FRTB for trading banks: What the EU relief through 2029 means for sales and trading

If you trade bonds, interest rate derivatives, equity derivatives, FX or structured products inside a European bank, the date is one your treasury keeps in mind and that becomes concrete fast at desk level: On 1 January 2027 the bank switches the capital calculation for its trading book to a new methodology. The capital charge rises – moderately on some trades, drastically on others. The European Commission published a draft on 22 April 2026 that scales that charge back over three years. The consultation closed on 19 May 2026; formal adoption is pending. What is at stake, what the relief actually saves in capital terms and what sales and trading should understand right now is what this article sets out.

1 · What this is about – in five sentences

Banks must hold capital against every trade they put on. How much they must hold derives from a regulatory calculation. The current calculation is dated and seen by many observers as too lenient – the replacement, called FRTB, arrives in Europe on 1 January 2027. It demands materially more capital for the same trades. Because the US and the United Kingdom do not adopt FRTB on the same schedule, the European Commission introduces a relief for the years 2027 to 2029 that scales the additional charge back at bank-individual level – so that European banks can hold on to their sales and trading franchises in international competition.

2 · FRTB in one paragraph

FRTB stands for Fundamental Review of the Trading Book. The Basel Committee on Banking Supervision finalised the framework in 2019, and the EU has anchored it in the Capital Requirements Regulation 3 (CRR3). Three things change against today's regime. First, risk factors are decomposed more finely – instead of standard add-ons for entire risk classes, banks compute sensitivity-based contributions per risk factor. Second, the Internal Models Approach (IMA) is governed by far stricter rules (see Section 6). Third, the boundary between trading book and banking book is tightened and becomes harder to reverse. In aggregate: same portfolio, same risks – significantly more capital. Industry calls it the FRTB shock; supervisors call it risk-adequate capitalisation. Both have a point.

3 · The problem: Competitors compute more slowly

FRTB is an international Basel standard. That means all G20 banking jurisdictions have committed to implementing it – but not at the same speed. In the US, the Federal Reserve published a first implementation proposal (Notice of Proposed Rulemaking, NPR) in July 2023 that contained FRTB at full sharpness. That proposal was never finalised. After the change of administration in January 2025, US bank supervisors effectively shelved it. On 19 March 2026 the Federal Reserve, the Office of the Comptroller of the Currency (OCC) and the Federal Deposit Insurance Corporation (FDIC) jointly published a new proposal – explicitly framed as capital-neutral. The consultation closes on 18 June 2026, with finalisation expected in the fourth quarter of 2026. There is no binding FRTB application date for US banks.

The United Kingdom takes a middle path. The Prudential Regulation Authority (PRA) of the Bank of England published its final rules in January 2026 (Policy Statement PS1/26). The FRTB standardised approach applies in the United Kingdom in parallel with the EU from 1 January 2027 – but the internal models approach (see Section 6) only from 1 January 2028. UK banks with model approval will therefore carry one year less of market risk capital than their EU peers on the same trade. The PRA cites uncertainty over US implementation explicitly as the reason.

For a European trading floor competing against JPMorgan in New York and against Barclays in London for the same corporate clients, hedge-fund mandates and market-making volumes, the implication is this: Without an EU relief, the same trade would carry a measurable capital surcharge on the bank's own balance sheet that the competition does not. That is the political core of the issue.

4 · The EU answer – and what it is explicitly not

The EU has already postponed FRTB twice – from the original 1 January 2025 to 1 January 2026, and then to 1 January 2027. A third postponement is no longer legally available. The underlying empowerment in Article 461a of the Capital Requirements Regulation (CRR) allows two one-year postponements – no more. Anyone hoping for another postponement is hoping in vain. FRTB arrives on 1 January 2027.

What the Commission does instead: It uses a second lever within the same empowerment. Article 461a CRR allows it not only to postpone FRTB but also to make »targeted amendments to the market risk framework for up to three years« – provided that the international level playing field is at risk. Which is what has happened. FRTB enters into force on 1 January 2027 as planned. Banks must from that date compute and report their market risk capital under FRTB. But the resulting values are then multiplied by factors that scale them back – through the end of 2029.

5 · How the relief works in mathematical terms

The relief consists of two combined multipliers. A multiplier is mathematically nothing more than a number by which the FRTB capital requirement is multiplied before the result becomes binding on the bank. A multiplier of 0.9 retains 90 per cent of the FRTB requirement as the binding capital figure – and relieves the bank by 10 per cent.

