates are more appropriate for certain assets as part of their MA attestation process.^{footnote [27]}

7.38 Two respondents commented on the PRA’s proposed expectation in paragraph 1A.7 of SS7/18 that a firm takes remedial action where differences exist between the assumptions underlying the MA and the risk profile of the firm’s MA portfolio. It was suggested that only material deviations should be considered appropriate for any further action and/or that firms may have different views to the PRA on the appropriate remedial action in certain circumstances.

7.39 Having considered the comments raised, the PRA has decided not to change the draft policy. Firms are expected to determine appropriate remedial action, considering the specific reasons for the deviation and its materiality. Any differences between the views of firms and the PRA in respect of appropriate remedial actions would be addressed as part of ongoing supervisory activity.

7.40 One respondent suggested that, rather than looking at deviations from the assumptions underlying the MA, an alternative approach to completing an MA attestation (ie assessing the ability to earn the MA with a high degree of confidence) could involve a detailed and rigorous credit assessment to assess the appropriateness of the allowance for credit risk.

7.41 The PRA notes that firms can use a range of approaches in order to complete the MA attestation. However, the PRA considers that an assessment against the assumptions underlying the MA is important in its own right and potentially informative as to areas where the FS/MA may be less appropriate for a given portfolio. The PRA has therefore decided not to change the draft policy in this area.

Capital add-ons in respect of the MA

7.42 The PRA did not propose any changes in its approach to capital add-ons in respect of the MA. For clarity, the PRA noted that it would expect a firm to consider the proposed conceptual and technical assumptions when determining if the risk profile of its MA portfolio deviates significantly from the assumptions underlying the MA. As noted in paragraph 7.3 above, the PRA is currently undertaking this consultation as part of CP5/24.

7.43 Three respondents commented on the PRA process for engaging with firms after a capital add-on has been applied. It was suggested that the PRA engages in an open dialogue with firms, considers the operational burden on firms involved in implementing a capital add-on and assesses appropriate review frequency and recalculation triggers of the capital add-on.

7.44 The PRA accepts these points as being important facets of the capital add-on process and considers that they are already incorporated in the PRA’s approach to capital add-ons – as set out in the Capital Add-On SoP.

7.45 One respondent sought clarification on whether the PRA would consider the level of spread on an asset or assets when calculating a capital add-on relating to the MA.

7.46 The PRA has decided not to change the draft policy. The PRA process with regard to capital add-ons will consider the specific circumstances in which the capital add-on is being applied and will engage with the affected firm. Given the range of circumstances in which a capital add-on may be merited, it is not possible for the PRA to provide a definitive list of all factors that may be relevant to a capital add-on calibration. However, the PRA does not expect that the prevailing level of spread on an asset will generally be a primary factor that is considered as part of the capital add-on process in respect of the MA.

7.47 Three respondents commented that capital add-ons may not provide an appropriate reaction to deviations from the assumptions underlying the MA. One respondent considered the existence of alternative safeguards such as the attestation and FS additions to mean that capital add-ons should not be necessary. Another expressed concern as to the effectiveness of capital add-ons for the MA given that the capital add-on would increase the SCR rather than adjust the TPs.

7.48 The PRA has considered these responses and has decided not to change the draft policy in these areas. The Solvency II framework already includes a provision for the PRA to set a capital add-on where there has been a significant risk profile deviation from the assumptions underlying the MA. Article 278 of the Commission Delegated Regulation (EU) 2015/35 has been restated in paragraphs 2.12 and 2.13 of the Capital Add-On SoP. The circumstances for setting capital add-ons in paragraph 2.1(e) of the Capital Add-On SoP largely replicate the circumstances set out in Article 37 of the Solvency II Directive.

8: Matching adjustment asset and liability information return data collection

8.1 This chapter provides feedback to responses relating to the proposals in Chapter 8 (Matching Adjustment Asset and Liability Information Return data collection) of CP19/23. It also contains the PRA’s final policy, as follows:

amendments to the Reporting Part of the PRA Rulebook (Appendix 2); and
updates to SS7/18: Solvency II – Matching adjustment (Appendix 3)

8.2 In Chapter 8 of CP19/23 the PRA proposed to:

introduce a new annual reporting requirement in the PRA rules for firms to provide portfolio metrics and detailed information on the assets and liabilities held in their MA portfolios;
introduce a new reporting template (the MALIR template (Appendix 11)); and
introduce a process that would allow firms to apply for a waiver from the requirement to submit a MALIR, or part thereof, in certain circumstances.

8.3 The PRA received 15 responses to Chapter 8 of CP19/23. Respondents generally welcomed the PRA’s proposals to formalise the MALIR and the added clarity that this gives regarding reporting requirements and timescales. However, several respondents said completing the MALIR would increase the burden on firms. The respondents specifically mentioned the new data fields (compared to the YE2022 ad hoc exercise) and the increased cashflow data. Respondents also made a number of other observations and requests for clarification. These responses are considered in more detail later in this chapter.

Changes to draft policy

8.4 After considering the responses, the PRA has made changes to the final MALIR instructions (Appendix 12) and template (Appendix 11). A summary of the changes is set out below:

the requirement for firms to submit cash flows extending beyond 50 years has been removed. Those cash flows will have to be discounted to the last month of the 50^th year;
substantive changes have been made to the asset type definitions for ‘Corporate Bonds’, ‘Covered Bonds’ and ‘Sale and Leaseback Loans on Commercial Properties’;
two new asset type definitions have been introduced to cover ‘Other Loans’ and ‘Other Sovereigns, Sub-sovereigns, Quasi-government/Supranationals’;
clarifications and minor updates have been made to several other asset type definitions to improve clarity;
‘Other/Unknown’ has been split into two separate drop-down options for the ‘Capacity Enhancing Assets’ field;
MALIR 2 – 2.8 and MALIR 2 – 2.9 in the proposed MALIR template set out in Appendix 9 of CP19/23 have been merged to be consistent with the Quantitative Reporting Templates (QRTs);
other minor changes have been made to improve clarity in some areas (including field names) and to update references to the QRTs as appropriate.

8.5 Any significant changes to the CBA arising from changes to the draft policy proposed in CP19/23 in respect of the MALIR are considered in Chapter 10 (Cost benefit analysis) of this PS. There are no material changes to the ‘have regards’ analysis.

Feedback to responses

8.6 The PRA has considered the responses received for Chapter 8 of CP19/23. Feedback to the responses has been grouped as follows:

the MALIR;
the MALIR template; and
a waiver process for the MALIR.

The MALIR

8.7 In CP19/23, the PRA proposed introducing a new rule requiring all firms with MA portfolios to submit a MALIR on an annual basis from year end 2024. Under the draft rule, separate MALIRs would be required to be submitted for each MA portfolio no later than 130 business days after a firm’s financial year end, in line with instructions and definitions that would be set out in the PRA Rulebook.

8.8 One respondent said the MALIR could be a barrier to firms making use of the MA, potentially reducing competition and product choice, and therefore negatively impacting the PRA’s primary and secondary objectives.

8.9 The PRA considers that firms should have this data available as part of good risk management and internal reporting, and firms with existing MA approvals will generally have provided similar data under the previous ad-hoc MALIR. Therefore, providing this data through the new MALIR should not present a significant additional burden. Furthermore, smaller firms or firms with more recent MA permissions would likely have fewer assets to include in the MALIR. Also, where firms with smaller MA portfolios are proposing to materially expand these portfolios in future, they would be able to build processes that are consistent with the MALIR at an earlier stage in their investment strategies. The PRA is committed to implementing the MALIR taking account of materiality and proportionality and notes that a waiver process is open to firms where the burden associated with the MALIR is considered to be disproportionate. Therefore, the PRA has decided not to change the draft policy.

8.10 One respondent asked the PRA to consider a later deadline for the annual MALIR submission in order to reduce the operational burden on firms. Two respondents stated that the proposed timescales are shorter than they first appear and that most work would need to be done during QRT reporting timescales. Another respondent was more supportive highlighting that the imposed deadline was 'very reasonable' and that a later deadline would not reduce the reporting burden as core resources would be tied up in MALIR work for a longer period. The same respondent noted a later deadline would also risk the data being out-of-date by the time it is received.

8.11 After considering the responses, the PRA has decided not to change the draft policy. The proposed deadline for the MALIR has been deliberately set after the deadline for other regular reporting in order to ease the operational burden on firms while balancing this against data relevance. The PRA considers that a later deadline would be of limited value as it would coincide with half-year reporting, thereby creating other potential areas of resource-stretch.

8.12 Two respondents asked the PRA to consider introducing a transition period before the MALIR becomes a requirement. One of those respondents proposed that the year end 2024 submission should be a dry run. The other asked that some additional flexibility for the first exercise be allowed as they would find it difficult to mobilise the operational capabilities to respond quickly to the final rules, especially to any major changes that might significantly deviate from the current consulted position.

8.13 The MALIR proposals consulted on were designed to minimise the changes relative to the last ad hoc request at year end 2022. Firms that participated in previous exercises should therefore be familiar with the process and the PRA does not consider it necessary to introduce a transition period or dry run. The PRA notes that submission of the year end 2024 MALIR will be more than one year from the date of this PS, giving firms at least six months to prepare. The PRA has therefore decided not to change the draft policy but firms with particular concerns around the year end 2024 MALIR should reach out to their usual supervisory contact to discuss.

The MALIR template

8.14 In CP19/23 the proposed MALIR template built on the template used for the last ad hoc data collection as at year end 2022, by proposing new fields primarily to collect data relevant to the MA reforms. The PRA further proposed that an excel-based template would continue to be used for the MALIR for a period of at least two years while work was undertaken to investigate the potential to move the MALIR collection to an alternative reporting interface.

