CP16/22 – Implementation of the Basel 3.1 standards: Market risk

Overview

6.1 This chapter sets out the Prudential Regulation Authority’s (PRA) proposals to implement the Basel 3.1 standards on market risk. These comprise: new requirements for determining which positions should be allocated to the trading book; a recalibrated version of the existing standardised approach as a simplified standardised approach (SSA); and two new calculation methodologies – a new advanced standardised approach (ASA); and a new internal model approach (IMA). The new market risk framework and methodologies are proposed to replace the existing calculation methodologies for market risk capital requirements. Table 1 summarises how the key proposals map to the existing market risk framework.

Table 1: Summary of how the key proposals map to the existing market risk framework

6.2 The proposals in this chapter would:

• amend the existing Trading Book (CRR) Part of the PRA Rulebook;
• introduce a new Market Risk: Simplified Standardised Approach (CRR) Part of the PRA Rulebook;
• introduce a new Market Risk: General Provisions (CRR) Part of the PRA Rulebook;
• introduce a new Market Risk: Advanced Standardised Approach (CRR) Part of the PRA Rulebook;
• delete the existing Market Risk Part of the PRA Rulebook, transferring paragraphs 3 and 4 of that Part to Market Risk: Simplified Standardised Approach (CRR) and Market Risk: General Provisions (CRR) respectively; and
• amend Supervisory Statement (SS) 13/13 ‘Market risk’.

6.3 The PRA’s proposals would implement the new market risk framework finalised by the Basel Committee on Banking Supervision (BCBS) in 2019. That framework was designed in response to the global financial crisis, which revealed material weaknesses in the Basel 2 market risk framework. Market risk capital requirements proved insufficient to absorb losses. As an immediate response, BCBS introduced a set of revisions to the market risk framework in July 2009, which we refer to as ‘Basel 2.5’. The Basel 2.5 amendments were a necessary short-term fix, but they made the framework significantly more complex, and did not address all of the issues revealed by the crisis. The Basel 3.1 standards introduce a comprehensive set of amendments to the market risk framework. The PRA played a key role in designing and agreeing the new market risk standards.

6.4 The proposals in this chapter are intended to improve the coherence of the framework, promote consistency across firms, and more comprehensively address the market risks posed by firms’ exposures. The proposals include a range of approaches of differing levels of sophistication to support proportionality. The proposals would:

• more clearly define the scope of the framework by introducing a stricter delineation between positions that should be allocated to the trading book and non-trading book, and specifying the treatment of internal hedges between the two books;
• retain a recalibrated version of the existing standardised approach as the SSA for firms with limited derivatives business. The updated calibration reflects market developments since the approach was initially introduced;
• introduce a new, more comprehensive standardised approach – the ASA. The PRA proposes that this would be used by firms that do not meet the criteria to use the SSA and that have not been granted supervisory permission to use the new IMA; and
• introduce a new IMA for firms that have been granted supervisory permission. This approach would replace the existing modelled approach.

6.5 The PRA proposes to add detail and apply targeted adjustments relative to the Basel 3.1 standards in several areas, described in more detail in the individual proposals below.^{footnote [1]} The most material areas are as follows:

• In the SSA:
  • introduce eligibility criteria that firms would need to meet to continue using the SSA;
  • amend SS13/13 to update the PRA’s expectations on the calculation of modified duration as an input to the SSA; and
  • incorporate the substantive elements of existing technical standards relating to the existing market risk standardised approach into PRA rules.^{footnote [2]}
• In the ASA:
  • clarify the calculation of gross jump-to-default for the default risk charge (DRC);
  • adjust the treatment of exposures to carbon emissions trading schemes; and
  • expand the range of allowable data sources to determine risk weights for positions in collective investment undertakings (CIUs).
• In the new IMA:
  • add detail on the calculation of capital requirements for non-modellable risk factors (NMRFs) and requirements for recognition of NMRFs in back-testing;
  • simplify modelling approaches for positions in CIUs, subject to tests to ensure they are appropriately conservative;
  • clarify the treatment of non-trading book foreign exchange (FX) and commodity positions; and
  • update SS13/13 to revise the PRA’s existing ‘risks not in value-at-risk’ (RNIV) framework, remove duplicative data standards, and remove requirements related to calculation methodologies made redundant by these proposals.

6.6 The proposals in this chapter are relevant to PRA-authorised banks, building societies, PRA-designated investment firms, and PRA-approved or PRA-designated financial holding companies or mixed financial holding companies (‘firms’). The proposals would not apply to UK banks and building societies that meet the Simpler-regime criteria and choose to be subject to the Transitional Capital Regime proposals.^{footnote [3]}

Scope of application

6.7 The Basel 3.1 standards introduce new requirements on the scope of application of the market risk framework in three key areas:

• the assignment of positions to the trading book or non-trading book (which determines whether a position is subject to the market risk or credit risk framework respectively);
• the treatment of internal hedges (ie the requirements for when a firm can internally transfer risks between the market, credit, and credit valuation adjustment (CVA) risk frameworks); and
• requirements for exemptions from market risk capital requirements for positions used to mitigate structural foreign exchange (SFX) risk.

Assignment of positions to the trading book or non-trading book

6.8 The assignment of positions to the trading book or non-trading book determines whether they are treated under the market risk or credit risk framework. Aligned with the Basel 3.1 standards, the PRA proposes to:

• prescribe lists of positions that would need to initially be assigned to either the trading book or the non-trading book, and require that firms obtain PRA permission to deviate from those lists. Positions assigned to the trading book would be subject to market risk capital requirements;
• set restrictions on any subsequent reassignment of positions between the trading and non-trading books, and require that, except in specific circumstances, firms cannot reassign a position between trading and non-trading book more than once;
• require that when a position is reassigned between the trading and non-trading book, firms hold a capital add-on equal to any capital reduction resulting from the reassignment until the position matures or expires; and
• retain the requirement that market risk capital requirements must be calculated for all FX and commodity positions, whether allocated to the trading book or non-trading book.

6.9 The PRA considers that introducing objective criteria for assigning positions to the trading book or non-trading book would support a more consistent treatment of similar risks across firms. The proposals would also help to ensure that positions moving between the trading book and non-trading book continue to have appropriate capital requirements.

