FRTB and Claude AI: SA, IMA, and Basel IV Market Risk Capital Workflows (2026)

FRTB: The Basel IV Market Risk Framework

The Fundamental Review of the Trading Book (FRTB), implemented under Basel IV, is the most significant overhaul of bank market risk capital in two decades. It replaced the Basel 2.5 VaR-based framework with two calculation approaches — the Standardised Approach (SA) and the Internal Models Approach (IMA) — and introduced new requirements for trading book / banking book boundary, P&L attribution testing, non-modellable risk factors (NMRF), and Default Risk Charge (DRC). For trading desks, risk teams, and bank capital management functions, FRTB creates substantial documentation, calculation, and reporting workload.

Claude is used in FRTB workflows for the documentation-intensive tasks: sensitivity methodology explanations, NMRF identification and assessment, P&L attribution testing analysis, capital ratio narrative for regulatory reporting, and management presentations explaining FRTB capital impacts. The quantitative calculations (sensitivity calculations, ES models) remain in risk systems, but Claude handles the surrounding documentation and analysis layer efficiently.

Standardised Approach (SA) Workflows

The FRTB SA calculates capital as the sum of sensitivities-based method (SBM) capital, residual risk add-on (RRAO), and DRC. For most trading books, the SBM is the largest component. It requires calculating delta, vega, and curvature sensitivities to prescribed risk factors (interest rates, credit spreads, equity, FX, commodity) and aggregating them using specified correlation assumptions within and across buckets.

• "FRTB SA delta sensitivity methodology for interest rate risk: We need to document our delta sensitivity calculation for GIRR (General Interest Rate Risk) under FRTB SA. Our portfolio includes: EUR fixed-rate sovereign bonds (duration 3-7 years), EUR interest rate swaps (5Y, 10Y, 30Y), and EUR cap/floor positions. Risk factors under FRTB SA GIRR: parallel shifts, steepener, fanning, and tenor-specific factors at 0.25Y, 0.5Y, 1Y, 2Y, 3Y, 5Y, 10Y, 15Y, 20Y, 30Y vertices. Write the sensitivity methodology documentation covering: (1) how we compute delta sensitivities for each instrument type to each tenor vertex, (2) the parallel/steepener/fanning decomposition, (3) DV01-based calculation approach, (4) how we aggregate across the bucket using the specified FRTB correlation matrices, (5) the resulting risk weight application. Format for inclusion in our FRTB SA methodology document."
• "FRTB SA RRAO identification: We need to assess which positions in our trading book are subject to the Residual Risk Add-On. Positions to assess: (1) exotic equity options with correlation payoffs (rainbow options, spread options), (2) FX knock-in/knock-out barrier options, (3) variance swaps on equity indices, (4) quanto options, (5) digital options across asset classes. RRAO applies to instruments with exotic underlyings or payoffs not captured by the SA SBM framework. For each instrument: (1) assess whether RRAO applies, (2) identify the RRAO notional base, (3) apply the prescribed risk weight (1% for exotic underlying, 0.1% for gap risk), (4) calculate the RRAO capital charge."
• "FRTB SA vs IMA capital comparison for approval: Our trading desk capital committee needs a comparison of our SA capital charge vs. our modelled IMA capital for Q2 2026. SA capital (from our systems): $142M. IMA capital (ES-based): $89M. IMA approval status: approved for FX and rates desks, pending for credit trading. Write the management presentation section covering: (1) SA vs IMA capital difference by desk, (2) which desks benefit most from IMA approval, (3) NMRF capital add-on breakdown, (4) capital benefit of completing the IMA approval process for credit trading, (5) P&L attribution test results and what they mean for IMA eligibility."

Internal Models Approach (IMA): P&L Attribution and NMRF

The IMA replaces Basel 2.5 VaR with an Expected Shortfall (ES) model at 97.5% confidence and introduces two critical new requirements: P&L Attribution (PLA) testing to verify that the risk model captures the actual P&L drivers, and Non-Modellable Risk Factor (NMRF) identification for risk factors without sufficient real price observations. Both create significant documentation requirements.