8.15 Five respondents questioned whether all the data fields in the proposed MALIR template were needed, particularly the additional ones not previously seen by firms. Three of the respondents asked for data fields to be removed from the template if they cannot be justified, particularly in relation to the PRA’s statutory objectives. One further respondent challenged the need for the PRA to collect this data only for MA portfolios. This respondent also commented on the merits of some of the data fields and suggested certain fields may be reconsidered at a future review point. Among them, the respondents either questioned or listed the following fields (in particular) as requiring further justification:

Name of Internal Methodology, which seeks information on the rating method used when an asset is rated internally;
Notional value, which is the same as defined in the QRTs ie the amount outstanding on the asset measured at par value;
SME, covering whether an asset involves lending to either a small or medium enterprise;
Climate Target/Green, covering whether an asset is ‘green’ or contributes to a climate target (eg the UK Government’s net zero targets);
Capacity Enhancing Asset, covering whether the asset directly contributes to UK economic growth via the financing of increased capacity in both capital and labour stock and tangible and intangible assets in the economy; and
Primary/Secondary Investment, covering whether the asset is the purchase of a new bond issue/origination of a new loan or purchase of a bond or loan from another investor.

8.16 Having considered the responses, the PRA has decided not to change the draft policy for the following reasons:

the MALIR focusses on MA portfolios as this is an area with high anticipated growth (from the BPA pipeline) and it is where investment flexibility is expected to expand following implementation of the MA reforms;
the PRA considers that it is important to collect the data fields above to understand the nature of firms’ investments and also to gauge the extent to which firms' investments evolve to reflect the MA reforms. These fields are also important in the context of the PRA’s secondary competitiveness and growth objective; and
it is important that the MALIR gives a holistic picture of firms’ asset holdings in the MA portfolio.

8.17 One respondent further noted that not all firms had systems, processes and controls for the efficient provision of the data required in the new fields that were proposed in CP19/23. This would lead to higher costs than the CBA estimates if internal resource was not available and external support was used.

8.18 The PRA recognises that the introduction of new fields could increase costs for several firms. However, the PRA considers that the burden of supplying this information is outweighed by the benefits set out above. Furthermore, the PRA expects that firms will be able to automate the completion of the fields after the first MALIR exercise. To ease automation, the PRA has made a very minor change to split ‘Other/Unknown’ into two separate drop-down options for the ‘capacity enhancing assets’ field.

8.19 One respondent said the MALIR template would result in a large submission in terms of file size and number of data rows. They further explained that this impacted checking and validation due to the need for additional calculations. In particular, they noted that the proposed increase in the period for reporting asset cash flows from 50 to 100 years would exacerbate this significantly.

8.20 After considering the response, the PRA has decided to change its draft policy so that only detailed cash flows up to 50 years will be required. Firms will then be expected to discount cash flows extending beyond that point and report them at the last month of the 50^th year. This is consistent with the approach that was taken for the year end 2022 ad hoc exercise.

8.21 One respondent requested an additional data field to indicate the proportion of an asset that is in each component of the MA portfolio (A, B or C).^{footnote [28]} They considered this would help avoid an asset being entered multiple times, once for each component it is used in.

8.22 The PRA notes that firms take different approaches in structuring their MA portfolios. Some firms hold different assets in the three components of their MA portfolios while others use a proportionate split (at least for some assets). Adding a new field which indicates the proportion of an asset that is in each component of the MA portfolio could increase the burden on firms that do not use a proportionate split. It would also potentially be more difficult for both the PRA and firms to split the data between components thereby risking inconsistencies. The PRA has therefore decided not to change the draft policy.

8.23 Four respondents said that the MALIR template included duplicate fields already reported in the QRT S.06.02. One also highlighted that validation was not a good reason for the PRA to collect duplicate data, while another asked that derivatives should be removed from the scope of the MALIR.

8.24 After considering the responses and to avoid different approaches being taken in the QRTs and in the MALIR, the PRA has merged MALIR 2 – 2.8 and MALIR 2 – 2.9 in the proposed MALIR template set out in CP19/23. This change combines Asset ID Code and Asset ID Code Type in a single field and other fields have had their reference numbers updated accordingly. However, the PRA has decided not to change the draft policy beyond this. As stated in CP19/23, the PRA recognises that the MALIR template overlaps with S.06.02^{footnote [29]} which also collects individual asset data. However, the MALIR also collects data fields not included in S.06.02 (such as asset cash flows) and covers a wider range of assets (such as derivatives and reinsurance). The PRA considers it important that the MALIR collects asset data for MA portfolios on a holistic basis (ie including duplicates, derivatives and reinsurance) to preserve its overall integrity and to allow it to be used effectively as a key supervisory tool. While not the primary reason for retaining duplication, duplicate fields do assist in the validation of MALIR data. The PRA intends to keep the collection of overlapping data under review and will, if appropriate, seek to remove duplicates in the future. However, this will need to be balanced against the need for stability of the MALIR template referred to below.

8.25 Two respondents said asset type definitions should be clarified to ensure consistent submissions across the industry. One of the respondents said the PRA may wish to set out more prescriptive definitions of some asset types given the increase in profile of the MALIR from an ad hoc request to a formal regulatory submission. The same respondent suggested that examples may be useful to help firms interpret the definitions and to decide on edge cases.

8.26 After considering the responses, the PRA has clarified the definitions of asset types where most ambiguity may exist. While most of the changes are minor, the PRA has made substantive changes to two of the asset type definitions and has also added two additional asset types to improve the consistency and usability of the data. A summary of these four changes is as follows:

the definition of ‘Corporate Bonds’ has been updated to only include bonds and not loans and, to facilitate this, a new asset type called ‘Other Loans’ has been created;
a new asset type called ‘Other sovereign, Sub-sovereign, Quasi-government/Supranationals’ has been created to capture assets that do not strictly meet definitions of asset types already in the MALIR;
the ‘Covered Bonds’ definition has been widened to include covered bonds of all CQS and not just CQS0 and CQS1; and
the definition of ‘Sale and Leaseback Loans on Commercial Properties’ has been broadened to include income strips and finance leases. The name of the asset type has also been changed accordingly.

8.27 In addition to the changes above, the PRA will provide feedback directly to firms based on the year end 2022 ad hoc MALIR submissions. Rather than providing examples in the instructions, the PRA considers that feedback to firms will be the best approach to providing clarifications as this can take account of firms’ actual holdings and any questions can be addressed directly.

8.28 Four respondents asked the PRA not to make changes to the MALIR template going forward, to avoid further costs incurred by firms in implementing the proposed changes. One respondent asked that the PRA give firms at least 12 months' implementation period for future changes.

8.29 The PRA recognises the need for stability in the MALIR template. As mentioned in CP19/23, the PRA intends that the format and content of the MALIR will remain unchanged from year end 2024 unless an important, unanticipated data change is needed. Should any changes be necessary in the future, the PRA will give firms notice to implement the changes. The precise timescale will be related to the nature of the issue in question. For example, changes to avoid errors or misinterpretation of the instructions may need a lead time of less than 12 months if the issue needs to be addressed before the next MALIR submission.

8.30 One respondent questioned whether moving from Excel to a different interface after two years is necessary. The respondent noted that an alternative interface seemed like a costly and unnecessary step if the data were simply going to be transferred to another spreadsheet within the PRA for analysis.

8.31 As stated in CP19/23, the PRA does not intend to move from Excel-based submissions for two years while it assesses whether a change is necessary. This will be done with appropriate engagement with affected firms, which will be given sufficient time to implement any changes. The PRA considers it important to assess the ongoing appropriateness of how data submissions such as the MALIR are collected and that any improvements are made as appropriate.

8.32 One respondent asked that the PRA provide an unlocked version of the Excel template for the firm to be able to use when completing the MALIR.

8.33 The Excel template is locked to reduce the risk of errors in submissions. The PRA will consider individual requests for an unlocked version of the template on a case-by-case basis. Firms may approach their usual supervisory contact to discuss their specific case.

8.34 One respondent asked for confirmation that their approach to estimating asset level MA benefit and FS using approximations in the MALIR template was appropriate. Another respondent acknowledged that more than one approach could be used to allocate MA benefit to individual assets. The respondent further suggested that the PRA should standardise the approach firms take via feedback on the year end 2022 ad hoc MALIR submissions. They noted that consistency across firms would be crucial when determining limits on MA benefit for assets with HP cash flows.

8.35 After considering the responses, the PRA has decided not to change the draft policy. MALIR 2 – 2.45 and 2 – 2.47 of the MALIR instructions outlines the approach firms should take when calculating the FS (%) and MA benefit (£) respectively for each asset. These instructions state that the total amount of MA benefit across the MA portfolio should be consistent with that achieved in the calculation of TPs and should reconcile with the MA Benefit reported in QRT S.22.01.^{footnote [30]} If firms are unsure about the approximations they intend to make for these calculations, they should discuss this with their usual supervisory contact in the first instance.

8.36 One respondent asked the PRA to confirm that the MALIR would not be subject to audit. The PRA confirms that the MALIR will not be subject to audit, under these proposals.

8.37 In addition to the points made by respondents above, the PRA has taken this opportunity to update the instructions to the template to improve clarity in some areas (including field names) and to update references to the QRTs to reflect changes made in PS3/24.

A waiver process for the MALIR

8.38 In CP19/23, the PRA recognised that completion of the MALIR on an annual basis may be overly burdensome in some cases given the size of a firm or its MA portfolio. In light of this, the PRA proposed introducing a waiver process for the MALIR whereby a firm could, on an MA portfolio basis, apply for an exemption from the MALIR reporting requirements, or part thereof. This proposed process would be consistent with the PRA’s standard waiver and modification process under section138A (and would be on a case-by-case basis).

8.39 Two respondents asked the PRA to clarify what materiality meant in the context of waivers for the MALIR, how it would be determined and who would make the decision. One respondent asked the PRA to consider granting a waiver from the MALIR if an MA portfolio had low materiality for a firm regardless of whether the firm was large. In contrast, two respondents asked the PRA to clarify that materiality of an MA portfolio would not be in relation to the firm or its balance sheet. Instead, they asked that materiality be assessed based on the size of a given MA portfolio relative to the total size of MA portfolios for the industry as a whole such that an MA portfolio that is material to a small firm could be granted a waiver or be excluded from the scope of the MALIR. One respondent asked that a waiver be considered appropriate for an MA portfolio in cases where there was one or more other MA portfolios that held similar assets within the same firm.