Treatment of internal hedges

6.10 Firms use internal hedges to manage risks that cross the trading book, non-trading book, and CVA portfolio. For example, to hedge a non-trading book position, a firm may enter into an interest rate derivative that is allocated to the trading book and then use an internal hedge to transfer the risk of that derivative from the trading book to the non-trading book. Aligned with the Basel 3.1 standards, the PRA proposes constraints on the recognition of internal hedges between risks in the trading book and non-trading book. Internal hedges would only be recognised where:

• for credit risk or counterparty credit risk hedges, the internal hedge would be recognised as unfunded credit protection in the credit risk mitigation framework (see Chapter 5 – Credit risk mitigation), and is exactly matched by a set of trading book positions with external third parties;
• for equity risk hedges, the internal hedge would be recognised as a hedge in the credit risk framework, and is exactly matched by a set of trading book positions with external third parties; and
• for general interest rate risk hedges, the hedges are conducted between the non-trading book and a dedicated internal hedge portfolio in the trading book that:
  • is separately capitalised from the rest of the trading book; and
  • only contains instruments that either directly arise from transactions with an external third party, or are exactly matched by a set of trading book positions with external third parties.

6.11 The PRA proposes that internal hedges between the trading book and the CVA portfolio would only be recognised where:

• the hedges would be recognised as eligible hedges in the CVA risk framework; and
• if the internal hedge is subject to curvature risk, default risk, or the residual risk add-on in the ASA, the internal hedge is exactly matched by a set of trading book positions with external third parties.

6.12 The PRA considers that these proposals would help to ensure that risks transferred between the trading book and non-trading book (or between the trading book and CVA portfolio) through internal hedges are either mitigated or have adequate capital requirements.

Exemptions from market risk capital requirements for positions used to mitigate structural foreign exchange risk

6.13 SFX risk is a risk that firms are exposed to when they have assets and capital resources that are denominated in a currency that is different to their reporting currency. In those situations, they are exposed to the risk of capital ratio volatility that is purely due to FX rate movements. Firms can mitigate SFX risk by holding unhedged FX positions such that the change in the value of foreign currency risk-weighted assets (RWAs) due to a movement in the FX rate is offset by a proportionate change in the value of the positions. Without exemptions for these FX positions, they would be subject to market risk requirements even though they are used to reduce capital ratio volatility.

6.14 The PRA proposes to supplement the existing requirements for SFX exemptions with two additional requirements, aligned with changes made in the Basel 3.1 standards:

• the size of the exempt position should not be greater than the size of position that neutralises a firm’s capital ratio sensitivity; and
• the exempted positions should be exempted for at least six months.

6.15 The PRA considers that the additional requirements would ensure exemptions are only applied to positions they are intended to cover (those which reduce volatility in the capital ratio). They also set a more objective requirement to demonstrate that positions are structural and not subject to frequent adjustment.

6.16 The PRA proposes that management and mitigation of SFX risk should not depend on the capital calculation methodology, and therefore the exemptions should be available to all firms. To clarify this and remove potential ambiguity in the current rules, the PRA proposes the requirements on exemptions for SFX positions are moved to Article 325a1(18) of the Market Risk: General Provisions (CRR) Part.

PRA objectives analysis

6.17 The PRA considers that the proposals set out in this section advance the PRA’s primary objective of promoting safety and soundness of firms. The proposals originate from the Basel 3.1 standards, which the PRA expects other jurisdictions will also implement. The proposals are intended to ensure that positions with market risks are appropriately capitalised by assignment to the trading book, and reduce the risk that subsequent reassignment of positions or internal hedging of risks lead to inadequate capital requirements. The proposed additional requirements on SFX exemptions would advance safety and soundness by limiting the size of exempted positions to those that meet the purpose of the exemption.

6.18 These proposals support the PRA’s secondary competition objective in that they would help to ensure a more consistent assignment of positions to the trading book and non-trading book across firms. The proposals would also facilitate more consistent recognition of internal hedges across firms.

‘Have regards’ analysis

6.19 In developing these proposals, the PRA has had regard to the FSMA regulatory principles, the aspects of the Government’s economic policy set out in the HMT recommendation letter from 2021 and the supplementary recommendation letter sent April 2022. Where the proposed new rules are CRR rules (as defined in section 144A of FSMA), the PRA has also taken into consideration the matters to which it is required to have regard when proposing changes to CRR rules. The following factors, to which the PRA is required to have regard, were significant in the PRA’s analysis of the proposals:

1. Proportionality (FSMA regulatory principles and Legislative and Regulatory Reform Act 2006):

• The PRA considers that the benefits from the proposed requirements on position assignment are proportionate to the burden imposed on firms. By setting more objective requirements, the assignment of positions to the trading book or non-trading book would be simplified. The PRA considers that clarifying the application of the existing SFX exemption framework, explicitly limiting the size and duration of positions exempted, would be consistent with the current intent and practice for SFX exemptions and therefore should not impose material additional burden on firms.

2. Relevant international standards (FSMA CRR rules):

• The PRA considers its proposals on the scope of application would be materially aligned with international standards.

3. The principle that the PRA should exercise its functions transparently (FSMA regulatory principles):

• The PRA considers that introducing clearer requirements on the scope of application would be a more transparent approach to firms compared to the current subjective nature of these requirements.

4. Efficient and economic use of PRA resources (FSMA regulatory principles):

• The PRA considers that proposing clearer requirements on position assignment and internal hedges would reduce the amount of supervisory resource needed to review firms’ individual practices.

Eligibility for different approaches

6.20 Consistent with the Basel 3.1 standards, the PRA proposes to introduce three approaches: the updated SSA; the ASA; and the new IMA. The PRA also proposes to retain the existing ‘derogation for small trading book business’. The PRA proposes to implement new eligibility requirements to use the different approaches, which would be included in Market Risk: General Provisions (CRR) Part.

6.21 The existing derogation for small trading book business permits firms to use the credit risk approach to measure market risk, where the size of their on- and off-balance sheet trading book business is less than 5% of total assets, and less than £44 million (without the PRA’s approval or notification). The PRA proposes to retain this derogation without amendment.

6.22 The SSA is a recalibrated version of the existing market risk standardised approach intended, in the Basel 3.1 standards, to be available to firms with limited market risks. The PRA proposes that firms meeting either of the following criteria can elect to use the SSA (without the PRA’s approval or notification, and subject to the restriction in paragraph 6.23 below):

• the firm’s aggregate market risk assets and liabilities are less than £440 million and less than 10% of total assets;^{footnote [4]} or
• the firm is eligible to use the derogation for small trading book business.

