Skip to content
QuantCalc Tools

Thematic Pillar Hub · 1 Published Briefings

Quantitative Risk & Capital Modeling

Quantitative Risk & Capital Modeling institutional intelligence, quantitative benchmarks, primary regulatory filings, and market structure analysis from QuantCalc Tools. Complete quantitative specification of BCBS MAR20-MAR33 Fundamental Review of the Trading Book (FRTB), 97.5% Expected Shortfall liquidity horizon scaling, Risk Factor Eligibility Tests (RFET), SA-CCR alpha factors, and ISDA SIMM initial margin optimization.

Curated by Elena Rostova, FRM, CQF (FRM, Certificate in Quantitative Finance (CQF)) · Fact-checked by QuantCalc Tools Institutional Review Board

Key Institutional Takeaways — Quantitative Risk & Capital Modeling

  • FRTB Internal Models Approach (MAR33) replaces 10-day 99% VaR with a liquidity-adjusted 97.5% Expected Shortfall (ES) cascade scaled across 10, 20, 40, 60, and 120 business-day liquidity horizons via marginal square-root-of-time weights (1.00, 1.414, and 2.449).
  • Cross-asset correlation benefits inside the Internals Model Capital Charge (IMCC) are restricted by the supervisory parameter rho = 0.50, blending unconstrained empirical ES with the linear sum of five standalone asset-class ES charges (GIR, CSR, EQ, COM, FX).
  • Under the Risk Factor Eligibility Test (MAR31.12), every modellable risk factor bucket requires >=24 verifiable real-price observations per year (with >=4 per 90-day window) or >=100 observations in 12 months; failing factors incur an un-diversifiable Non-Modellable Risk Factor (NMRF) Stressed ES charge with rho_NM = 0.60.
  • Trading desks must pass quarterly P&L Attribution Tests (PLAT, MAR32.20) requiring Spearman rank correlation r_s > 0.80 and Kolmogorov-Smirnov distance KS < 0.09 to remain in the Green Zone; falling into the Red Zone (r_s < 0.70 or KS > 0.12) forces immediate fallback to the FRTB Standardized Approach.
  • ISDA SIMM v2.6/v2.7 computes 99% 10-day MPOR bilateral initial margin across four strictly additive product silos (RatesFX, Credit, Equity, Commodity), creating strategic cross-margin optimization opportunities against SA-CCR (1.4 * (RC + PFE)) and CCP clearing via multilateral compression.

All Quantitative Risk & Capital Modeling Research Briefings (1)

2026-10-22Elena Rostova, FRM, CQF

Basel III Endgame FRTB & ISDA SIMM v2.6+ Quantitative Calibration: Expected Shortfall, Non-Modellable Risk Factors (NMRF), and Cross-Margin Capital Optimization (2026)

Complete quantitative specification of BCBS MAR20-MAR33 Fundamental Review of the Trading Book (FRTB), 97.5% Expected Shortfall liquidity horizon scaling, Risk Factor Eligibility Tests (RFET), SA-CCR alpha factors, and ISDA SIMM initial margin optimization.

Interactive Quantitative Calculators & Models

All Calculators →

Frequently Asked Questions — Quantitative Risk & Capital Modeling

How does the BCBS MAR33 Liquidity-Adjusted 97.5% Expected Shortfall formula avoid running separate 120-day historical simulations?

Under MAR33.4, banks compute all Expected Shortfall shocks over a uniform 10-business-day base horizon (ES_10d) and partition risk factors into five nested subsets Q(RF, j) corresponding to supervisory liquidity horizons LH_j in {10, 20, 40, 60, 120} days. For each horizon j >= 2, risk factors with shorter liquidity horizons are held constant while longer-horizon risk factors are shocked over 10 days, scaled by sqrt((LH_j - LH_{j-1}) / 10), and aggregated via root-sum-of-squares.

What triggers a Non-Modellable Risk Factor (NMRF) capital charge and why is it punitive for exotic derivatives desks?

A risk factor becomes an NMRF if it fails the Risk Factor Eligibility Test (RFET) under MAR31.12—requiring at least 24 committed trades or firm executable two-way quotes per year (with at least 4 per 90-day period) or 100 observations per year—or if it represents uncaptured interpolation basis between observable curve buckets. NMRFs are carved out of the diversifiable ES engine and stressed over a 12-month worst-case window with a 0.60 linear correlation constraint, often accounting for 30% to 55% of exotic desk capital.

Why can't an equity index total return swap offset the interest rate delta of an interest rate swap under ISDA SIMM v2.6+?

Under BCBS-IOSCO Uncleared Margin Rules, ISDA SIMM partitions all trades into four mutually exclusive Product Silos: RatesFX, Credit, Equity, and Commodity. While sensitivities net across risk classes within a single product silo, total SIMM Initial Margin is the strict linear sum of the four product silos. Therefore, interest rate delta inside an Equity silo trade cannot offset interest rate delta inside a RatesFX silo trade unless the exposures are restructured or cleared.

How does posting segregated ISDA SIMM Initial Margin impact SA-CCR Exposure at Default (EAD) and Margin Valuation Adjustment (MVA)?

Under SA-CCR (CRE52), EAD equals 1.4 * (RC + PFE). For the party receiving initial margin, segregated collateral reduces the PFE multiplier toward the 5% floor. However, for the posting party, bankruptcy-remote third-party custodial initial margin does not reduce Net Independent Collateral Amount (NICA) against Replacement Cost and simultaneously generates a permanent Margin Valuation Adjustment (MVA) funding drag over the life of the netting set.

Explore Related Thematic Research Hubs on QuantCalc Tools

Quantitative Risk & Capital Modeling — Institutional Research & Regulatory Hub | QuantCalc Tools | QuantCalc Tools