8.40 After considering the responses, the PRA has decided not to change the draft policy. The PRA will consider applications for waivers on a case-by-case basis, in accordance with its usual practice, and it is not possible, or appropriate, to envisage all circumstances where a waiver may, or may not, be granted in advance. As stated in paragraph 8.23 of CP19/23 and confirmed by the amendments to paragraph 8.1A of SS7/18, the PRA will take into account a number of factors when making its decision (including but not limited to the materiality of the portfolio, size of firm, and the nature of the asset holdings in the portfolio). The relevant firm should seek to robustly support their application as to why they consider the requirement, for which they are seeking a waiver, to be unduly burdensome or not achieve the purpose for which the rules were made, in order to meet the statutory tests contained in section 138A, for a waiver to be appropriate.

9: Notching

9.1 This chapter provides feedback to responses relating to the proposals in Chapter 9 (Notching) of CP19/23. It also contains the PRA’s final policy, as follows:

amendments to the Matching Adjustment Part of the PRA Rulebook (Appendix 1);
updates to SS7/18 - Solvency II: Matching adjustment (Appendix 3);
updates to SS8/18 - Solvency II: Internal Models – modelling of the matching adjustment (Appendix 4); and
updates to SS3/17 - Solvency II: Illiquid unrated assets (Appendix 5).

9.2 In Chapter 9 of CP19/23 the PRA proposed to:

introduce a requirement that the FS applied by firms with an MA permission must reflect, where appropriate, differences in the credit quality of their assets by rating notch for the purposes of calculating their TPs. If this is not possible for some assets, firms would be required to take this into account in their MA attestation process;
require firms to derive a more granular FS by rating notch by linearly interpolating the TI published by the PRA for each relevant CQS;
introduce an expectation that firms justify any differences in the granularity at which the credit quality of their assets is reflected in their TPs and internal models used to calculate SCR, and sets out the factors the PRA would expect firms to consider when doing this;
introduce an expectation that interim remedies should be used, where a firm considers that its risk profile requires it to increase the granularity at which credit quality is reflected in its internal model, but that developing its model may take some time; and
introduce requirements and expectations for the purposes of assessing the appropriateness of firms’ internal credit assessments by rating notch, which are confirmed in Chapter 4 of this PS.

9.3 The PRA received 16 responses to Chapter 9 of CP19/23. Respondents generally welcomed the proposals but made several observations and requests for clarification which are set out later in this chapter. In particular, respondents asked for more clarity around timescales for implementing notching in the MA calculation for the purpose of calculating TPs. Respondents also asked when internal models would be expected to be updated (where applicable) and the timescales by which notched ratings would be expected to be available.

Changes to draft policy

9.4 After considering the responses, the PRA has:

introduced a new expectation in the first bullet point of paragraph 5.7A of SS7/18 to further explain what it means for a notched rating to be ‘available’; and
clarified that the requirement for the MA calculation to reflect notching will not apply until 31 December 2024, although firms will be able to voluntarily include notching in their MA calculations from 30 June 2024.

9.5 There are no significant changes to CBA or ‘have regards’ analysis arising from the changes to draft policy proposed in CP19/23 in respect of this PS chapter.

Feedback to responses

9.6 The sections below have been structured along the same lines as Chapter 9 of the CP. The responses have been grouped as follows:

mandatory application of a notched FS;
implementation of a notched FS in the TPs calculation;
differences in the granularity at which credit quality is reflected in TPs and internal models;
increasing the granularity at which credit quality is reflected in internal models – operational considerations; and
internal ratings and their validation.

Mandatory application of a notched FS

9.7 The PRA proposed that when calculating the MA for the purpose of their TPs, firms would be required to adjust the FS to reflect, where appropriate and possible, differences in the credit quality of exposures by rating notch. In the few cases where this is not possible, CP19/23 proposed that the firm must use the FS for the CQS to which the exposure is mapped. In these cases, the appropriateness of the FS and MA would then be explicitly considered as part of the attestation process as set out in Rule 6.5(2) of the Matching Adjustment Part of the PRA Rulebook.

9.8 Three respondents supported the PRA’s proposals for notching to be mandatory in the calculation of TPs, with one noting it should lead to better risk management and another saying it would be inappropriate for some firms to apply notching while others did not.

9.9 In light of these responses, the PRA will implement its proposal and make notching mandatory for the purpose of calculating TPs.

Implementation of a notched FS in the TPs calculation

9.10 For the purposes of the TPs calculation the PRA proposed that:

it would continue to publish TI in respect of the FS at the level of CQS;
the FS would be adjusted to reflect differences in credit quality by rating notch for assets mapped to CQS 1 – CQS 5 (inclusive) where such assets do not use the PRA-published tables for ‘Central Government and Central Bank bonds’;
the existing CQS level TI published by the PRA would be applied unadjusted to the middle rating notch within each CQS;
for the remaining rating notches, firms would apply linear interpolation to the published TI assuming that intermediate notches are evenly spread between consecutive CQS pairs; and
linear interpolation would be applied in respect of the PD component of the FS and at least the overall FS.

9.11 Most respondents supported the PRA’s proposed implementation approach for notching with broad recognition that it was pragmatic and achieved a balance between simplicity, transparency and technical robustness. One respondent specifically welcomed the clear requirements from the PRA for the use of notching in the base FS calculation, although it was recognised more generally that implementation of this approach would involve a moderate level of effort by most firms. Some respondents also noted that although the implementation of notching would likely increase the FS, it was a positive risk management step.

9.12 One respondent asked if the PRA would publish notched FS data as part of the TI it produced, rather than requiring firms to interpolate the FS tables. The respondent suggested that this would ensure consistency of application by firms. Another respondent echoed this preference for the PRA to publish notched data but in the context of wider comments they made on the PRA’s proposal for implementing notching as set out in CP19/23.

9.13 The PRA recognises that consistency of approach could be an issue if the implementation of notching were to rely on a complex modelling or interpolation framework. However, one of the reasons the PRA opted for linear interpolation was on the grounds of simplicity and transparency. The PRA therefore does not consider consistency of application to be an issue that is likely to emerge in practice and has decided not to change the draft policy in this area.

9.14 Another respondent argued that the proposed approach to introduce notching in CP19/23 was seriously flawed and it would be better to not introduce notching than to pursue this in practice for the following key reasons:

the shape of default probabilities by credit quality is more exponential than linear and use of linear interpolation could create a disjoint, for example between the CoD and any notch-level base transition matrix;
the approach proposed an excessively granular FS calibration by rating that could be detrimental to future evolution of the MA regime; and
the approach could create computational and other difficulties for internal models.

9.15 The same respondent set out a preference for the PRA to use a wholly different method for implementing notching based on bottom-up data-driven analysis, fewer CQS buckets, and updated CQS mappings that recognised the exponential progress of CQS bucket definitions.

9.16 The PRA recognises the points made by the respondent, with many of these having been discussed in the Notching SEG (NSEG). ^{footnote [31]} The PRA’s proposed approach for implementing notching was deliberately simplified to allow an increase in the FS-granularity while avoiding the data limitations and complexity associated with alternative approaches. The PRA has therefore decided not to change the draft policy. In more detail:

Data-driven approach: The NSEG considered possible methods for implementing notching and concluded that a data-driven, bottom-up approach would be a credible method. However, further investigations highlighted that there was insufficient data to do this robustly. It was observed that although some data by rating notch was available, it may be incomplete, have limited history or require the use of multiple sources, creating issues around consistency. These issues meant that either the input data, the outputs derived, or both, would likely require smoothing or other judgements to be applied in the calibration of the FS tables to facilitate consistency of outputs. In addition, it was noted that existing limitations arising from approximations within the FS calculation – particularly in respect of the Long-Term Average Spread (LTAS) element of the FS – would likely be exacerbated if moved to a more granular basis. Interpolation allowed the current FS calibration to be undisturbed, with granularity then being built around it. While the PRA recognises the theoretical arguments for a data-driven method, it considers this is not feasible in practice due to data limitations and so has decided not to change the draft policy to use interpolation for implementing notching.
Shape and use of interpolation: The PRA notes the respondent’s comments that the PD may exhibit a more exponential shape between CQS. However, the same is not necessarily true of the LTAS which dominates the FS in the majority of cases. The PRA therefore continues to consider that a linear progression – which also benefits from simplicity of approach – is preferable and has decided not to changes the draft policy set out in CP19/23. The PRA also notes the comments on possible disjoints that could arise between some FS parameters and other relevant assumptions. While interpolation is not without limitations, the PRA considers it is preferrable to a more detailed approach that would involve significant judgement and potentially give rise to material levels of spurious accuracy. The PRA has therefore decided not to change the draft policy based on linear interpolation that was set out in CP19/23.
Granularity of approach: The PRA recognises that granularity would be a key consideration if a data-driven approach was being used (including for derivation of an updated transition matrix). As set out in CP19/23, the PRA proposed an interpolation-based approach for implementing notching, for which issues relating to granularity were less pertinent. That said, the PRA explicitly excluded CQS buckets that are ordinarily not notched or where notching could introduce spurious accuracy from the scope of notching. The PRA has therefore decided not to change the draft policy in this area.
Internal models: The PRA notes the respondent’s comments in respect of internal models. In CP19/23, the PRA was deliberately not prescriptive as to how notching should be reflected in internal models with firms free to model at a level of granularity that they considered was appropriate to their risk profile. The PRA recognised that obtaining stressed data and developing appropriate modelling methodologies could present several challenges and so the proposed policy was intended to give sufficient flexibility to allow firms to navigate such challenges in a manner that was appropriate to their specific circumstances. The PRA has therefore decided not to make changes to the draft policy in this area.

9.17 Three respondents requested further clarity from the PRA regarding timelines for implementing notching in the TPs calculation.

9.18 The PRA recognises that implementation of notching will require firms to update their systems and processes to allow the MA to be calculated on a notched basis. In recognition of this, the PRA is changing the draft policy such that the requirement to reflect notching in the calculation of TPs will now be effective from 31 December 2024, rather than 30 June 2024 as was proposed in CP19/23. This change will not preclude firms from implementing notching from 30 June 2024 should they wish to do so and explaining this, as appropriate, in any relevant public disclosures. However, there will be no requirement for notching to be reflected in the calculation of TPs until 31 December 2024.