6.23 Consistent with the Basel 3.1 standards, the PRA additionally proposes that firms with correlation trading portfolios (CTP securitisations) should be prohibited from using the SSA due to the complexity of correlation trading.

6.24 The ASA is a comprehensive standardised methodology, intended in the Basel 3.1 standards to be available to all firms. Aligned with those standards, the PRA proposes that firms can elect to use the ASA without the PRA’s approval or notification.

6.25 Consistent with the Basel 3.1 standards, the PRA proposes that firms need to seek permission to use the IMA, with permissions granted at trading desk level.

6.26 The PRA proposes that firms should be allowed to use a combination of IMA and ASA to calculate market risk capital requirements. However, a firm that uses the SSA would need to do so for all market risk positions, and a firm that uses the small trading book derogation would need to do so for its entire trading book.

PRA objectives analysis

6.27 The PRA considers that retaining the existing small trading book derogation, and a recalibrated version of the existing standardised approach (the SSA), while introducing the new ASA and IMA, advances the PRA’s primary objective of safety and soundness. The derogation and SSA would provide operationally simple but conservative approaches for firms with limited market risks. The ASA is a more risk-sensitive methodology, more suitable than SSA for measuring more complex trading risks, and therefore would be made available to all firms. Finally, the IMA would provide an appropriate level of risk-sensitivity for firms with material market risks while being subject to an additional safeguard of PRA scrutiny of models by means of the permissions process. The proposed framework would advance the PRA’s primary objective by enhancing firms’ capture of market risk in their capital requirements, according to the size and complexity of their financial market activities.

6.28 The PRA also considers that the proposals would support its secondary competition objective by providing greater comparability of outcomes between firms across different approaches. The ability to gain permission to use the IMA at trading desk level would also reduce barriers to smaller firms using the most risk-sensitive approach to calculating capital requirements.

‘Have regards’ analysis

6.29 In developing these proposals, the PRA has had regard to the FSMA regulatory principles, the aspects of the Government’s economic policy set out in the HMT recommendation letter from 2021 and the supplementary recommendation letter sent April 2022. Where the proposed new rules are CRR rules (as defined in section 144A of FSMA), the PRA has also taken into consideration the matters to which it is required to have regard when proposing changes to CRR rules. The following factors, to which the PRA is required to have regard, were significant in the PRA’s analysis of the proposals:

1. Relevant international standards (FSMA CRR rules):

• The PRA proposals would align with international standards.

2. Proportionality (FSMA regulatory principles and Legislative and Regulatory Reform Act 2006):

• The PRA considers that the proposal to implement the four different approaches, each with a different level of operational complexity and risk-sensitivity, would provide an appropriate range of alternatives that are proportionate to the levels of market risk faced by different firms.

3. Different business models (FSMA regulatory principles):

• The PRA considers that the proposal to implement four different approaches also provides a range of alternatives for firms, depending on their degree of market risk-related activities. In particular, the relatively simpler SSA and derogation for small trading book business would be available to larger firms with limited and simpler trading activities.

4. Efficient and economic use of PRA resources (FSMA regulatory principles):

• The PRA has taken into account efficient and economic use of its resources. The proposals would limit the use of resources by restricting approval processes to reviewing firms’ own models for use under the IMA, and setting objective thresholds for the use of the SSA.

Simplified standardised approach (SSA)

6.30 The PRA proposes, aligned with the Basel 3.1 standards, to retain the existing market risk standardised approach, recalibrated to reflect market conditions and events since it was first introduced, for firms with limited derivatives business (as defined in the section ‘Eligibility for different approaches’).

Recalibration of the current standardised approach to be used as the SSA

6.31 Consistent with the Basel 3.1 standards, the PRA proposes to recalibrate the existing standardised approach by applying the following multipliers to the capital requirements calculated under the existing approach for each risk class:

Table 2: SSA multipliers

6.32 The PRA considers the proposed recalibration would ensure capital requirements are sufficiently conservative for the risks faced by firms, while limiting operational burden by maintaining the methodologies within the existing standardised approach.

Other amendments to the SSA

6.33 The PRA proposes to make a number of consequential amendments to improve the overall coherence of the SSA rules, including:

• incorporating the requirements contained in existing technical standards on non-delta risks^{footnote [5]} and on the definition of market^{footnote [6]} into PRA rules;
• updating SS13/13 to include the substance of the existing guidelines on modified duration;^{footnote [7]} and
• amending cross-references to other parts of the CRR that are being amended (particularly to the credit risk standardised approach).

6.34 As part of incorporating technical standards into PRA rules, the PRA proposes not to directly prescribe lists of:

• appropriately diversified stock indices that may be exempted from equity specific risk;^{footnote [8]} and
• closely correlated currencies, for which a lower risk weight may be applied.^{footnote [9]}

6.35 Instead, to improve the responsiveness of the rules to changes in the market and support the efficient use of supervisory resource, the PRA proposes to replace those lists with the current criteria used by the PRA to identify appropriately diversified stock indices or closely correlated currencies. Firms would be able to self-identify appropriately diversified stock indices and closely correlated currencies according to the same criteria currently applied by the PRA.

PRA objectives analysis

6.36 The PRA considers that retaining a recalibrated version of the existing standardised approach is consistent with its primary safety and soundness objective. Once recalibrated, the SSA would remain generally more conservative than the new ASA for firms with limited and less complex market risks, and the resulting capital requirements would better reflect the market risks observed since it was first implemented.

6.37 Retaining the SSA would support the PRA’s secondary competition objective by providing a simple approach that could be used by smaller firms with limited and less complex market risks.

‘Have regards’ analysis

6.38 In developing these proposals, the PRA has had regard to the FSMA regulatory principles, the aspects of the Government’s economic policy set out in the HMT recommendation letter from 2021 and the supplementary recommendation letter sent April 2022. Where the proposed new rules are CRR rules (as defined in section 144A of FSMA), the PRA has also taken into consideration the matters to which it is required to have regard when proposing changes to CRR rules. The following factors, to which the PRA is required to have regard, were significant in the PRA’s analysis of the proposals:

1. Proportionality (FSMA regulatory principles and Legislative and Regulatory Reform Act 2006):

• The PRA considers that its proposal would be proportionate. Recalibrating the existing standardised approach using high-level multipliers rather than changing the methodology would avoid the operational burden of implementing a new approach on firms using it.

