FRTB regulation – PRA revises the IMA for market risk

The internal model approach for market risk (IMA) has proven to be one of the more complex areas of Basel 3.1 regulation. As a crucial element of the Fundamental Review of the Trading Book (FRTB), the IMA rules offer a more sensitive view of market risk than the alternative Advanced Standardised Approach (ASA). Despite the final rules published in January 2026, the PRA has proposed a number of adjustments to encourage uptake, improve international alignment and promote a more proportionate approach.

PLAT monitoring extended to three years 

Under FRTB, use of the IMA for market risk is determined through the PLAT, which consists of the Spearman correlation test and the Kolmogorov-Smirnov test. These compare a trading desk’s modelled profit and loss against its actual profit and loss over a one-year period. 

In CP9/26, the PRA has proposed extending the PLAT to three years to ensure accurate calibration of the pass/fail thresholds. During this period, firms will need to report test outcomes, but a failed PLAT will not trigger a move to the ASA until after the initial three-year window closes. This will essentially ease firms into the regime, without raising capital requirements early in the post-implementation period. 

Article

Basel 3.1 and SDDT regimes – PRA issues final rules

Read more

Changes to the RFET under FRTB 

The RFET determines which risk factors are included in the expected shortfall (ES) model, using both quantitative and qualitative measures. If a factor that fails either test over a 12-month period, then it must be assessed through the non-modellable risk factor (NMRF) framework instead. The PRA has proposed two key changes, as outlined below. 

Treatment of illiquid risk factors 

Under the initial rules, the firms needed to provide 24 verifiable prices for risk factors with a liquidity horizon greater than 20 days. It’s dropped this figure to 16 to avoid double counting illiquidity and needlessly pushing risk factors into the NMRF framework. Allowing more risk factors to stay within the ES model can improve proportionality and operational effectiveness of the framework. 

Pro-rated data for new issuances 

The RFET works on a 12-month basis, meaning new issuances would fail the test even if they have a high volume of pricing data for modelling. The PRA proposes that firms can pro-rate the data, allowing desks that actively trade new instruments to continue to use the IMA approach, resulting in more risk-sensitive market risk management and potentially lower capital requirements. 

In addition to the above, the PRA also plans to create more proportionate rules for modelling constraints, and some audit requirements for prices verified by regulated exchanges or trading platforms. Moving forward, firms’ senior management may need greater assurance over pricing data quality, independent price verification and governance over market data used in the IMA. 

New intermediate category of NMRF 

The RFET aims to capture risk factors that are non-modellable. However, the Basel Committee on Banking Supervision’s (BCBS) Quantitative Impact Study (QIS) data showed significantly more risk factors being classified as NMRF than anticipated, with the PRA concluding that the RFET can inadvertently capture factors that are modellable but simply illiquid. This pushes some factors into the NMRF framework, resulting in higher capital requirements. 

As such, the PRA has proposed an intermediate category of NMRFs to allow those factors to stay under the ES: 

• Type 1 for risk factors that fail the quantitative part of the RFET but pass that qualitative aspect – moving forward they’ll will be subject to an NMRF capital-add, with a zero-correlation assumption, on to reflect the limited data available on verifiable pricing. 
• Type 2 for risk factors that fail both the quantitative and qualitative elements of the RFET, and will be capitalised under the NMRF framework. 

The PRA also proposes several operational simplifications including: aligning NMRF and ES model stress period; reducing the Type 2 NMRF calculation frequency from daily to monthly; and removing the distinction between idiosyncratic and non-idiosyncratic NMRFs. 

Article

RWA accuracy: what recent ECB fines mean for Basel 3.1

Read more

A gradual move to the IMA for market risk  

Under the initial Basel 3.1 market risk rules, firms using both ASA and IMA added the two capital figures together as a simple sum. When desks moved to IMA, they lost their diversification benefit with the rest of the ASA portfolio, which could artificially inflate total capital requirements and create a barrier to gradual IMA adoption. 