Differences in the granularity at which credit quality is reflected in TPs and internal models

9.19 The PRA proposed introducing an expectation that firms should justify any differences in the granularity at which credit quality is reflected in the FS for the purposes of calculating the TPs and the SCR. The PRA also proposed factors that firms should consider when justifying any differences in granularity, including availability and credibility of data and the modelling approach used.

9.20 Several respondents were supportive of the flexibility of approach that the PRA had proposed in this area and considered the points set out in SS8/18 to be sensible.

9.21 Five respondents said that the introduction of notching in base should not necessitate a change in firms' internal models. All five requested that the PRA be pragmatic and proportionate in determining if notching needed to be reflected in firms' internal models, particularly given the lower materiality (as a percentage of overall SCR) of such a change relative to other aspects of firms’ MA in stress modelling. One respondent specifically asked that the level of evidence (initial and ongoing) needed to justify any differences in granularity should not be onerous. Another respondent implicitly made the same point on evidence and raised a concern about the potential for increased complexity in internal models. They suggested that the list of considerations in paragraph 4.21A of SS8/18 was too long and should focus instead on the impact of notching on the base balance sheet and ensuring the modelling approach used by the firm did not lead to any increase in FS from notching in base being automatically backed out in the SCR.

9.22 In CP19/23, the PRA did not propose to require firms to model the FS on a notched basis in the SCR. Paragraph 4.21A of SS8/18 provided firms with a framework that they could use to determine if they needed to reflect notching in their internal models and the justification that they would need if they concluded that this was not required. The PRA is supportive of firms being proportionate in this area, provided the internal model meets the relevant requirements and calibration standards. The PRA has also considered the points made on the extensiveness of the list of considerations in paragraph 4.21A and considers that having a more extensive list is helpful given certain considerations may be more pertinent for some firms than others. The PRA has therefore decided not to change the draft policy.

9.23 Six respondents asked that firms should be able to use an appropriate and pragmatic method of their choosing when updating their internal models to allow for notching. Of the respondents, five favoured being able to use interpolation in some form and two expressed strong concerns over the feasibility of detailed bottom-up modelling (eg through implementation of a notched spreads and notched transition calibration), with one explicitly asking that firms were not expected to undertake such an approach.

9.24 The PRA deliberately did not propose a particular methodological approach for use in internal models in CP19/23. Firms can adopt an approach of their choosing, including an approach based on linear interpolation of the stressed FS, provided they can demonstrate that this meets the internal model requirements and calibration standards. The PRA has therefore decided not to change the draft policy.

9.25 In relation to the proposed paragraph 4.35A of SS8/18, one respondent stated that it might not be appropriate to adjust all historical transition data where there were low volumes of historical data.

9.26 The PRA confirms that paragraph 4.35A was not intended to require firms to adjust transition data as a matter of course. It was instead alerting firms to consider the potential lack of historical transition data and the associated implications when developing/validating their models. The PRA has therefore decided not to change the draft policy.

9.27 One respondent suggested that paragraph 2.2 of SS8/18 should be moved to SS7/18 or deleted (if it is already incorporated in SS7/18) due to its focus on the base balance sheet.

9.28 Paragraph 2.2 of SS8/18 was intended to provide context. The PRA acknowledges that this paragraph talked about the FS/MA in base. However, the PRA considers it important to clearly set out the base FS/MA calculation at the start of SS8/18 as this is what firms are then required to stress. The PRA has therefore decided not to change the draft policy.

9.29 One respondent asked that simplifications in internal models – including using a whole-letter rather than a notched approach for stresses – should not automatically lead to a capital add-on.

9.30 In CP19/23, the PRA proposed that where a firm considered that its risk profile required it to increase the granularity at which credit quality is reflected in its internal model, it would need to develop a methodology that met the relevant internal model requirements. In CP19/23 the PRA proposed that if developing a model was not straightforward for a firm then it should consider other possible remedies in the interim. The PRA did not intend to imply that an addition to the SCR or the imposition of a capital add-on would be automatically required. However, the PRA recognised that an increase may be required to the internal model SCR if it did not meet the internal model calibration standards. This supported the key point that firms need to ensure their internal models were giving appropriate capital outcomes, including where simplifications were used. Remedies where an internal model did not meet the calibration standards could include an addition to the SCR or a capital add-on. Given this, the PRA has decided not to change the draft policy.

9.31 One respondent stated that, if notching is mandatory in base, then it should be mandatory in stress. However, this was predicated on the approach in base being changed as they considered the proposed approach to be materially flawed.

9.32 The PRA considers that, regardless of the method used in base, it would be unduly burdensome for the level of prudential risk to mandate the use of notching in firms’ internal models. Also, it would be inconsistent with how internal models are intended to operate, where it is for firms to justify that their approach meets the relevant internal model requirements and calibration standards, and not for the PRA to prescribe a specific approach that must be used. The PRA has therefore decided not to change the draft policy.

Increasing the granularity at which credit quality is reflected in internal models – operational considerations

9.33 The PRA proposed that if updating their internal models to reflect notched ratings, firms should consider, in addition to existing considerations set out in SS8/18, basis risk when comparing their own portfolios to the calibration data set and granularity of historic transition data. The PRA also proposed that, where developing a model is not straightforward, the firm would be expected to consider other possible remedies until it had completed the necessary development, including potentially increasing the capital requirement calculated by the internal model, in order to ensure that the SCR complied with the core calibration standards at all times.

9.34 One respondent specifically asked for clarity as to when notching would be expected to be included in internal models. Along with one additional respondent, they also asked that a reasonable transition period was allowed before imposing any capital add-on in respect of potential basis risk caused by inconsistencies in rating granularity between the base and stressed MA calculation. Two respondents also asked that proportionality and materiality were taken into account when determining if a capital add-on was required while model development was ongoing with a different respondent asking how any increase in capital should be determined and another suggesting an interpolation approach could be useful in setting a capital add-on.

9.35 The PRA deliberately did not set a timescale for reflecting notching in the internal model when firms considered this was necessary in order to continue to meet the internal model requirements. This flexibility would allow firms to update their models (if required) in line with their own priorities and model development plans. In determining if a capital add-on is required in the interim, it will be necessary for the PRA to consider if there is a significant deviation in the risk profile of a firm from the assumptions underlying the SCR. Further details around capital add-ons can be found on the PRA website^{footnote [32]} which will be updated to reflect near final rules and policy set out in PS2/24.

9.36 One respondent stated they would not expect changes to reflect notching in TPs or internal models to necessitate a new MA application or a model change application respectively.

9.37 The PRA notes that whether any changes made to reflect notching require a new MA application and/or a major model change application, will depend on the nature and extent of any changes being made by the firm. The PRA would not expect a new MA application to be required unless the firm was making changes beyond updates to its MA calculation and ratings processes to reflect notching. In the case of internal models, any changes should be assessed by the firm against its model change policy. If there is any uncertainty, firms are encouraged to reach out to their usual supervisory contact to discuss.

Internal ratings and their validation

9.38 Chapter 4 of this PS covers credit ratings under the MA. This includes responses and feedback in respect of proposed requirements and expectations set out in CP19/23 for the purposes of assessing the appropriateness of internal credit assessments by rating notch. The proposals in CP19/23 recognised that internal credit assessments were heavily reliant on expert judgement, and that validating such assessments on a notched basis could be challenging. The PRA therefore focussed its expectations in this regard on assessing potential bias in internal credit assessment outcomes (relative to CRA issue ratings) rather than asset by asset outcomes in isolation.

9.39 The PRA also noted in CP19/23 that several firms would likely have to enhance their current internal credit assessment processes to allow them to produce internal ratings on a notched basis and to comply with paragraph 2.5A in SS3/17. The proposal to not specify a granular approach to internal rating validation (ie focussing on overall bias rather than individual rating differences) and the PRA’s recognition of the challenges with validating internal credit assessments at a notched level were welcomed by respondents. However, it was noted by one respondent that rating on a notched basis could require significant effort from firms.

9.40 Two respondents specifically commented on time being needed to obtain notched internal ratings, with one requesting a pragmatic approach was taken in respect of assets that are more difficult to rate on a notched basis (eg restructured Lifetime mortgages).

9.41 Having considered these responses, the PRA has updated its expectations to state that firms have six months from the date at which an asset first becomes an ‘assigned asset’^{footnote [33]} for a notched rating to become ‘available’. For avoidance of doubt this means that any ‘assigned asset’ in the MA portfolio on 30 June 2024 would be expected to have a notched rating available from 31 December 2024. If an asset becomes an ‘assigned asset’ after 30 June 2024 then a notched rating would be expected to be available, no later than six months from the date when the asset became an ‘assigned asset’. The PRA has also expanded the first bullet point of paragraph 5.7A of SS7/18 to set out that, in cases where a notched rating is not ‘available’ within six months, firms should be able to explain to the PRA why they are unable to rate the asset(s) on a notched basis.

9.42 Two further responses covered points consulted in Chapter 4 of CP19/23 on Credit Ratings and the MA.

One respondent expressed concerns about the PRA's proposal (outlined in proposed updates to SS3/17) for firms to develop a validation framework for internal credit assessments. The respondent considered that implementing this framework would require them to make amendments to their existing internal rating and validation frameworks, resulting in a significant increase in governance costs.
The same respondent urged the PRA to provide exemptions from the requirement to obtain independent external assurance on internal ratings for firms that hold a small volume of private credit or have a relatively small MA portfolio. The respondent argued that by considering these exemptions, the PRA could reduce the regulatory burden on firms while still maintaining appropriate risk management practices.

9.43 The PRA notes that these points relate more generally to credit ratings and go beyond considerations specifically in respect of notching. In light of this, the PRA has addressed these points in Chapter 4 of this PS, together with other comments received on credit ratings.

10: Cost benefit analysis

10.1 This chapter provides feedback to responses relating to Chapter 10 (Cost benefit analysis) of CP19/23. It also contains the PRA’s final CBA resulting from final policy set out in the earlier chapters of this PS.

10.2 In CP19/23, the PRA set out that, within the framework of the legislation, the proposals would advance the PRA’s primary objectives of safety and soundness and policyholder protection. The PRA considered the proposals would do this through:

enhanced senior management responsibility and improved management of MA portfolios and their inherent risks, resulting from the attestation process and requirement to demonstrate compliance with the PPP;
greater assurance from the attestation process that the FS reflects all risks retained by firms and that the MA can be earned with high degree of confidence;
ensuring that the additional risks from assets with HP cash flows are allowed for appropriately within the MA;
maintaining the quality of matching between asset and liability cash flows through the introduction of additional controls for assets with HP cash flows; and
improved monitoring and supervision of firms by the PRA, in turn helping to realise the benefits of other policy proposals.