2. Different business models (FSMA regulatory principles):

• The PRA considers that retaining the SSA offers a simpler approach for firms with business models that inherently are not focused on trading activities.

3. Relevant international standards (FSMA CRR rules):

• Retaining the SSA and recalibrating it using high-level multipliers aligns with the Basel 3.1 standards.

Advanced standardised approach

6.39 The Basel 3.1 standards introduce a new standardised approach (the ASA). The market risk capital requirements calculated under the ASA are calculated as the sum of three separate capital requirements:

• the Sensitivities-based Method (SbM) capital requirement;
• the residual risk add-on (RRAO); and
• the default risk charge (DRC).

6.40 The PRA considers that the ASA methodology represents a significant improvement in the measure of market risk under the standardised approach. Therefore, the PRA proposes to implement the ASA as set out in this section. As outlined in the section ‘Eligibility for different approaches’, the approach would be available for all firms.

Sensitivities-based method (SbM) capital requirement

6.41 The SbM calculates capital requirements based on the sensitivity of the value of a firms’ positions to a specified set of market risk factors. At a high level, the PRA proposes that the SbM capital requirement methodology is implemented consistently with the Basel 3.1 standards, and therefore calculated through a number of steps:

• positions would be allocated to one or more of seven risk classes according to types of risks firms are exposed to (eg general interest rate risk, equity risk);
• each risk class would have a predefined set of risk factors. Risk factors are market variables relevant to the risk class, such as specific interest rates or FX rates, the movements of which would affect the value of positions;
• firms would calculate the sensitivity of each position to movements in the value of each risk factor, and multiply them by a prescribed risk weight – equivalent to the potential movement of that risk factor in stressed market conditions. Risk weights are scaled to take into account the relative liquidity of different risk factors. The resulting risk-weighted sensitivities can be considered to be the potential loss to the firm’s positions from the movement in that risk factor in a stressed market; and
• the risk-weighted sensitivities would be aggregated using prescribed formulae and correlations to allow for diversification benefit.

6.42 To address the risk that correlations can fluctuate in periods of stress, the final aggregation step would be performed three times, assuming specified high, medium, and low correlations between risk factor movements. The above steps would be followed separately for each risk class and the risk class-level capital requirements would then be aggregated as a simple sum. The total SbM capital requirement would be the largest of the capital requirements calculated for the three correlation scenarios.

Residual risk add-on (RRAO)

6.43 The RRAO is intended to ensure capital requirements are adequate to address complex or exotic risks not considered in the SbM or DRC. Aligned with the Basel 3.1 standards, the PRA proposes that the RRAO is calculated as the sum of gross notional amounts of positions multiplied by 1% for those with exotic underlyings,^{footnote [10]} and by 0.1% for those with other residual risks.^{footnote [11]}

6.44 Consistent with the Basel 3.1 standards, the PRA proposes to set out a non-exhaustive list of positions that would be considered to have exotic underlyings or be exposed to other residual risks.

Default risk charge (DRC)

6.45 The Basel 3.1 standards’ DRC is intended to set capital requirements for default risk from credit and equity positions. It is calibrated to a similar degree of conservatism as the credit risk framework, but recognises a greater degree of offsetting between long and short positions. The methodology involves calculating the ‘gross jump-to-default’ exposure arising from each credit and equity position, applying a default risk weight (based on credit rating), and calculating the overall capital requirement after recognising a degree of offsetting benefit between long and short positions.

6.46 To support a clear and consistent interpretation of the ASA DRC framework, the PRA proposes to introduce a minor adjustment to the Basel 3.1 standards by requiring firms to calculate the ‘gross jump-to-default’ of an instrument as the difference between:

• the current market value of an instrument position; and
• the market value of the instrument assuming an instantaneous default of the underlying credit or equity instrument with recovery equal to a percentage of the notional or face value of the instrument. The percentage would be determined as 1 minus the loss given default (LGD), with LGD prescribed in the PRA rules as per the Basel 3.1 standards.

6.47 The PRA considers that its proposed calculation of ‘gross jump-to-default’ improves the clarity and consistency of the DRC calculation. The PRA expects that its alternative definition would only rarely lead to different outcomes relative to the Basel 3.1 standards.

Question 38: Do you have any comments on the PRA's proposed definition of ‘gross jump-to-default’ in the ASA default risk charge?

Treatment of carbon emissions trading schemes

6.48 The Basel 3.1 standards do not provide a separate treatment for carbon emissions certificates. The PRA considers that, at present, there is insufficient evidence to propose a separate treatment. However, it proposes to provide a framework so the treatment could be easily amended in the future as carbon markets evolve. The PRA proposes to keep the treatment of carbon trading under review as evidence accumulates. As such, the PRA proposes to:

• introduce a distinct commodities bucket for carbon emissions certificates with a distinct risk weight and tenor basis correlation parameter; and
• set the risk weight and correlation for carbon emissions certificates identical to the Basel 3.1 standards for other commodities (ie a 60% risk weight and 99% tenor basis correlation).

Question 39: Do you have any comments on the PRA's proposal for carbon emissions certificates? What additional information could be considered for the calibration of risk weights and correlations, particularly relating to any historical period of stress?

Treatment of collective investment undertakings (CIUs)

6.49 The Basel 3.1 standards set out three approaches to calculating capital requirements for CIUs in the ASA:

Look-through approach (LTA): Where a firm knows the exact holdings of a CIU, firms can treat the holdings of the CIU as if they were on the firm’s own balance sheet.

Mandate-based approach (MBA): Where a firm knows the investment mandate of the CIU (but not the actual holdings), firms can calculate their exposure by assuming that the CIU invests in a portfolio consistent with its mandate that generates the maximum possible capital requirement. Because the actual holdings may differ substantially from the mandate, no hedging or diversification benefit can be recognised between the capital requirements calculated under the MBA and other market risk positions.

Fall-back approach (FBA): Where a firm does not have the information for the LTA or MBA, or chooses not to use them, they would need to use the FBA. This approach makes no assumptions about the holding of the fund, and therefore a conservative risk weight of 70% is applied to the firm’s position. Given the lack of knowledge of the fund positions, no hedging or diversification benefit is permitted with other market risk positions.

6.50 The PRA proposes to implement the above three approaches. In the case of the MBA, firms would require permission to apply this approach.