The first multiplier is bank-individual. Each bank, at the reference date 31 March 2027, compares its FRTB market risk capital requirement against the same requirement under the current Basel 2.5 rules. Where FRTB demands more than Basel 2.5, the difference is compensated through the bank-individual multiplier. Concretely: If a bank would have to hold EUR 1,000 million of market risk capital under FRTB but only EUR 600 million under Basel 2.5, the multiplier becomes 0.6 (= 600/1,000). The bank then holds EUR 600 million instead of EUR 1,000 million, neutralising the FRTB increase temporarily.

The second multiplier is a flat 0.9. It is mandatory for all banks in the standardised approach and operates on top of the bank-individual multiplier. Both work cumulatively. A bank whose bank-individual multiplier is, say, 0.75 reaches an effective factor of 0.75 × 0.9 = 0.675 through the combination – that is, the bank holds 67.5 per cent of the FRTB requirement as bound capital.

6 · The three sweeteners for model banks

In addition to the multipliers, the act contains three technical interventions affecting only banks that operate under the Internal Models Approach (IMA). IMA means: The bank computes market risk capital not under a rigid supervisory schedule (which is the standardised approach), but using an internal model approved by the supervisor and fed by the bank's own risk measurements. The internal model is usually more capital-efficient – but FRTB regulates it so tightly that many banks have questioned whether to use it at all.

NMRF: Risk factors without market data

Non-Modellable Risk Factors (NMRF) are risk factors for which the internal model is not allowed to derive an independent estimate because there is not enough market data. Example: A specific 10-year credit spread position on a mid-cap industrial whose bond trades only sporadically. The model does not see enough data points, so the bank must apply an NMRF surcharge to that factor – and that surcharge is painfully high under FRTB. During the relief phase, the NMRF requirement is reduced to 35 to 45 per cent of its regular value. This is the single most important relief for banks operating in less liquid risks (credit, emerging-market rates, volatility products).

PLAT: The model reality check

The Profit and Loss Attribution Test (PLAT) verifies how well the internal model explains actual daily trading results. If the model deviates too far from realised P&L, the test fails – and the affected trading desks must fall back to the less capital-efficient standardised approach. That is potentially expensive and operationally risky for a bank. During the relief phase, PLAT is downgraded from a binding test to a monitoring instrument only. A failure during the phase carries no consequences. This allows banks to keep using the internal model even when a single desk fails the test temporarily – which is the more realistic case in the early years than the clean textbook outcome.

RRAO: The bonus for structured product houses

Residual Risk Add-On (RRAO) is an additional capital surcharge that FRTB imposes for a narrowly defined group of exotic instruments: Bermuda options, spread options on Constant Maturity Swaps, instruments with realised volatility as underlying. The logic: These products carry residual risks that the standard risk schedule does not capture. During the relief phase, the RRAO for these exotic underlyings is suspended on a time-limited basis. For the few European houses with material exposure in this segment – Deutsche Bank, BNP Paribas, several specialised investment banks – this can move the profitability of entire business lines.

7 · Which houses benefit most

The relief does not act on all banks equally. The differentiation matters because it explains who has driven the consultation politically and who pushes hardest on the final calibration in the negotiations.

The largest beneficiaries are the three trading-intensive houses BNP Paribas, Deutsche Bank and Société Générale. Risk.net calculated in November 2025 that these three would carry risk-weight increases of 2.5 to 2.8 times their current Basel 2.5 level under the pure FRTB standardised approach. BNP Paribas confirmed this in its own consultation response of 17 December 2025: The three most-affected EU banks would face an FRTB capital requirement averaging more than double the current level. These banks are the primary target group of the bank-individual multiplier.

Mid-sized universal banks with moderate trading books – Commerzbank, the larger Landesbanken – see primarily the 0.9 multiplier and, where applicable, the bank-individual multiplier, but no material NMRF, PLAT or RRAO effects. Their relief is noticeable but not outsized. Houses with small trading books – savings banks, cooperative banks, mid-sized private banks – see primarily the 0.9 multiplier plus the explicit permission to continue using the Simplified Standardised Approach above the existing threshold instead of the full FRTB standard. For them, the act is a meaningful simplification but not a game-changer.

8 · What industry and supervisors say

The industry speaks with an unusually clear voice. ISDA and the Association for Financial Markets in Europe (AFME) submitted a joint response on 19 May 2026 – the final consultation day – welcoming the dual multiplier mechanic and flagging two areas for clarification. The European Banking Federation (EBF) had taken a broader line a month earlier, pointing out that European banking regulation runs up to 66 per cent above Basel minimums – through gold-plating and overlapping requirements. From an industry standpoint, the FRTB relief is a partial correction, not a gift.