10.3 The PRA considered the proposals would facilitate effective competition, international competitiveness and growth, through:

a contribution to a more level playing field resulting from improved consistency of approach between firms;
reduced barriers to investment as a result of the proposed streamlined MA application approach for suitable assets;
greater clarity about how the investment flexibilities will be applied in practice, which is a key part of facilitating productive investment and supporting medium-to-long-term growth; and
improved incentives for private insurance and saving provision resulting from a broader range of liabilities being MA eligible.

10.4 The PRA received 13 responses to Chapter 10 of CP19/23. Respondents made a number of observations which are set out later in this chapter. Key themes from responses included implementation and ongoing cost estimates being potentially underestimated; costs being considered disproportionate for firms with smaller MA portfolios and the impact of proposals on the PRA’s primary and secondary objectives.

Changes to draft policy that impact the CBA

10.5 Having considered the responses received to CP19/23, the PRA has made a number of policy changes, as described elsewhere in this PS. The PRA considers that the following changes to policy relative to that in CP19/23 reduce the burden on firms and/or give rise to additional benefits:

simplifications to the analysis process expected for corporate bonds within the MA attestation requirements;
revised calibration for Matching Tests 4 and 5 applicable to assets with HP cash flows;
the expansion of MA liability eligibility criteria to include in-payment GDAs;
the commitment by the PRA to monitor and publish regular reports on the MA framework covering application review timelines and decision rates;
changes to the MALIR template to improve the consistency and usability of the data including removal of cashflow projections beyond 50 years, changes to asset type definitions and greater alignment with the QRTs; and
other simplifications and clarifications to assist with completion of the template including updates to field references.

10.6 The PRA considers that the policy to include mandatory FS additions for assets with HP cash flows within the scope of external audit may give rise to additional costs for firms.

10.7 Overall, the PRA does not consider that the policy changes are significant enough to revise the CBA, and the CBA within CP19/23 remains valid.

Feedback to responses

10.8 The PRA has considered the responses to Chapter 10 of CP19/23. Feedback to the responses has been grouped into the following broad headings:

CBA baseline;
assessment of costs and benefits;
benefits;
costs;
impact on asset allocation and annuity pricing; and
CBA summary.

CBA baseline

10.9 The PRA set out the baseline for the CBA in CP19/23 as being the current onshored legislative framework as supplemented by PRA Rulebook material in force, together with the anticipated (at the time of the CP publication) legislation in line with the November 2022 statement. The IRPR Regulations were made and then laid by Government in December 2023, after the CP was published, and were in line with the approach anticipated by the PRA at the time of publication.

10.10 Five respondents to Chapter 2 considered that the thresholds on the additional matching tests for firms holding assets with HP cash flows were too low.

10.11 Having considered the response, the PRA notes that its baseline assessment in CP19/23 considered that firms would be closely cash flow matched (ie that firms would already have sufficient capacity to facilitate investment in assets with HP cash flows). The PRA has reviewed this position and recognised that firms are typically less well cash flow matched than assumed in the baseline for the CBA in CP19/23. Therefore, the PRA has updated its view of the baseline, and correspondingly recalibrated the proposed matching tests thresholds to ensure that sufficient capacity is afforded to firms to facilitate investment in assets with HP cash flows, per the intent of the original policy proposal (see Chapter 2 of this PS for further details).

Assessment of costs and benefits

10.12 In CP19/23, the PRA assessed the costs and benefits of its proposals against its CBA baseline, including quantitative estimates of costs where possible using information obtained from firms. The PRA considered that the benefits of its proposals outweighed the costs.

10.13 The PRA received the following responses regarding the assessment of costs and benefits in the CBA.

Two respondents agreed that the benefits outweigh the costs. One other respondent considered that while there were flaws in the PRA’s assessment, the cost estimates were not so wrong as to invalidate the CBA.
One respondent asked that the PRA publish a quantitative view of the appropriate level of policyholder protection that should be achieved by the MA framework (ie a baseline) and assess the impact of the proposals in the CP against this baseline. The same respondent considered that the PRA assumed that different policyholder protection tools were interchangeable. The respondent considered that more work was needed to assess this assumption and assure the public that the revised balance would secure an appropriate degree of protection, including in stressed circumstances.

10.14 Having considered the responses, the PRA notes that its approach to advancing policyholder protection operates through a range of factors which cover different dimensions of policyholder protection, as set out in the PRA’s approach to insurance supervision.^{footnote [34]} Additionally, the PRA notes that the tools the PRA uses to achieve an appropriate degree of protection for policyholders may vary depending on a range of factors, many of which may not be quantitative in nature. The PRA considers that some of its policy tools may be complementary or affect more than one dimension of policyholder protection, and therefore may not be exchangeable for others. The PRA may use a suite of policy tools depending on the circumstances which, in aggregate and within the legislative framework, is considered sufficient to secure an appropriate level of policyholder protection. Therefore, the PRA has decided not to change its CBA approach.

Benefits

10.15 In CP19/23 the PRA set out its assessment of the benefits of the proposals. In addition to the benefits set out in paragraphs 10.2 and 10.3 above, the PRA considered that the key benefits of the proposals were to:

improve clarity of expectations on the calculation of best estimate cash flows, the methodology for determining the FS addition for assets with HP cash flows, the proposed controls to the quality of matching for assets with HP cash flows and the implementation of a more granular FS to reflect differences in credit quality through notching;
improve consistency across firms’ approaches once embedded, and reduce supervisory burden in the longer term through a proportionate and more efficient focusing of resources, improved risk sensitivity from mandatory notched credit ratings and a regular, structured approach to MALIR data collection;
reduce some capital costs for firms through, among other things, proportionate regulatory treatment of breaches of MA conditions; and
support greater diversification of asset portfolios as the proposals target a broader range of asset classes.

10.16 The PRA received the following responses on the benefits of the proposals.

Two respondents considered that the benefits had been overstated, with one of the respondents highlighting specific benefits on which they disagreed with the PRA. Areas of disagreement related to the investment flexibility and MA attestation proposals. These included improved matching between asset and liability cash flows relative to the baseline, reduced barriers to investment as a result of the streamlined MA application approach for suitable assets, firms having improved confidence in making investment decisions, and increased transparency and accountability resulting from the attestation proposals.
One respondent disagreed with the application of the FCA study mentioned in CP19/23^{footnote [35]} in respect of the attestation process raising risk management standards.
Another respondent agreed that the MA attestation requirement should result in greater assurance over the sufficiency of the FS and quality of the MA.

10.17 The PRA has reviewed these responses and expects that the implementation of the MA attestation requirement for firms will contribute to the improvement of the overall risk management of their MA portfolios (which would be of particular benefit given the wider range of assets that firms can include in their MA portfolios). While some supervisory processes already apply to MA portfolios, when these are combined with the attestation process the PRA considers this will improve firms’ transparency and accountability in this area.

10.18 Relative to the initial CBA, some of the final policy areas will increase benefits to firms through:

reduced implementation and ongoing compliance costs relating to the MA attestation requirement due to the simplified analysis process expected for corporate bonds, relative to the initial CBA. This may have a beneficial impact on firms whose MA portfolios are predominantly invested in less complex assets, including mutual firms. However, given the indicative nature of the cost estimate that was included in the initial CBA, the PRA considers that any refinement to the estimate would not be meaningful;
potential for firms to use capital more efficiently with the inclusion of GDAs within MA eligible liabilities, along with a small reduction in best estimate liabilities;
improved transparency through the PRA’s commitment to monitor MA application review times;
additional investments into assets with HP cash flows for some firms as a result of the revised calibration for Matching Tests 4 and 5; and
halving the number of cash flow fields in MALIR by removing the requirement for firms to submit detailed cash flows beyond 50 years, avoiding the file size of MALIR submissions becoming potentially unmanageable.

10.19 The PRA does not consider that the policy changes are significant enough to revise the CBA and has concluded that the CBA within CP19/23 remains valid.

Costs

10.20 The PRA estimated total ongoing costs from year 1 ranging between £7 million and £9 million each year, with an additional £2 million to £3 million for implementation costs across all 19 firms with existing MA approvals, excluding one-off cost estimates or those that apply at an asset level.

10.21 The PRA received the following responses on cost estimates.

Although one respondent noted the firm level cost estimates did not look unreasonable, seven respondents expressed concern that the implementation and ongoing cost estimates for firms may have been underestimated. Six of these respondents highlighted specific policy areas that were expected to incur higher costs than estimated, including MA attestations, MALIR and implementation decisions.
Two respondents provided quantitative evidence as part of their response, which covered cost estimates for MA attestation, MALIR and notching policy proposals.
Three respondents noted that information on implementation considerations would be useful for assessing firm costs.
Four respondents were concerned that costs may be disproportionate for firms with smaller MA portfolios. One respondent also questioned the cost impact for mutuals.

10.22 Having considered the responses, the PRA has made changes to the MA attestation and MALIR requirements as set out above, but the PRA considers its initial CBA remains valid.

In respect of attestation, the PRA considers that it is an important supervisory measure that is consistent with the package of Solvency II reforms as set out in the November 2022 statement. The PRA also considers that the benefits of the proposed policy on attestations (as adjusted through the changes in this PS) continue to exceed the costs given the significant impact that the MA has on firms’ solvency positions. The PRA agrees that firm implementation costs will depend upon their existing risk management processes but considers that the attestation work (and costs) will settle down to a steady level after the introduction phase.
The PRA notes that this PS along with the April statement sets out further details on the implementation of the final policy.
The PRA recognises that, consistent with the PRA’s approach to proportionate supervision, there may be situations, usually on grounds of proportionality or impracticability, where relevant reasons and supporting evidence may be provided to support a waiver or modification of the rules, which would be considered by the PRA on a case-by-case basis.