6.51 Additionally, the PRA proposes to implement a fourth approach for the treatment of CIUs – an external party approach (EPA). Under the proposed EPA, where a firm has access to a risk weight for the CIU that is calculated by an external party, the firm may use that risk weight for their position in the CIU provided that:

• the external party knows the exact holdings of the CIU and calculates the risk weight each reporting period in accordance with the LTA;
• the external party’s risk weight calculation is externally audited, and the firm verifies the appropriateness of the external party’s risk weight calculation; and
• no hedging or diversification benefit is permitted between the position in the CIU and other market risk positions.

6.52 The PRA considers its proposal to implement the EPA would offer a proportionate approach that provides more risk-sensitive capital requirements than the MBA and FBA, while being less operationally burdensome for firms. It would also be at least as conservative as the LTA, and therefore ensure capital requirements are aligned with the risk of the CIU position.

Question 40: Do you have any comments on the PRA's proposals to include the EPA for the treatment of CIUs in the new ASA?

Treatment of non-trading book foreign exchange (FX) and commodity positions

6.53 The PRA proposes to introduce new rules that clarify how non-trading book FX and commodity positions should be reflected in the ASA. The proposal would be aligned with the Basel 3.1 standards, but provide additional prescription.

6.54 For FX risk positions in the non-trading book, the PRA proposes that firms may elect to use either the accounting or, if calculated at least quarterly, the fair value of those positions for the purposes of calculating capital requirements under the ASA. Aligned with the Basel 3.1 standards, the PRA proposes that under either option, firms should update the FX component of non-trading book positions at least monthly.

6.55 For commodity positions in the non-trading book, the PRA proposes that firms should recalculate the fair value of their positions at least monthly, and use the last available fair value in calculating capital requirements under the ASA.

6.56 The PRA considers this additional prescription would ensure a consistent treatment of such risks across firms.

PRA objectives analysis

6.57 The PRA considers that the proposed ASA advances its primary safety and soundness objective. It would improve the risk-sensitivity of capital requirements relative to the existing standardised approach, and more accurately capture default risks, residual risks, and exotic risks. This would better align capital with risk for firms using the ASA.

6.58 Regarding the treatment of carbon emissions trading certificates, the PRA considers that there is insufficient evidence at present to justify a less conservative treatment relative to the Basel 3.1 standards, and therefore has not adjusted the calibration based on primary objective concerns. In particular, there is a lack of data on how such markets may behave in periods of significant stress.

6.59 Regarding the treatment of CIUs, the PRA considers that the proposed EPA advances its primary safety and soundness objective, since the risk weight is calculated consistently with the LTA, but hedging and diversification recognition are constrained, reflecting that firms may not know the actual holdings of the CIU.

6.60 The PRA considers that the ASA furthers its secondary competition objective by introducing a credible alternative to internal models that would be accessible to firms with smaller trading portfolios. The improved risk capture and sensitivity of the ASA would mean that capital requirements are more commensurate with the actual risks faced by such firms.

‘Have regards’ analysis

6.61 In developing these proposals, the PRA has had regard to the FSMA regulatory principles, the aspects of the Government’s economic policy set out in the HMT recommendation letter from 2021 and the supplementary recommendation letter sent April 2022. Where the proposed new rules are CRR rules (as defined in section 144A of FSMA), the PRA has also taken into consideration the matters to which it is required to have regard when proposing changes to CRR rules. The following factors, to which the PRA is required to have regard, were significant in the PRA’s analysis of the proposals:

1. Proportionality (FSMA regulatory principles and Legislative and Regulatory Reform Act 2006):

• The PRA considers its proposals would support proportionality by introducing a risk-sensitive alternative to the IMA. Regarding the calculation of gross jump-to-default, the proposal would limit the burden for firms implementing the approach by setting out a clearly defined approach to be applied to all positions. Regarding the treatment of CIUs, the EPA is designed to provide a more proportionate alternative to the MBA and FBA while being less operationally burdensome to implement compared to the LTA and MBA.

2. Relevant international standards (FSMA CRR rules):

• The PRA considers the proposals to be aligned with international standards. In areas where the proposals differ from the Basel 3.1 standards, the PRA considers its proposals to achieve equivalent outcomes. The proposal for calculating ‘gross jump-to-default’ for DRC is aligned to the Basel 3.1 standards because it leads to a generally equivalent calculation. The proposed treatment of carbon emissions trading schemes is aligned with the Basel 3.1 standards. The PRA considers that its proposal to introduce the EPA remains aligned with the Basel 3.1 standards as the EPA would lead to capital requirements that are at least as conservative as under the LTA.

3. 2050 net-zero target in the Climate Change Act 2008 (HMT recommendation letters / FSMA CRR rules):

• The PRA considers that its proposal on the treatment of climate emissions trading schemes is consistent with the government’s commitment to achieve a net-zero economy by 2050. The PRA concludes that its proposal for a more granular treatment of emissions certificates is warranted (at the cost of marginally increasing the complexity of the ASA) to provide scope for future adjustments to risk weights or correlations if sufficient evidence were to emerge that it was required.

4. Competitiveness (HMT recommendation letters):

• The PRA considers that its proposal on the treatment of CIUs supports the competitiveness of the UK as an attractive domicile for internationally active financial institutions trading in CIUs. It would provide an additional approach that is aligned to the Basel 3.1 standards, while being operationally simpler than the approaches in the Basel 3.1 standards.

5. The principle that the PRA should exercise its functions transparently (FSMA regulatory principles):

• The PRA considers that its proposal for the treatment of non-trading book FX and commodities positions would improve the clarity of its rules, ensuring a consistent treatment of these risks is applied across firms.

Internal model approach (IMA)

6.62 The Basel 3.1 standards introduce a new market risk IMA to replace the existing framework. Under the new IMA, the market risk capital requirement would be the sum of three separate components:

• an expected shortfall (ES) calculation which incorporates the variable liquidity of different risk factors;
• a default risk charge (DRC); and
• a separate capital requirement for non-modellable risk factors (NMRFs).^{footnote [12]}

6.63 Use of the new IMA would be subject to permission, which would be granted at the level of firms’ trading desks. The Basel 3.1 standards introduce related requirements for the structure and eligibility of trading desks to use internal models.