Supervisory positions are split. The European Central Bank (ECB) had positioned itself in the spring of 2025 against a further FRTB postponement – arguing that a blanket delay penalises banks that had prepared for the date. José Manuel Campa, Chair of the European Banking Authority (EBA), framed the supervisory line more nuanced in an August 2025 interview with Les Échos: A further postponement was legally available, but he was »equally opposed to any deregulation«. This tension – supervisors sceptical of relief, industry and the Commission running on competitiveness arguments – defines the detail negotiations on final calibration.

9 · Status quo: What happens between today and application

The consultation has closed. The next step is the formal adoption of the act by the Commission – internally by college decision or written procedure. The Commission had announced 19 May 2026 as the adoption date, but as of the time of writing neither a press release nor a document in the Official Journal confirming adoption is available. That is procedurally normal – the formal adoption can slip by a few weeks without affecting substance.

After formal adoption, the three-month scrutiny window follows: European Parliament and Council can object. For Delegated Acts with broad consensus – and the multiplier mechanic is on balance well received by industry – an extension or even a veto are highly unlikely. A realistic expectation: Official Journal publication in the third quarter of 2026, entry into force in parallel with FRTB application on 1 January 2027.

The EBA must publish a report on FRTB implementation in third countries (US, United Kingdom and others) by 10 July 2026 at the latest. That report will be the empirical basis for a later, permanent legislative initiative by the Commission – an indication that the three-year relief is not the endgame. If the US has finalised its Basel III implementation by the end of 2029, the relief can lapse as planned. If the US continues to diverge, a second, then permanent, adjustment of the CRR3 market risk framework will be on the table – with the material risk that the EU permanently deviates from the Basel standard.

10 · What changes concretely by business line

The table below maps the impact by the product classes a trading desk actually runs. It is illustrative – the precise bank-individual effect depends on portfolio mix, model approach and the output floor interaction (see Section 11).

Business line	Raw FRTB effect	Relief relevant	Net effect for the desk
Government bonds and rates	Moderate rise (sensitivity-based standardised approach stricter than today)	Both multipliers apply in full	More profitable during the relief phase than today
Investment grade credit	Material rise (credit spread risk more heavily weighted)	Multipliers plus NMRF relief for illiquid issuers	More profitable than today, particularly in the secondary market
High yield / emerging markets	Sharp rise (NMRF dominates)	NMRF reduction to 35 to 45 per cent is the main driver	Business retained that would otherwise have been cut back
Equity and equity derivatives	Moderate to material rise	Multipliers apply; PLAT monitoring helps IMA houses	Broadly neutral to slightly positive
FX cash and forwards	Limited rise (FX risk treated conservatively)	Multipliers apply in full	Market relief in major currencies, tight margins in EM crosses
Structured products / exotics	Sharp rise (RRAO and NMRF act heavily)	RRAO suspension for defined exotic classes is the key	Business lines remain viable that would otherwise have been wound down
Commodities	Material rise (sensitivity-based)	Multipliers, no specific sweetener	Competitive disadvantage versus US houses partly remains

11 · The one open issue – output floor

A detail question decides for many banks whether the relief works in net terms or evaporates through the output floor. The output floor, also part of CRR3, sets a floor: The total risk-weighted-asset calculation of a bank using internal models must not produce less than 72.5 per cent of what the standardised approaches would produce (phased in, currently 50 per cent, rising through 2030). Because the output floor is based on FRTB standardised-approach risk weights, FRTB-driven increases in the standardised approach feed indirectly through the floor onto banks that model internally.

The unresolved question: Does the bank-individual multiplier reach the floor calculation, or does it apply only to the capital requirement before the floor is applied? BNP Paribas raised this explicitly in its consultation response as a technical point requiring clarification; ISDA and AFME picked it up in their May joint statement. If the multiplier does not reach the floor, the relief can largely run dry at standardised-approach-heavy banks – because the floor pulls market risk capital up before the multiplier scales it down. The final text of the act must decide this. Banks should prepare their capital planning for both variants.

12 · What sales, trading and trading treasury should do now

The relief is substantial, but it comes with detail requirements that cannot be hashed out in the last quarter before 1 January 2027. Four steps are central from the perspective of the trading franchise.