10.23 One respondent suggested there would be additional costs arising due to external audit fees relating to judgements underpinning FS additions and the appropriateness of the MA. The PRA notes that the inclusion of mandatory FS additions for assets with HP cash flows within the scope of external audit will increase costs to firms relative to the initial CBA. The PRA considers that the additional audit work required will be proportionate to firms’ exposure to assets with HP cash flows, and that the marginal cost to firms will be low when assessed against the audit work currently performed. The PRA considers the initial CBA remains valid, noting that voluntary FS additions are not within the scope of external audit.

10.24 One respondent noted that they could not see any allowance for implementation costs for new applicants for the MA and queried whether this was out of scope of the CBA. The PRA notes that firms that do not already have MA approval will not incur any additional costs from the new regime. Instead, new applicant firms will benefit from the new regime. Proposed changes to the application process would lead to reduced costs relative to the baseline (if their assets/liabilities would currently be eligible). Otherwise, the new regime offers the opportunity for some applicants to apply for MA permission where that would not previously have been possible – an option that firms could take up if it is beneficial to them. The PRA has determined that no change to the CBA is required.

10.25 One respondent considered that the PRA’s supervisory cost estimates may have been underestimated. The respondent suggested that an underestimated level of supervisory effort could increase the risk of inadequate supervision and poor regulatory outcomes, together with a further slowdown in approval processes and delays to firms’ realisation of the benefits of the reforms.

10.26 Since the publication of CP19/23, the PRA has further reviewed its resourcing needs as it has further developed its intentions on the implementation of the proposals. The PRA considers that the original CBA, including the supervisory cost estimates, remain valid, noting that resources are expected to be re-directed based on supervisory priorities and focus.

Impact on asset allocation and annuity pricing

10.27 One respondent commented that the attestation requirements were significantly onerous and would discourage firms from diversifying their asset risk profile and new productive asset investment. Another respondent agreed that greater investment flexibility would support greater diversification for asset holdings within MA portfolios. Another respondent expressed concern that compliance costs would wipe out any potential reductions in the cost of bulk or individual annuities.

10.28 The PRA notes that the proposals around attestations set a higher bar for the analysis expected of firms who wish to include in their MA portfolios assets whose characteristics differ from the assets on which the basic FS has been calibrated. However, given the significant benefit the MA provides to firms, the PRA considers it appropriate that firms should be accountable for demonstrating that the higher MA benefit available on such assets can still be earned with a HDC. The PRA considers that there is no inherent bias in terms of the onerousness of the attestation requirements against productive assets versus any other illiquid asset class and considers that the responses to the CP did not provide evidence that a bias did exist. The PRA notes that bulk annuity prices depend on a range of factors, some of which may have offsetting impacts. The potential benefits highlighted in the CBA arise as a result of the broader range of MA eligible assets under the draft proposals.

CBA summary

10.29 The PRA received the following responses on the PRA’s consideration of its secondary objectives in its CBA.

Five respondents expressed concern about the impact of the proposed package of reforms on the PRA's secondary objectives.
Two respondents expressed concern that the proposals could have an adverse impact on the PRA's secondary competition objective. Specifically for attestation, one respondent argued that standardised wording for attestations would not lead to a ‘level playing field’, given larger firms with more resources would be able to do more analysis and justify a higher MA benefit. Furthermore, an asset with a high MA benefit in a smaller MA portfolio consisting of other less complex holdings could be considered as one with an average MA benefit in a larger MA portfolio consisting of more mature illiquid asset holdings. The respondent commented that this could lead to firms attesting with the same wording, but with materially differing underlying analysis and conclusion around the resulting MA benefit.
Two respondents expressed concern that the proposals would not advance the PRA’s secondary objectives.
One respondent considered that compliance costs could be a significant barrier to new entrants, and firms looking to offer innovative products and services.

10.30 Having considered the responses, the PRA has decided not to alter its CBA. The PRA considers that by adapting the MA rules for the features of insurance business in the UK and the financing demands of the wider economy, the proposals, will allow the life insurance sector to play a bigger role in productive investment in the UK economy, while continuing to offer their policyholders the level of security determined by legislation. This will advance the secondary competitiveness and growth objective in a sustainable way.

10.31 The PRA considers that the attestation proposals would reduce the risk of an inappropriate FS being applied, contributing to a level playing field. The PRA also notes that within the attestation analysis, since the reference benchmark for determining whether an illiquid asset is an outlier is expected to be an equivalent corporate bond (as opposed to other illiquid assets that a firm holds), this ensures some degree of consistency in the attestation requirement for different firms. In addition to this, the proportionate implementation of MALIR, along with it being required of all firms, will ensure firms are treated consistently helping to facilitate effective competition. All of the above aspects were considered in the CBA.

10.32 One respondent noted that the proposals did not sufficiently balance the PRA's primary and secondary objectives. The respondent considered the increase in regulatory burden would be disproportionate to the additional risk. Another respondent called for the PRA to carry out quantitative analysis as part of its CBA to estimate the impacts of the MA reforms on competitiveness, particularly focussing on how they may affect smaller insurers.

10.33 The PRA acknowledges that some features of the proposed reforms, for example the MA attestation and safeguards for assets with HP cash flows, are designed specifically to advance the PRA’s primary objectives. The PRA reiterates that pursuing its secondary objectives cannot come at the expense of its primary objectives.^{footnote [36]} Nevertheless, for the reasons set out in the original CBA, and as updated in this chapter, the PRA continues to believe the final package of reforms set out in this PS will also advance the PRA’s secondary objective for competition, as well as its secondary objective for competitiveness and growth.

11: Have regards analysis

Introduction

11.1 This chapter provides feedback to responses relating to the proposals in Chapter 11 (Have regards analysis) of CP19/23.

11.2 Chapter 11 of the CP covered in detail the factors which the PRA is required to have regard to, and which were significant in the PRA’s analysis of the proposals set out in the CP. These factors included competition, growth and competitiveness, climate and environmental targets and regulatory best practice. The PRA had regard to other factors as required, with other ‘have regards’ covered at a high level where these were considered to not have a significant bearing on the proposals.

11.3 The PRA received eight responses to Chapter 11 of CP19/23. Respondents made a number of observations which are set out later in this chapter. The key themes that emerged included requests for a more proportionate treatment to recognise differences between businesses and the impact of the proposed package of reform on the attractiveness for international financial services.

Feedback to responses

11.4 The PRA has considered the responses to Chapter 11 of CP19/23. Feedback to the responses has been grouped into the following broad headings:

Competition (FSMA regulatory principles);
Growth and competitiveness (HMT recommendation letter); and
Other ‘have regards’ (Financial Services and Markets Act 2023 (FSMA 2023) and Legislative and Regulatory Reform Act 2006 (LRRA)).

11.5 This chapter also comments on changes to the draft policy (described elsewhere in this PS) that impact on the CP assessment against the factors to which the PRA is required to have regard, where relevant.

Competition (FSMA regulatory principles)

11.6 CP19/23 set out the approach taken within the PRA’s proposals to recognise differences between businesses, recognising that the size and composition of MA portfolios can vary materially between firms. The proposals were designed to be proportionate in their impact.

11.7 Two respondents asked for a more proportionate approach for firms with a small MA benefit compared to their overall balance sheet, including exemptions from some policy areas. These respondents also requested that the PRA further consider the unique nature of mutuals and their governance structure. Two further respondents requested a more proportionate approach for new and smaller firms. One respondent requested that the PRA implement the final MA policy only for firms that pose significant risks to the PRA's primary objective.

11.8 Having considered the responses, the PRA has made an additional change to the final MA attestation policy as set out in Chapter 6 of this PS, which involves a simplified approach for corporate bonds. This would apply a more proportionate treatment for firms whose MA portfolios are predominantly invested in less complex assets, or firms with a small MA benefit compared to the overall balance sheet, including mutual firms. The PRA considers that given the potential impact of any crystallisation of uncertainty over the ability to earn the MA, its policy should apply to all firms that seek to make use of the MA. Nevertheless, the PRA recognises that there may be situations, usually on grounds of proportionality or impracticability, where relevant reasons and supporting evidence may be provided to support a waiver or modification of the rules, which would be considered by the PRA on a case-by-case basis.

Growth and competitiveness (HMT recommendation letter)

11.9 The PRA considered that the proposals in CP19/23 could have significant benefits for facilitating investment in productive assets and sustainable growth. The proposed reforms were also designed with the Government’s SRF in mind, in which regulators set the detailed requirements which apply to firms directly through their rulebooks, operating within a framework established by Parliament. The proposals, relating to both liability and asset eligibility, were designed to be more tailored to the business models of UK firms.

11.10 Two respondents expressed concern that the proposals would not improve the relative attractiveness of the UK to internationally active financial services firms and activity or align with the Government’s strategy to promote competitiveness. One respondent considered that the lower cost of capital of non-UK domiciled reinsurers would result in a further loss in investment to the UK economy with profits being passed overseas. One respondent expressed concern that the PRA’s proposals failed to align with the Government’s aim to deliver SRF because they would be applied uniformly to all firms. One respondent expressed concern that the proposals would not go far enough to materially boost the role the UK life insurance sector plays in increasing productive and sustainable investment in the UK economy.

11.11 Having considered the responses, the PRA notes that strong and reliable standards, together with healthy competition in the financial sector, and consideration of the UK’s long-term output and growth, collectively underpin the success of the UK as an international financial centre. The PRA considers that these factors help maintain trust among domestic and foreign firms and enable the financial sector to support the real economy.

11.12 The proposals in CP19/23, together with the changes outlined in this PS, will also increase the range of eligible assets for UK firms. This may reduce the propensity for firms to use reinsurance to transfer liabilities offshore, effectively increasing the competitiveness of the UK as an insurance centre. The PRA considers that it has taken a proportionate approach where appropriate, rather than a uniform approach, with exemptions and waivers applying for policy areas. The PRA also notes its ongoing responsiveness to the alignment of investment needs of the industry and financing needs of the UK economy, for example through the launch of 'sandbox’ SEGs to explore further suggestions raised by some respondents which fall outside the scope of CP19/23.