6.64 The PRA considers that the improvements incorporated in the new IMA, including the better capture of liquidity risk in the ES calculation, and a prescribed treatment of NMRFs, would provide a more comprehensive, coherent, and consistent approach to calculating market risk capital requirements compared to the existing IMA. Therefore, the PRA proposes to implement the new IMA and related requirements on firms’ trading desk structure as set out in this section. As outlined in the section ‘Eligibility for different approaches’, the IMA would be available to firms who receive permission from the PRA, and would be granted at trading desk level, aligned with the Basel 3.1 standards. As previously indicated,^{footnote [13]} the PRA would expect firms to submit final pre-application materials for new IMA permissions at least 12 months in advance of the PRA’s proposed Basel 3.1 implementation date (see Chapter 1 – Overview).

Requirements for the trading desk structure for IMA permissions

6.65 The Basel 3.1 standards set requirements on the trading desk structure that firms are required to have in place when applying for permission and to use the IMA. The requirements are intended to promote consistent approaches across firms while aligning trading desks with the operational structure of firms’ trading businesses. The PRA considers the requirements are important to ensure consistency across firms, and to provide clarity on how firms should prepare to request IMA permission. Therefore, the PRA proposes to implement the Basel 3.1 standards in this area without amendment.

Internal model approach (IMA) calculation

Expected shortfall

6.66 The Basel 3.1 standards introduce a new ES model to calculate the risk of losses in firms’ trading positions due to movements in market variables (referred to as risk factors). Whereas value-at-risk (VaR) and stressed value-at-risk (SVaR) – the two models used in the existing framework – set capital requirements using the estimated loss that will not be exceeded over a given time frame to a certain level of confidence, ES considers both the size and likelihood of losses that might occur above a defined confidence level. The ES model is calibrated to a historical period of stress.

6.67 The new ES model includes two further enhancements relative to the existing approach:

• it recognises that correlations between different risk areas behave unpredictably in periods of stress, so diversification between broad risk factor types (eg equity, commodity) is restricted; and
• it better captures risk factor illiquidity in periods of stress by prescribing a ‘liquidity horizon’ ranging from 10 to 120 days for each risk factor. The ES is calculated based on estimated risks over a 10-day horizon, and scaled up to the prescribed liquidity horizon to determine the final ES risk measure for each risk factor.

6.68 The PRA proposes to implement the ES model, aligned with the Basel 3.1 standards, including two tests on the accuracy and conservatism of a firm’s ES model that would need to be passed in order for it to be used to calculate capital requirements. The PRA’s proposals would require that for each trading desk, a firm’s model would need to continually pass ‘back-testing’ and ‘profit and loss attribution tests’ (PLAT) for the desk to be treated under the IMA. The proposals also prescribe back-testing requirements at the overall trading book portfolio level, aligned with the Basel 3.1 standards:

• Back-testing is a test of model conservatism. It counts the number of times that actual losses exceed the estimate from the firm’s model (a back-testing ‘exception’). The proposals would require that when a trading desk model exceeds a specified number of back-testing exceptions over a 12-month period, it would no longer be permitted to be treated under the IMA.
• PLAT are tests of model accuracy. PLAT applies two statistical tests that compare the time series of daily profit and loss (P&L) calculated using the risk factors and pricing models in the ES model, to the actual daily P&L of the trading desk. Simplifications in the model will cause the two time series to differ – if the tests show the differences exceed specified thresholds, this indicates there are material simplifications in the model and the desk would no longer be permitted to be treated under the IMA. Aligned with the Basel 3.1 standards, the PRA proposes to delay the application of this test for the purposes of calculating market risk capital requirements (but not for reporting on them) until one year after the proposed IMA rules are implemented.

Default risk charge

6.69 The DRC in the Basel 3.1 standards is intended to measure the jump-to-default risk of credit and equity positions in a firm’s trading book. Relative to the existing framework’s model for default risk, the DRC would reduce unwarranted variability in modelled capital requirements by removing consideration of migration risk – captured under the ES model in the new framework – and providing greater specification of input parameters. It also requires both credit and equity portfolios to be modelled, replacing the existing option for firms to not include equity positions.

6.70 The PRA considers the DRC would improve consistency in capital requirements across firms, and by removing modelling of migration risk it would reduce the potential for the model to overlap with risks included in the ES model. It therefore proposes to implement the DRC, aligned with the Basel 3.1 standards.

Non-modellable risk factor (NMRF) framework

6.71 The Basel 3.1 standards introduce a new NMRF framework. The framework recognises that trading desks with permission to use the IMA may be exposed to risk factors for which there is limited market data and therefore not well suited to being modelled. The NMRF framework is intended to ensure that those risks have adequate capital requirements. The PRA has existing expectations for firms to consider capital add-ons for these types of risks, and it considers that introducing a consistent framework in rules would enhance the approach and ensure greater consistency across firms. It therefore proposes to implement the NMRF framework in the IMA approach as set out in paragraphs 6.72 to 6.78.

Data quality standards for NMRFs

6.72 The PRA proposes, aligned with the Basel 3.1 standards, that risk factors would only be permitted to be included in firms’ ES models if they have at least 24 observations in the preceding 12 months, with at least four observations in any 90-day period in those 12 months. The PRA also proposes to include a series of qualitative criteria that the risk factor would need to meet. Risk factors not meeting any of the criteria would be excluded from a firm’s ES model and would be separately capitalised through the NMRF framework using the approach described in paragraphs 6.73 to 6.76.

Capital requirements for NMRFs

6.73 Consistent with the Basel 3.1 standards, the PRA proposes that each NMRF would need to be capitalised based on an individual stress scenario that is at least as conservative as a standalone ES calculation for that NMRF. NMRFs would then be aggregated assuming a 36% correlation assumption between different NMRFs to recognise a degree of diversification benefit.

6.74 The PRA proposes to further clarify the Basel 3.1 standards by requiring that firms develop and document methodologies to calculate the capital requirement for individual NMRFs. Firms would be required to ensure their methodologies are adequately conservative, with a high degree of confidence, and consider any potential limitations in their calculations, including limitations arising from sparse data, portfolio non-linearities, or reliance on proxies.

6.75 The PRA proposes that the stress period used to calculate NMRF capital requirements for each risk category would be the 12-month period that either:

• maximises the sum of NMRF capital requirements for all NMRFs in that risk category; or
• maximises the ES calculation for modellable risk factors in the same risk category, where a firm is able to demonstrate that applying the same stress period for NMRFs would not result in a materially different outcome to a period that maximises the NMRF capital requirement.

