Part III: The G-SIB Surcharge Re-Wired

In Part I, we established that the four pillars of the March 19, 2026 package form one coordinated capital architecture. In Part II, we decomposed the four risk stripes — credit, FRTB, operational risk, and CVA — and revealed that the 17-percentage-point spread in CET1 outcomes across G-SIBs is driven entirely by business model. We showed that the G-SIB surcharge NPR accounts for −3.8% of the cumulative CET1 impact — the single largest source of capital reduction in the package — and that understanding why required a dedicated deep-dive into the surcharge mechanics.

Part III delivers that deep-dive. It moves from the strategic frame and risk-stripe analytics of the first two instalments into the precise engineering of the G-SIB surcharge — and makes the case that most commentary is still treating this NPR as a side-show when it is, in fact, where a significant portion of capital leverage now sits. For a Category I firm with RWAs in the low trillions, every 50-basis-point change in the G-SIB surcharge translates into billions of CET1. Getting the mechanics right is not a technical curiosity — it is a strategic capital allocation question.

Five mechanics that deliver the 3.8%

The paper unpacks the five interlocking changes in the G-SIB surcharge NPR: the Method 2 coefficient recalibration with automatic annual GDP indexation; the STWF reset back to 20% with RWAs removed from the denominator; the shift from point-in-time year-end snapshots to daily/monthly averages; the move from 50bp to 10bp surcharge increments; and technical refinements to specific systemic indicators. Each is traced through its mechanical impact on surcharge scores — and each affects different G-SIBs asymmetrically depending on funding structure, trading book composition, and cross-jurisdictional footprint.

Key points of the paper:

• STWF is the quiet driver of the 3.8%. The short-term wholesale funding indicator had drifted from its 20% design target to approximately 30% of Method 2 scores — not because funding risk had increased, but because the interaction of an RWA-based denominator with fixed coefficients and FRTB-driven RWA inflation turned STWF into a stealth amplifier of trading-book capital. For Goldman Sachs and Morgan Stanley, the same risk was being hit twice — once through RWAs, again because those RWAs appeared in the STWF denominator. The NPR removes RWAs from the denominator entirely and recalibrates the coefficient, restoring STWF to what it was always supposed to measure: pure reliance on short-term wholesale funding.
• Average-based indicators end the December diet permanently. The move from a single December 31 snapshot to daily/monthly averages eliminates the incentive for year-end balance-sheet window-dressing — the repo compression, prime brokerage reshuffling, and derivatives notional management that has distorted Q4 pricing and Treasury market liquidity for over a decade. Short-term shrinkage on December 31 barely moves the needle under an annual average. Quarter-long and year-long positioning is what matters now.
• A new feedback loop links FRTB to the G-SIB surcharge in real time. FRTB-driven trading book decisions — simplifying structured credit positions, shifting from bilateral to cleared execution, reducing gross notionals — now continuously reshape FR Y-15 systemic indicators, which feed into G-SIB surcharge scores, which in turn affect eSLR buffers. Under the old point-in-time system, these were managed once a year. Under averaging, the loop runs daily. FRTB implementation at a G-SIB is no longer a market risk silo — it is a capital architecture problem that must be owned across Risk, Treasury, and the Front Office simultaneously.
• Franchise-specific impacts diverge sharply. Trading-heavy G-SIBs (Goldman Sachs, Morgan Stanley, Citi) see the largest STWF relief but must replace the December playbook with continuous repo and PB path management. Universal banks (JPMorgan, Bank of America) benefit from the STWF de-linkage — growing traditional lending no longer carries a secondary G-SIB penalty — but face the most operationally significant transition to daily averaged indicator management given the scale of their repo and Treasury operations. Custody-dominated institutions (BNY Mellon, State Street) find their largest capital levers in intraday liquidity and SFT businesses, not FICC trading — and the real question is whether they use the de-duplication headroom to lean into low-risk intermediation or allow capital to passively exit.

The CRO playbook

The paper closes with a concrete five-point playbook: rebuild the internal G-SIB model around the new STWF and averaging mechanics; embed the FRTB–G-SIB–eSLR feedback loop into trading-book strategy; redesign repo and prime brokerage economics for time-weighted capital; run binding-constraint analysis under the new four-pillar stack; and treat the comment period as a design lever — not a formality — with quantitative, desk-level evidence on STWF calibration, averaging methodology, and indicator construction.

Fill in the form below to receive the full paper — including the STWF mechanical breakdown, the FRTB → FR Y-15 → G-SIB → eSLR feedback loop analysis, the franchise-by-franchise impact assessment for all eight U.S. G-SIBs, and the CRO playbook for year-round balance-sheet strategy.