Other ‘have regards’ (FSMA 2023 and LRRA)

11.13 Under the ‘Climate and environmental targets (FSMA 2023)’ section in the CP, the PRA mentioned that the proposals would not directly seek to incentivise firm investment into ‘green’ assets; however, many insurance firms have indicated their view that the Solvency II reforms could improve the ability of the sector to contribute to the Government’s net zero targets. The PRA noted that firms would have to consider the extent to which the FS is sufficient compensation for all retained risks, including environmental risks related to climate change.

11.14 One respondent queried how an attestation process with a specific focus on environmental risks related to climate change should be carried out, considering that this contradicted the expectation that the attestation process for corporate bonds would be straightforward.

11.15 As described in Chapter 6 of this PS, the PRA has updated its expectations relating to the attestation process for corporate bonds in response to CP19/23 responses. The PRA also notes that firms may make use of scenario modelling already performed (for example in relation to climate risks) if relevant to assist with an assessment of whether an increased risk allowance is needed for any particular asset.

11.16 The PRA did not receive any comments for other ‘have regards’.

12: General points raised by respondents

Introduction

12.1 This chapter sets out the PRA’s feedback to any relevant, general points raised in response to Chapter 1 (Overview) of CP19/23 and other general areas that are not aligned to the specific proposals in the CP and the PRA’s final decisions, where applicable.

12.2 The PRA received 21 responses that raised general points in this respect, which covered a number of areas. Some of the overarching comments received are addressed in the summary of responses in Chapter 1 of this PS. The remaining general comments are addressed in this chapter.

Feedback to responses

12.3 The PRA has considered the general responses received to CP19/23 and has set out its feedback below. The responses have been grouped as follows:

treatment of currencies in the MA calculation;
implementation;
future developments;
engagement with the PRA on the Solvency II reforms;
external audit of the MA calculation;
application of PPP to assets with HP cashflows;
consistency of definitions between PRA Rulebook and IRPR Regulations;
PRA risk appetite;
weakening of regulation; and
comments outside of the scope of CP19/23.

Treatment of currencies in the MA calculation

12.4 The PRA received comments from respondents on how to reflect assets and liabilities in different currencies in their MA calculation and/or the MALIR. These responses were not related to the policy proposals in CP19/23 but raise important points on which the PRA sets out its response below.

12.5 As part of their responses on Chapter 9 of CP19/23, one respondent stated that firms should be allowed to use the FS for the currency that is best reflective of the risk profile of the asset in question.

12.6 The PRA considers that the risks associated with holding assets in different currencies should be part of firms’ risk management processes. The PRA also expects firms to consider the appropriate treatment of risks for exposures in different currencies, including the appropriateness of the FS/MA, as part of their attestation processes and has updated paragraph 5.35 of SS7/18 to make this clear.

12.7 Two respondents raised questions regarding the appropriate treatment of liabilities in different currencies in an MA context. As part of their comments on Chapter 8 of CP19/23, one respondent asked the PRA to confirm whether it was satisfied with an approach that involves separate MA calculations for each currency, with the results of these calculations then being aggregated to give an overall MA for the entire MA portfolio. Separate to this, in their response on Chapter 7 of CP19/23, another respondent commented that the approach that firms should take if they have liabilities of different currencies within an MA portfolio was, in their view, ambiguous.

12.8 The PRA notes that there is a range of potential calculation approaches that firms could take where their MA portfolio contains liabilities in more than one currency. The PRA considers that the key focus should be on ensuring the resulting MA is appropriate for the MA portfolio, taking account of the assumptions underlying the MA set out in Chapter 1A of SS7/18. Firms’ considerations of the appropriateness of the MA should also take account of the approach to cash flow matching where the presence of liabilities in different currencies can give rise to additional risks. Where a firm has concerns or questions regarding the approach it has used in its MA calculation or for the MALIR, it should discuss these with its usual supervisory contact in the first instance.

Implementation

12.9 The PRA outlined the expected implementation timetable in Chapter 1 of CP19/23. At the time the CP was published further implementation details were not available. Responses relating to the implementation of specific policy are addressed in the relevant chapters of this PS.

12.10 Seven respondents requested the PRA engaged early and clarified practical implementation details. Respondents mentioned that this would be beneficial as it would:

reduce the chances of early development work being at risk;
provide the PRA with early industry feedback on the feasibility of implementation in the proposed timescales; and
enable firms to plan resourcing and activities for 2024 to invest in assets that the Government has indicated they would like insurers to support.

12.11 Five respondents requested that the PRA take a pragmatic initial approach to implementation on 30 June 2024. One respondent asked for clarity on which elements of the policy required PRA permission to implement and when firms can apply for such permissions.

12.12 In light of the responses received on implementation of the MA reforms, the PRA published its April statement to provide some clarifications requested by firms in advance of the publication of this PS. Chapter 1 of this PS summarises the details on when the different elements of the proposed policy are expected to be implemented and, where applicable, relevant chapters have expanded upon this.

12.13 Two respondents were concerned with the tight implementation timelines and proposed that the implementation of the reforms should be moved to 31 December 2024 rather than 30 June 2024. One respondent mentioned that the PRA should take more time if necessary to make informed public policy decisions.

12.14 Consistent with the IRPR Regulations coming into force, the PRA has agreed an implementation date of 30 June 2024 with the Government and remains committed to this date to ensure firms will be able to take advantage of these MA reforms in advance of 31 December 2024.

12.15 One respondent asked the PRA to clarify when firms may be able to apply for permission to include assets with HP cash flows in MA portfolios. Another respondent asked the PRA to clarify when firms should engage with the PRA on the SCR calculation for assets with HP cash flows, and if there should be a concurrent internal model change application.

12.16 The PRA confirms that firms can apply for permission to include such assets in MA portfolios from 30 June 2024, in line with the IRPR Regulations. The PRA considers that firms should make internal model changes in accordance with their internal model change policies as approved by the PRA. The PRA has not set out any additional internal model expectations specific to these assets.

12.17 One respondent recommended ongoing engagement after the implementation date to ensure outstanding issues or concerns with the interpretation of expectations are closed. Another respondent queried whether ongoing compliance would be followed up with individual supervision teams directly or a central PRA team.

12.18 The PRA confirms that firms should continue to engage with their supervisors on ongoing compliance with the MA policy.

12.19 One respondent queried how existing approvals would need to be updated to reflect the reforms, including the interaction with the 'automatic conversion' of existing approvals to permissions referred to in paragraph 1.52 of CP12/23.

12.20 As mentioned in paragraph 1.12 of this PS, the Government made a statutory instrument which ensures that firms with existing MA approvals can continue to apply the MA post 30 June 2024.

Future developments

12.21 Three respondents mentioned that they considered it beneficial for the PRA to review and make further enhancements to the MA regime in the coming years.

12.22 The PRA notes that further work is ongoing on some matters related to the MA, for example the ‘sandbox’ SEG. Other areas could also be identified as needing further work in due course. The outcome of such work could lead to future changes to the MA regime in the coming years. In addition, the PRA will keep implementation of the MA reforms under review and may consider that enhancements are required following analysis of relevant experience data, including industry feedback. As mentioned in the November 2022 statement, the PRA will also undertake an evaluation of its assessment of the impact on its statutory objectives of the Solvency II reforms, and its assessment of whether further changes are needed. These findings will be provided to the Government in order for it to take them into account during its five-year review in accordance with IRPR Regulations.

Engagement with the PRA on the Solvency II reforms

12.23 Throughout this consultation process, and the wider Solvency II review process, the PRA has endeavoured to ensure an appropriate level of engagement with firms. Eight respondents appreciated the PRA’s pre-consultation engagement with six of these specifically mentioning the helpfulness of the SEGs. Four respondents requested that the approach of using SEGs should be used for future policy development, which the PRA will consider going forward.

External audit of the MA calculation

12.24 One respondent asked for clarification on whether an auditor was required to consider if assets and liabilities included in an MA calculation fell within the scope of the PRA approved MA application.

12.25 The PRA considers that in line with the footnote to paragraph 4.2H of SS11/16, auditors are expected to form their own view about whether assets and liabilities are in scope of the PRA approved MA application. The PRA does not expect auditors to review any decisions or approvals made by the PRA on the MA application.

Application of PPP to assets with HP cash flows

12.26 In the CP, the PRA proposed an expectation in SS1/20 for firms to limit their investments appropriately where there is insufficient data to quantify the risks. One respondent asked for clarification as to whether paragraph 2.33 of CP19/23 and paragraph 3.24 of SS1/20 only refer to assets with HP cash flows.

12.27 The PRA confirms that both of the paragraphs mentioned above link to paragraphs 3.11 and 3.14 of SS1/20 and hence do not just apply to assets with HP cash flows (or even only to assets in an MA portfolio). The PRA notes that the proposed paragraph 3.24 of SS1/20, which it consulted on as part of CP19/23, applies to all assets, not only those in MA portfolios and that it is consistent with paragraph 3.14 of SS1/20 which already sets out that internal quantitative investment limits should have regard to the nature and quantification of the risks associated with the assets.

12.28 The PRA notes that assets with HP cash flows in MA portfolios are a good example of where sparsity of data may also increase uncertainty over the quality of matching and hence the ability to earn the MA which should lead firms, all else being equal to lower internal quantitative investment limits.

Consistency of definitions between PRA Rulebook and IRPR Regulations

12.29 One respondent raised that the IRPR Regulations define 'assigned assets' to implicitly cover components A and B of the MA portfolio while the draft changes to the PRA Rulebook define a similar term, ‘assigned asset’, to implicitly cover component A only. The respondent suggested it would be helpful if similar terms were not used in the PRA Rulebook and IRPR Regulations to mean different things.

12.30 The PRA considers that it does not have control over the terms used in the IRPR Regulations but has sought to harmonise language between the IRPR Regulations and the PRA Rulebook to the extent possible while retaining consistency with terms used more widely within the PRA Rulebook. As noted in the rules set out in Appendix 1 of this PS, the IRPR Regulations refer to the ‘assigned portfolio of assets’ which has the same meaning as the ‘relevant portfolio of assets’ in the PRA Rulebook. The exception to this is Regulation 5(5) of IRPR Regulations and replicated in Rule 4.7 of the Matching Adjustment Part of the PRA Rulebook, where ‘assigned portfolio of assets’ is referring to ‘assigned assets’ as set out in Rules 4.3 and 4.4 of the Matching Adjustment Part of the PRA Rulebook, which are themselves replications of Regulations 5(1) and 5(2) of the IRPR Regulations respectively.