6.76 The proposals would also add detail to the requirements set out in the Basel 3.1 standards, to support consistent implementation. They prescribe a methodology for firms to determine a time series of 10-day risk factor returns for calculating NMRF capital requirements. Where a firm is unable to identify an appropriate individual stress scenario for an NMRF, the PRA proposes a fall-back approach where the firm would determine NMRF capital requirements by applying a shock to the NMRF that is:

• for positions with a finite maximum loss, the shock to the risk factor that would lead to that maximum loss; and
• for positions with an infinite maximum loss, the greater of:
  • a qualitatively-determined shock to the NMRF that would not be exceeded in a future stress period with 99.95% certainty; or
  • the shock that would lead to the maximum historically observed loss on that NMRF.

Recognition of NMRFs in back-testing

6.77 For the purposes of the proposed back-testing requirements for the IMA, the PRA proposes to specify how NMRFs should be treated. The Basel 3.1 standards require NMRFs to be excluded from models used for back-testing at both trading desk level and portfolio level. Recognising that under this approach back-testing exceptions could be due to NMRFs being excluded from the model rather than poor model performance, the Basel 3.1 standards contemplate supervisors permitting firms to ignore such exceptions where the capital requirements for a single NMRF are greater than the entire daily loss. The PRA proposes to implement a more risk-sensitive and flexible approach than the Basel 3.1 standards that it considers would ensure capital requirements are proportionate to the risk of firms using insufficiently conservative models. The proposal would require that:

• when performing back-testing at the trading desk level, firms may elect to include NMRFs in their model; and
• when performing back-testing at the overall trading book portfolio level, firms would need to exclude NMRFs from their model – firms would be permitted to ignore back-testing exceptions at portfolio level that are caused by NMRFs if approved by the PRA.

6.78 The proposals would require firms to report the NMRFs they elect to include in their model at trading desk level and to regularly report back-testing performance at the overall portfolio level.

Question 41: Do you have any comments on the PRA's proposals to recognise NMRFs in your model for the purposes of back-testing at the trading desk level? To what extent would you be able to incorporate NMRFs into your model for back-testing?

Treatment of collective investment undertakings

6.79 The Basel 3.1 standards prescribe that firms may only apply the IMA to positions in CIUs where the firm looks through to the CIU’s underlying positions and models them as though they are held on the firm’s balance sheet. The PRA considers that the ability to look through to the underlying holdings of a CIU is an important component of risk estimation. However, to reduce the operational burden on firms, the PRA proposes a simpler approach that it considers would achieve an outcome that is at least as conservative as the Basel 3.1 standards.

6.80 The PRA proposes to allow firms to apply the IMA to positions in CIUs, without separately modelling each underlying position, where:

• the firm is able to look through to the underlying positions of the CIU; and
• the firm, at a minimum annually, demonstrates to the PRA that the outcomes of its modelling of the CIU are consistent with or clearly more conservative than modelling by looking through to the underlying positions of the CIU.

Question 42: Do you have any comments on the PRA's proposal to allow firms a greater degree of modelling flexibility for CIUs in IMA?

Treatment of non-trading book FX and commodity positions

6.81 Similar to its proposals for the ASA, the PRA proposes to clarify how non-trading book FX and commodity positions should be reflected in the IMA. The proposals would be aligned with the Basel 3.1 standards, but provide additional prescription to ensure consistent and transparent treatment of these risks across firms.

6.82 To include non-trading book FX risk positions in their IMA calculations, the PRA proposes that firms should update the value of those positions to reflect changes in FX rates at least daily.

6.83 To include non-trading book commodity positions in their IMA calculations, the PRA proposes that firms should also update the fair value of those positions at least daily.

6.84 For the purposes of back-testing and PLAT, the PRA proposes to clarify that:

• for non-trading book FX positions that are not fair-valued and whose value moves linearly with respect to FX rates, firms would be required to include the effect of changes in FX rates in actual and hypothetical P&L calculations. Firms would be permitted to elect to include the effect of changes in all risk factors that determine the value of those positions; and
• for non-trading book commodity positions, consistent with the proposed requirement to fair-value non-trading book commodity positions for the IMA, firms would be required to include the full change in the fair value of such positions in actual and hypothetical P&L calculations.

Replacing the PRA’s risks not in value-at-risk framework with a risks not in model framework

6.85 In addition to the proposed implementation of the IMA, the PRA proposes to introduce a rule that would require firms to hold additional capital requirements for material deficiencies in risk capture in their internal models. The PRA proposes to set out its expectations for meeting this requirement by amending its existing RNIV framework in SS13/13 and convert it into a new ‘risks not in model’ (RNIM) framework through amendments to:

• specify the scope of the RNIM framework to cover risks not included either in a firm’s ES model or NMRF framework, and risks not included or adequately capitalised in firms’ DRC models;
• be clear that firms are expected to identify all model limitations and missing risks. Firms would be expected to maintain a centralised inventory to track limitations and assumptions that may have an impact on the output of market risk models;
• specify that while all model limitations and missing risks should be identified, firms would only be required to hold RNIM capital add-ons for any material risks not adequately captured by internal models;
• include an expectation that, where appropriate, firms should calculate RNIM capital add-ons in accordance with the PRA’s proposed requirements on calculating capital requirements for individual NMRFs; and
• add a procedure whereby firms seek agreement from the PRA to recognise limited diversification or offsetting benefits between specific RNIMs where empirically justified.

6.86 Additionally, the PRA proposes that:

• subject to receiving an explicit waiver from the PRA, firms may be permitted to recognise the contributions of RNIMs in the PLAT test; and
• firms would be required to continue holding RNIM capital add-ons for at least 12-months for trading desks that have reverted to the ASA by reason of failing the desk-level back-testing or PLAT requirements.

6.87 The PRA considers that its proposed conversion of the RNIV framework into the RNIM framework would remove potential overlaps between RNIMs and NMRFs. This is specifically by clarifying that firms are required to hold RNIM capital add-ons only for material risks not adequately captured in internal models, and by potentially recognising some diversification or offsetting benefits across RNIMs. Similarly, the PRA considers that its proposal to have a process to potentially allow the contributions of RNIMs in PLAT removes potential overlap of requirements and would be proportionate and consistent with the prudential intention of the PLAT test as a test of model accuracy.

