FRTB Explained: The Fundamental Review of the Trading Book

The Fundamental Review of the Trading Book, almost universally shortened to FRTB, is the Basel Committee's overhaul of how banks calculate capital requirements for market risk. It was designed after the 2008 financial crisis exposed serious weaknesses in the previous framework, and it remains one of the more technically demanding topics in bank risk management, and a regular feature of FRM and PRM exam syllabuses.

What problem FRTB was built to fix

The pre-crisis market risk framework had several well-documented weaknesses. The boundary between a bank's "trading book" (positions held for short-term trading) and its "banking book" (positions held to maturity) was loosely defined enough that banks could shift exposures between the two to get more favourable capital treatment. Value-at-Risk (VaR), the standard risk measure at the time, consistently understated the risk of extreme, "tail" market moves. Liquidity assumptions were static and unrealistic, typically using a flat 10-day horizon regardless of how genuinely liquid a position actually was. And risk models varied so much between banks that outputs weren't meaningfully comparable. FRTB was built to address all four issues at once.

Two ways to calculate capital: SA and IMA

Under FRTB, every bank must be able to calculate its market risk capital using the Standardised Approach (SA), which applies prescribed risk weights and correlations set by regulators. The SA itself has three components: a sensitivities-based approach, a default risk charge, and a residual risk add-on for exotic or hard-to-model exposures. The SA is mandatory for every bank, and it also acts as the fallback for any trading desk that fails to qualify for, or loses approval for, the more sophisticated alternative.

That alternative is the Internal Models Approach (IMA), which lets banks use their own risk models, subject to meeting strict, continuously monitored performance standards. Approval under IMA is no longer granted firm-wide; it's assessed and can be revoked at the level of an individual trading desk, which significantly raises the bar for maintaining model approval across a large trading operation.

Expected Shortfall replaces VaR

One of FRTB's most significant technical changes is replacing VaR with Expected Shortfall (ES) as the core risk measure under the Internal Models Approach. Where VaR estimates the loss that won't be exceeded at a given confidence level, Expected Shortfall measures the average of all losses beyond that point, giving regulators and banks a clearer picture of how bad the tail of the loss distribution actually is. This is more computationally demanding than VaR, but it was a deliberate response to VaR's well-known tendency to understate genuine tail risk.

P&L attribution testing and non-modellable risk factors

Two further mechanisms give FRTB real teeth. The profit and loss attribution (PLA) test requires banks to compare, on a regular basis, the P&L their risk model would have predicted against the P&L their front office actually generated; too much unexplained divergence and a desk can lose its IMA approval. Separately, non-modellable risk factors (NMRFs) are risk factors, often for less liquid markets, that a bank can't demonstrate sufficient reliable pricing data for. Where a risk factor is classified as non-modellable, the bank must hold a specific stress-calibrated capital add-on for it, rather than folding it into the general internal model.

Where implementation actually stands

FRTB has had a long and uneven rollout since it was finalised by the Basel Committee. In the EU, an adjusted version of the framework is set to apply for a three-year period from 1 January 2027, subject to the usual parliamentary and council scrutiny process. In the UK, the internal models approach specifically has been delayed further, to 1 January 2028, with UK banks continuing to rely on existing model approvals in the meantime. In the US, no firm FRTB implementation date has been set; the comment period on the related Basel III proposal ran through mid-2026, with final timing still to be confirmed. The practical result is that FRTB implementation dates now differ meaningfully by jurisdiction, and anyone working across an international trading operation needs to track more than one timeline.

Why FRTB matters beyond the regulatory requirement

FRTB sits alongside Expected Shortfall as one of the core quantitative risk concepts underpinning modern market risk management, and it builds on the same Basel capital framework covered in our guide to Basel IV's evolution. For finance professionals working toward FRM, PRM, or a career in market risk, understanding how SA and IMA differ, why Expected Shortfall replaced VaR, and how the PLA test and NMRF add-ons actually bite in practice is genuinely useful groundwork, not just an exam topic.

FAQ

Is the Standardised Approach mandatory for all banks? Yes. Every bank must be able to calculate capital under the SA, even if it also uses the Internal Models Approach for some desks, since SA acts as the fallback whenever IMA approval isn't in place.

Why did Expected Shortfall replace VaR under FRTB? VaR consistently understated the risk of extreme market moves beyond its confidence threshold. Expected Shortfall measures the average loss in that tail, giving a more complete picture of downside risk.

Has FRTB fully come into force yet? Not everywhere. The EU's adjusted framework starts in January 2027, the UK's internal models approach is delayed to January 2028, and the US has yet to confirm a firm date, so implementation currently varies by jurisdiction.

FRTB is dense, but the underlying logic is straightforward: force banks to measure tail risk more honestly, tighten model governance down to the desk level, and stop capital arbitrage between the trading and banking books. That's a meaningful shift from the pre-crisis framework it replaced.