• "P&L Attribution test explanation for audit committee: Our FRTB IMA P&L attribution test for Q2 2026 shows: Spearman correlation between risk-theoretical P&L and hypothetical P&L = 0.89 (threshold ≥ 0.80 = Green zone). Kolmogorov-Smirnov test statistic = 0.085 (threshold < 0.09 = Green zone). Both metrics in Green zone → desk retains IMA eligibility for the quarter. One desk (EM credit) shows Spearman correlation = 0.78 → Amber zone, requires enhanced monitoring and explanation. Write the audit committee memo explaining: (1) what P&L attribution testing is and why FRTB requires it, (2) what the Spearman and KS test metrics measure, (3) why the EM credit desk is in Amber and what it means for our capital, (4) what remediation steps we are taking."
• "NMRF identification and capital assessment: We need to document our NMRF identification process for our credit default swap portfolio. Under FRTB, a risk factor is non-modellable if we cannot demonstrate at least 24 real price observations in the previous 12 months with no gap longer than 1 month. Risk factors in question: (1) single-name CDS spreads for EM corporate issuers with less frequent trading (5Y CDS on 3 Brazilian corporates), (2) CDS index tranches (correlation inputs — CDX.IG and iTraxx tranches), (3) spread curves beyond 10Y tenor for IG single-name CDS. For each: assess modellability using our observation dataset, document the rationale, and calculate the NMRF stress scenario capital using the prescribed floor (the IMA ES at the liquidity horizon)."

Default Risk Charge (DRC)

The DRC captures jump-to-default risk in credit instruments and replaces the Incremental Risk Charge (IRC) from Basel 2.5. It applies to positions with credit risk including corporate bonds, sovereign bonds, equity (via issuer default), and CDS. The DRC is calculated at 99.9% confidence over a one-year horizon and cannot be offset by hedges unless they are recognized under the DRC netting rules.

• "DRC methodology memo for IG corporate bond portfolio: Our DRC calculation covers a $2.1B IG corporate bond portfolio (long-only fund client). DRC components: (1) gross JTD loss for each obligor = bond notional × LGD, where LGD = 0.75 for senior unsecured, (2) net JTD after recognizing eligible CDS hedges per FRTB DRC netting rules, (3) bucket assignment by credit quality: IG, HY, defaulted, (4) within-bucket aggregation with 50% net long / net short offset recognition for IG, (5) DRC capital = weighted sum across buckets. Write the DRC methodology section for our FRTB IMA documentation: cover gross JTD calculation, LGD parameters, hedge eligibility and netting rules, bucket aggregation, and the capital floor at 50% of gross DRC."

FRTB Regulatory Reporting and Capital Ratios

FRTB market risk capital forms part of Pillar 1 RWA under Basel IV. Banks must disclose SA capital quarterly and disclose IMA capital (where approved) in Pillar 3. The narrative supporting the capital disclosures — explaining quarter-over-quarter changes, NMRF movements, and material risk drivers — is where Claude accelerates the reporting cycle.

• "FRTB Pillar 3 MR1 table narrative: Our Q2 2026 MR1 table shows: SA market risk capital $142M (Q1: $128M, +$14M QoQ). IMA market risk capital $89M (Q1: $84M, +$5M). Total market risk RWA $1.78B. Write the Pillar 3 narrative covering: (1) drivers of the $14M SA capital increase — FX positions grew as we added EM exposure, GIRR positions increased ahead of anticipated rate volatility, (2) the $5M IMA increase driven by higher ES in the rates book, (3) NMRF capital: $8.2M (stable QoQ), (4) DRC: $12.1M (down $3M following portfolio repositioning). Tone: factual, disclosure-quality, suitable for publication in Pillar 3 report."

FRTB Implementation Resources

For banks and dealers implementing FRTB or preparing for examination on their FRTB framework, the Compliance & Risk category includes tools for capital ratio analysis, regulatory documentation, and stress testing narrative. For market risk VaR and Greeks calculation, the Quant & Trading category has the Options Greeks Calculator and Portfolio VaR Engine. Related reading: Bank Regulatory Capital AI covers CET1, RWA, and Basel III leverage ratio workflows.