The PRA had initially implemented a cap to prevent firms paying higher capital than they would have under full ASA. It is now going further by proposing an adjustment to how ASA and IMA capital are aggregated. Firms will calculate the difference between the ASA for all positions and the ASA for IMA positions, preserving the diversification benefit of desks that move to IMA. The existing cap is replaced with a permission-based cap at the full ASA level, though the PRA does not expect it to be widely used. 

A simpler approach for CIUs 

The initial Basel 3.1 market risk rules allowed firms to apply IMA to CIUs without modelling each position separately. However, they must be able to look through all underlying holdings and demonstrate (at least annually) that their CIU modelling resulted in a capital requirement that was equal to, or higher than, individually modelled positions. 

This approach could be operationally heavy for firms and small changes to CIU composition could introduce capital cliff-edges. As such, the PRA’s proposed a 90% look-through threshold (assessed quarterly), with the remaining 10% capitalised under the ASA’s fallback approach. Index-tracking funds can be treated as a direct position in the tracked index, provided the annualised return difference between a CIU and its tracked index is below 1% over 12 months. 

Minor adjustments and clarifications 

In addition to the above, the PRA’s made a few other minor adjustments and clarifications to the IMA framework under Basel 3.1: 

• Firms modelling with a reduced set of risk factors (subject to PRA notification), must demonstrate that their coverage captures at least 75% of the variability of the full set. 
• General interest rate risk (GIRR) internal hedges desks can use the same stress period and reduced risk factor set as other IMA desks, but their own funds requirement must still be calculated on a standalone basis. 
• Where a listed closed-ended fund also meets the definition of a CIU, it is excluded from the CIU boundary requirements. 

The PRA is also updating reporting and disclosure requirements to meet the revised rules.  

Article

Top themes for banking and capital markets in 2026

Read more

What do the FRTB regulation changes mean for firms? 

The updated FRTB market risk rules are broadly positive and could make a significant impact for firms implementing Basel 3.1. For larger global banks, the IMA will be easier to apply operationally, with smoother transitions between the ASA and IMA. Greater international alignment will also help cross-border groups take a consistent, coordinated approach to compliance costs. Meanwhile, mid-sized firms can benefit from improved proportionality, although the rules remain complex. 

Impact on capital requirements 

The PRA's changes could result in a modest reduction in capital requirements, but the bigger impact is a more stable and proportionate implementation. The longer PLAT monitoring period reduces the risk of desks reverting to the ASA, helping firms protect their IMA investment and associated infrastructure. The revised NMRF classification also means more risk factors can remain within the ES model, producing more risk-aligned capital outcomes. 

Model governance and validation

The revised RFET and NMRF classification rules will require enhanced data quality frameworks and systems to demonstrate verifiable prices and assess risk factor eligibility. Firms will need to embed effective processes for the extended PLAT monitoring and back-testing regime, with model validation cycles aligned to updated regulatory expectations. Internal audit functions will need to review these processes to make sure they’re fully embedded within the broader risk management framework, and that they align with SS1/23 expectations. 

Next steps for FRTB market risk management 

A more proportionate IMA may encourage wider adoption, prompting strategic reassessment of the wider FRTB regulation implementation programme. With the go live date remaining 1 January 2028, firms have significant work ahead to meet the deadline. To get started, firms can benefit from a gap analysis to see whether strategic adoption of IMA (on a full or hybrid basis) is preferable to the business model and capital structures. For firms weighing that decision, re-running the cost-benefit case in light of the ASA/IMA diversification adjustment and the shift from partial caps to a permission-based cap will be essential. 

Other key actions include: 

• Assess how the revised NMRF Type 1 and Type 2 split, the lower RFET threshold and the aligned stress period would affect capital requirements and the modellable/non-modellable boundary for current desks. 
• Review CIU holdings against the proposed 90% look-through and 1% index-tracking thresholds. 
• Map the draft reporting and disclosure changes to existing reporting systems 
• Review cross-border approaches to ensure alignment across all territories and monitor for any additional international divergence. 

The consultation closes on 18 September 2027 for firms wishing to help shape the future regulatory landscape. For further information on Basel 3.1 and FRTB, contact Kantilal Pithia or Vivian Lagan.