PRA risk appetite

12.31 One respondent mentioned that the PRA’s risk appetite for exposing policyholders to risks appeared much lower in comparison to The Pensions Regulator (TPR) and Department for Work and Pensions (DWP) risk appetites.

12.32 The PRA considers that its risk appetite is not directly comparable to other regulatory bodies given the PRA operates in a different regulatory environment, with different statutory objectives set by Parliament. Specifically, the PRA, in line with its statutory objectives, needs to consider an appropriate degree of protection for policyholders and the potential impact on firms’ safety and soundness from the use of the MA. Furthermore, the Government stated at the outset of the Solvency II Review, that while there were opportunities to tailor the Solvency II regime much more to the needs of the UK, the Government and the PRA continued to support the fundamental principles and framework underlying Solvency II, and the need for the prudential regulatory regime to ensure high standards of policyholder protection and promote the safety and soundness of insurers. It also noted that the calibration of the capital standard in Solvency II has contributed to the insurance sector’s overall resilience to both recent events and PRA stress tests and is compatible with emerging international standards.^{footnote [37]}

Weakening of regulation

12.33 One respondent felt the Government's Solvency II regulatory policy intention was to weaken regulation and the PRA needed to have a stronger stance on this by making its position clear on whether it has accepted weaker regulation.

12.34 The PRA considers that, in line with the speech given by the CEO of the PRA to the Association of British Insurers (ABI) on 20 February 2023,^{footnote [38]} it has designed the MA policy in CP19/23 and this PS within the framework of Government’s legislation on the MA which has been passed by Parliament. The PRA considers that the reforms to the MA will allow the life insurance sector to play a bigger role in productive investment in the UK economy while also advancing its primary objectives of safety and soundness and policyholder protection as well as its secondary objectives within the framework set by the MA legislation.

Comments outside of the scope of CP19/23

12.35 The PRA did not consider general comments from respondents that did not relate directly to the proposals. However, the PRA has sought to provide clarity in response to some of these issues below, where appropriate.

12.36 One respondent challenged the appropriateness of the FS calibration.

12.37 The proposals in CP19/23 have been prepared in accordance with the IRPR Regulations which maintains the existing calibration of the FS. The PRA has therefore not further considered the calibration of the FS as part of the MA reforms.

12.38 One respondent expressed concern that different firms were interpreting the rules and regulations on the liability eligibility conditions for inclusion in MA portfolios differently, thereby leading to each firm implementing different solutions with varying degrees of benefit.

12.39 The PRA considers that there are already differences in the details of firm approaches and firms may choose to follow different practices according to their own policies. The PRA had also not made any proposals in CP19/23 in this aspect, other than those specifically described in Chapter 3 of this PS.

12.40 One respondent noted that they would continue to reinsure the majority of the longevity risk from new business. The same respondent expressed concern that the regulatory regime may hinder product and investment development that could come with the expected increase in demand for annuities.

12.41 The PRA notes that the proposals in CP19/23 were limited to changes relating to the MA, developed in accordance with the PRA’s objectives and the IRPR Regulations. Firms’ reinsurance approaches and wider changes to the regulatory regime are therefore beyond the scope of this PS.

This was previously published under PS3/24.
Solvency II Review – Matching adjustment reform implementation considerations for 30 June 2024.
Since the April statement was published, the Government has made relevant legislation. See paragraph 1.12 of this PS.
The Financial Services and Markets Act 2000 (Disapplication of Modification of Financial Regulator Rules in individual Cases) Regulations 2024.
Sections 138J (5) and 138K(4) of FSMA.
Regulation 4B (3) – (7).
Regulation 5 of The Financial Services and Markets Act 2023 (Commencement No. 4 and Transitional and Saving Provisions) (Amendment) Regulations 2023 (2023/1382).
In CP5/24 the PRA proposed to amend its existing SoP: The PRA’s approach to the publication of Solvency II technical information (the ‘TI SoP’) to restate paragraphs (3) to (5) of Regulation 4B in, and to not restate paragraphs (6) and (7) within, PRA policy materials.
In paragraph 1.48 in Chapter 1 (Overview) of CP19/23, the PRA explained that it expected the impact of the proposed reforms on mutuals to be no different from the impact on other firms.
Sections 138J (3) and 138J(4) of FSMA.
Section 138J(2)(d) of FSMA.
For further information please see Transitioning to post-exit rules and standards.
This was originally numbered 2.12D in the consultation draft of SS7/18.
This was originally numbered 5.22 in the consultation draft of SS7/18.
Paragraphs 5.24, 5.24A, 5.24B and 5.24C were originally numbered 5.24 in the consultation draft of SS7/18.
Final policy on MLAs has since been published in PS2/24.
PRA’s approach to supervision of the banking and insurance sectors.
As provided for in Insurance – Allocation of Responsibilities Rule 3.1(4) of the PRA Rulebook.
Regulation 4(4) of the IRPR Regulations.
PS31/18 – Solvency II: Equity release mortgages, paragraph 2.16.
Rule 9.1 of the Matching Adjustment Part of the PRA Rulebook.
Rule 4.17 of the Matching Adjustment Part of the PRA Rulebook.
Version published as part of PS2/24 and effective from 31 December 2024. This policy statement and the associated consultation paper set out the justification for the PRA’s proposals as regards the use of capital add-ons.
Also set out in Rule 4.11 of the Matching Adjustment Part of the PRA Rulebook.
Version published on 2 October 2023, applicable as of 1 January 2024.
Note this SoP is currently subject to an ongoing consultation in CP5/24.
Note SS8/18 includes material relating to PRA expectations on the use of proceeds from the sale of defaulted assets – see paragraph 5.18. While those expectations apply in stressed scenarios, firms may find these useful when considering recovery rates generally.
Components A, B and C of the MA portfolio are defined in paragraph 4.5 of SS7/18.
This template reference will change to IR.06.02 effective from 31 December 2024 as part of PS3/24. The IR.06.02 template can be found in Appendix 2 of PS3/24.
This template reference will change to IR.22.01 effective from 31 December 2024 as part of PS3/24.
A joint industry and PRA group to gather a broad range of information on, and options for, the development of the new supervisory measures in respect of Notching. The NSEG’s focus was on how to effectively deliver the UK Government’s intention to legislate to allow the use of notched ratings in the FS methodology and calibration.
Capital Add-ons: Policies relating to Capital Add-Ons for Solvency 2 insurers.
An asset in the relevant portfolio of assets that falls within the scope of Rule 4.4(1) of the Matching Adjustment Part of the PRA Rulebook.
PRA’s approach to supervision of the banking and insurance sectors | Bank of England.
General insurance pricing attestation multi-firm review | FCA.
The PRA notes its recent consultation on its approach to policy in CP27/23 - The Prudential Regulation Authority’s approach to policy | Bank of England. As part of that CP, the PRA is considering quantitative metrics at a high level, including comparison with other jurisdictions.
HM Treasury Review of Solvency II: Call for Evidence.
Fundamental Spreads - speech by Sam Woods.

Appendices

Appendix 5: Update to supervisory statement 3/17 – Solvency II: Illiquid unrated assets

Appendix 6: Update to supervisory statement 1/20 – Solvency II: Prudent Person Principle

Appendix 7: Update to supervisory statement 11/16 – Solvency II: External audit of, and responsibilities of the governing body in relation to, the public disclosure requirement

Appendix 10: Respondents to CP19/23 who have consented to the publication of their names (pdf)

Note: All content for this PS is included on this webpage. If you would like a PDF version of the PS, please use the button available below.

Download a PDF version of PS10/24

Prudential Regulation // Supervisory statement

15 November 2024

Prudential Regulation // Supervisory statement

15 November 2024

Prudential Regulation // Prudential regulation

15 November 2024

Prudential Regulation // Consultation paper

15 November 2024

Necessary cookies	Analytics cookies
Yes	Yes
Yes	No

Our use of cookies

Necessary cookies

Analytics cookies

PS10/24 – Review of Solvency II: Reform of the Matching Adjustment

Related links Related links

1: Overview

Background and developments since CP19/23

Improving business flexibility

Being more responsive to the level of risk

Enhancing firms’ responsibility for risk management

Summary of responses

Changes to draft policy

Accountability framework

Structure of the PS

Implementation

2: Investment flexibility

Introduction

Changes to draft policy

Feedback to responses

Criteria for inclusion of assets with HP cash flows

Clarification of expectations in relation to fixed assets in MA portfolios

Perimeter between fixed assets and assets with HP cash flows

Asset restructuring

Controls on the quality of matching

Total exposures of additional risks

Additional matching tests

FS additions for assets with HP cash flows

Standard Approach for FS additions

Determination of best estimate cash flows for assets with HP cash flows

Modelling assets with HP cash flows under stress

3: Liability eligibility

Introduction

Changes to draft policy

Feedback to responses

Restatement of existing MA liability conditions

Expanded liability eligibility conditions

4: Credit ratings under the MA

Introduction

Changes to draft policy

Feedback to responses

Investment in SIG assets

Modelling the FS for SIG assets in internal models

Proportionality of requirements and expectations for internal credit assessments

Assurance on internal credit assessments

Responsibility for the internal credit assessment function

Other responses

5: MA permissions, breaches, and consequential rule changes

Introduction

Changes to draft policy

Feedback to responses

Restatement of existing regulations into PRA Rules and new MA SoP

PPP

Streamlined MA application approach

Breaches of MA eligibility conditions

Delegated authority to submit MA applications

6: Matching adjustment attestation

Introduction

Changes to draft policy

Feedback to responses

FS and MA attestation wording

Treatment of corporate bonds

Analysis level and offsets

High degree of confidence that the MA will be earned

Suggestions on FS and MA attestation wordings

FS additions guidance

Spread volatility, credit rating and the MA

EVT for ERMs in attestation

FS additions in stress

Principles-based guidance versus benchmarking

Miscellaneous suggestions

Attestation frequency and responsibility

Out-of-cycle attestations

Attestation responsibility

Public disclosure requirement and external audit

Public disclosure requirement

External audit

Implementation

7: Assumptions underlying the MA

Introduction

Changes to draft policy

Related links