6.88 The PRA considers that its proposal to require firms to continue holding RNIM capital add-ons for trading desks failing back-testing or PLAT would ensure that capital requirements for desks failing back-testing or PLAT do not see an automatic reduction in capital requirements as a result of the removal of RNIM capital add-ons, which may not be adequately captured under the ASA.

Amendments to SS13/13 ‘Market risk’

6.89 As a consequence of implementing the new IMA, the PRA proposes to amend SS13/13 to delete previous guidance relating to the existing IMA that ceases to be relevant under the IMA.

6.90 The PRA also proposes to amend Chapter 12 of SS13/13 to update the existing expectation on ‘Significant Influence Function (SIF) attestation’ with a reference to the PRA’s senior managers’ framework.

PRA objectives analysis

6.91 The PRA considers that introducing the new IMA as outlined above advances its primary objective of safety and soundness. The proposals address significant shortcomings identified by the BCBS and PRA in the existing framework, by:

• implementing a new risk measure (ES) that more effectively considers tail risks relative to the existing model;
• constraining diversification benefits between broad risk factor types, to better recognise that correlations between very different risks behave unpredictably in periods of stress;
• recognising risk factor illiquidity in periods of stress by prescribing different ‘liquidity horizon’ to risk factors;
• reducing pro-cyclical IMA capital requirements, by removing the current VaR risk measure which is calibrated to the most recent 12-month period; and
• restricting the use of IMA where firms cannot demonstrate that they are reasonably able to model risks, via the NMRF framework and back-testing and PLAT requirements.

6.92 The proposals to implement the NMRF framework would ensure that firms take into account material data limitations and potential losses in stressed periods when calculating individual NMRF capital requirements. The PRA considers that requiring permission to exclude back-testing exceptions due to NMRFs at the portfolio level, and defining reporting requirements on NMRFs included in back-testing at trading desk level, advances its primary objective by providing safeguards around how firms are taking account of NMRFs in the back-testing process.

6.93 Regarding the treatment of CIUs, the PRA considers that its proposals would ensure modelling approaches reflect the risks of underlying holdings, while reducing the operational burden on firms.

6.94 The PRA considers that its proposed introduction of an RNIM framework to include DRC model deficiencies advances its primary safety and soundness objective by ensuring that all model deficiencies are subject to review, and if material, capital add-ons. The PRA considers that its proposal to require firms to continue holding RNIM capital add-ons for trading desks failing back-testing or PLAT would also advance its primary safety and soundness objective by ensuring that overall capital requirements for desks that fail back-testing or PLAT would not have an automatic reduction in capital requirements, where such risks may not be adequately captured under the ASA.

6.95 The proposals would support the PRA’s secondary competition objective by restricting the use of IMA where modelling is not prudent, helping ensure that lower capital requirements for firms with modelling permission are only achieved where modelling is appropriate. By allowing model permission to be granted at trading desk level, the proposals would also reduce barriers to smaller firms being able to use the approach.

‘Have regards’ analysis

6.96 In developing these proposals, the PRA has had regard to the FSMA regulatory principles, the aspects of the Government’s economic policy set out in the HMT recommendation letter from 2021 and the supplementary recommendation letter sent April 2022. Where the proposed new rules are CRR rules (as defined in section 144A of FSMA), the PRA has also taken into consideration the matters to which it is required to have regard when proposing changes to CRR rules. The following factors, to which the PRA is required to have regard, were significant in the PRA’s analysis of the proposals:

1. Proportionality (FSMA regulatory principles and Legislative and Regulatory Reform Act 2006):

• The PRA considers that the proposals would provide a risk-sensitive approach for firms with substantial market risk, which would ensure the regulatory burden is commensurate with the risks that such firms are exposed to. The proposed additional prescription on identification of stressed periods would minimise the NMRF framework’s operational burden, and the proposed approach to the treatment of NMRFs in back-testing would overcome potentially unduly conservative outcomes from the Basel 3.1 standards’ approach. The burden that would arise from the PRA’s proposed treatment of CIUs remains proportionate to its benefits by offering a degree of additional flexibility to firms modelling CIUs.

2. Relevant international standards (FSMA CRR rules):

• The PRA considers that the proposals are materially aligned with international standards. The proposed methodology for calculating capital requirements under the NMRF would clarify the approach in the international standards while retaining a similar level of conservatism. The proposed treatment of NMRFs in back-testing, while more flexible than international standards, aligns with the intent of international standards by ensuring that firms would not have to hold duplicative capital requirements for the same risks. Regarding the treatment of CIUs, the PRA considers that its proposal to allow firms to take an operationally less burdensome approach to modelling remains aligned with the outcome of the international Basel 3.1 standards.

3. Efficient and economic use of PRA resources (FSMA regulatory principles):

• The PRA considers the benefits of improved risk capture justify the supervisory resource required to review and approve firms’ applications to use IMA, which include ensuring that the relevant modelling standards are met. Due to the relative importance of the NMRF framework, the PRA considers it an efficient use of supervisory resources to review the methodologies developed by firms.
• The PRA considers that its proposals to prescribe how firms identify stress periods and risk factor returns for NMRFs would remove the need for supervisors to review individual firms’ methodologies, minimising the use of supervisory resources.
• The PRA’s proposal to permit firms to incorporate NMRFs in models for back-testing at the trading desk-level would also limit the need for supervisory resource to review firm requests to ignore exceptions.
• In contrast, the PRA considers that requiring supervisory approval to exclude back-testing exceptions at the overall trading portfolio level provides important visibility and assurances around how firms are including NMRFs in their models, and the related supervisory resource would be limited, given the smaller number of calculations required for portfolio-level back-testing.

4. Competitiveness (HMT recommendation letters):

• The PRA considers that its proposals on the treatment of CIUs and of NMRFs in back-testing support the competitiveness of the UK as an attractive domicile for internationally active financial institutions. Allowing firms to include NMRFs in their back-testing models would potentially reduce exceptions that could arise from NMRFs (where they have been capitalised already). For CIUs, the proposals minimise the operational burden of modelling CIUs.

5. The principle that the PRA should exercise its functions transparently (FSMA regulatory principles):

• The PRA considers that its proposal for the treatment of non-trading book FX and commodities positions would provide improve clarity of its rules, ensuring a consistent treatment of such risks is applied across firms.