U.S. Banking Agencies Propose Sweeping Overhaul of Capital Framework Part I: GSIBs and Category I Banking Organizations

SUMMARY ANALYSIS: WHAT BANKS SHOULD KNOW NOW

The proposals. The first proposal (the "ERBA Proposal") would replace the existing dual-calculation framework with a single standardized methodology, the so-called Expanded Risk-Based Approach ("ERBA"), while fundamentally restructuring capital requirements for credit risk, operational risk, market risk, and credit valuation adjustment ("CVA") risk. (The agencies issued a second joint proposal, revising the current standardized approach in the capital rules, but this proposal does not apply to GSIBs.) The Fed's proposal would recalibrate the GSIB surcharge framework to better align surcharges with actual systemic risk.

Net aggregate capital relief, with significant firm-level variation. The combined proposals are projected to reduce CET1 requirements for Category I and II holding companies by approximately 4.8% (~$42.1 billion), but a standard deviation of approximately 8.7% across the GSIB population means individual outcomes will vary widely. GSIBs should model impacts across all major risk-weighted asset categories and their interaction effects on the stress capital buffer, GSIB surcharge, total loss-absorbing capacity ("TLAC"), and long-term debt requirements.

Fundamental simplification through the single-stack ERBA. Eliminating the dual-calculation requirement and advanced approaches simplifies compliance, but the new explicit operational and CVA risk requirements will increase risk-weighted assets for the most trading-intensive GSIBs on a standalone basis.

Potential benefit from a models-based approach for organizations with large trading businesses. The ERBA Proposal increases market risk-weighted assets by approximately 20%, and CVA and operational risk charges for trading activities would increase overall trading-related risk-weighted assets by approximately 31%—partially offset by proposed stress test revisions. Firms should assess desk-by-desk impacts and weigh the costs and benefits of seeking supervisory approval for the models-based approach versus defaulting to the standardized measure.

Fundamental restructuring of GSIB surcharge framework. The combination of a one-time coefficient reduction, annual GDP indexing, short-term wholesale funding ("STWF") recalibration, and narrower surcharge bands represents the most significant revision to the framework since 2015, with the average surcharge projected to decline from 2.7% to 2.3%. STWF recalibration effects will vary materially across GSIBs, and the shift to 10-basis-point increment bands will translate future score changes more directly into surcharge adjustments.

More favorable treatment of mortgage servicing assets. The proposal would eliminate the existing threshold-based deduction of mortgage servicing assets ("MSAs") from CET1 capital and replace it with a uniform 250% risk weight applicable to all MSA holdings. The agencies note that, by removing a strong disincentive for mortgage origination and servicing, this change is expected to significantly incentivize these banking organizations to reenter residential mortgage markets long dominated by nonbank lenders.

Partial (and disputed) Basel III alignment. The proposal deviates from the international standard in both directions—most notably by rejecting the output floor for market risk—raising questions about competitive parity and the risk that other jurisdictions may respond by relaxing their own frameworks. 

EXPANDED ANALYSIS

The ERBA Proposal

A single set of risk-based capital requirements. The most structurally transformative element of the ERBA Proposal is the elimination of the dual-calculation requirement and its replacement with a single framework: the ERBA. The ERBA is a fully standardized methodology—eliminating internal model use for credit and operational risk entirely—but retains model-based approaches for market risk where high-frequency data permit robust calibration.

Credit risk. Key changes for Category I banking organizations include the following:

• Residential Mortgages: Risk weights would be determined by loan-to-value ratios, ranging from 20% to 70% for standard residential mortgage exposures, and from 30% to 105% for mortgages dependent on cash flows generated by the underlying property.
• Corporate Exposures: The ERBA would differentiate capital requirements for corporate borrowers based on assessed creditworthiness rather than the single flat 100% risk weight in the current standardized approach. The agencies estimate that approximately 53% of Category I banking organization balance-sheet corporate exposures would qualify as investment-grade at a 65% risk weight, resulting in a meaningful reduction in credit risk-weighted assets for investment-grade corporate lending portfolios.
• Retail Exposures: Risk weights for retail lending would be calibrated to reflect borrowers' repayment history, adding a dimension of risk sensitivity absent from the current standardized approach.
• MSAs: The proposal would eliminate the existing threshold-based deduction of MSAs from CET1 capital. Currently, MSAs above 10% of CET1 capital must be deducted. Under the proposal, all MSAs would be assigned a 250% risk weight, as is the current risk weight for MSAs only up to the 10% threshold. This would act as an incentive for banks to increase their participation in residential mortgage markets.

Operational risk. The ERBA would introduce an explicit standardized operational risk capital requirement—a significant structural addition, as the current standardized approach that binds most Category I banking organizations today includes no operational risk capital requirement. The requirement would be calibrated to each institution's business volume, with a 70% reduction in the contribution of investment management, investment services, and nonlending treasury services income to reflect their historically lower observed operational risk.

Market risk and credit valuation adjustment risk. The market risk framework would be revised for all Category I and II depository institution holding companies, as well as other banking organizations meeting the proposed significant trading activity threshold of more than $5 billion in aggregate trading assets and liabilities (raised from the current $1 billion threshold) or trading activity equal to or greater than 10% of total assets. Key structural features include:

• Replacement of the VaR-based measure of market risk with an expected shortfall-based measure to better capture tail losses under stressed conditions.
• Introduction of a standardized approach for market risk as the default methodology for all covered banking organizations, alongside a trading desk-level models-based approach requiring supervisory approval.
• Notably, the proposal would not incorporate the Basel III output floor for market risk; instead, the standardized approach output would serve as a cap (rather than a floor) on firms' market risk capital requirements—a material downward deviation from the international standard.

The CVA framework would apply to Category I and II holding companies with significant trading activity and other banking organizations with significant trading activity and at least $1 trillion in notional derivative exposure; client-facing derivative transactions are exempt, providing targeted relief for commercial end users.

Interaction with stress testing and impact on TLAC. The adjustments to risk-weighted assets under the Expanded Risk-Based Proposal would also flow through to TLAC and long-term debt requirements for GSIBs. The combined effect of the Expanded Risk-Based Proposal and the GSIB Surcharge Proposal is estimated to reduce aggregate TLAC requirements for GSIBs by approximately $46 billion (or 2.6%) and long-term debt requirements by approximately $17 billion (or 2.3%), providing meaningful additional flexibility to these firms over their TLAC composition.

Standalone capital impact. On a standalone basis, the ERBA Proposal is projected to increase aggregate CET1 capital requirements for Category I and II bank holding companies by approximately 1.4% and tier 1 capital requirements by approximately 1.6%, driven primarily by the inclusion of new operational and CVA risk-weighted assets and an approximately 20% increase in market risk-weighted assets.

The GSIB Surcharge Proposal

The Fed separately proposed targeted amendments to the GSIB surcharge framework, addressing longstanding structural criticisms of the current method 2 calculation and its divergence from method 1 scores. The average applicable GSIB surcharge across the eight current U.S. GSIBs is projected to decline from 2.7% under the current framework to approximately 2.3% under the proposal.

One-time downward adjustment to method 2 coefficients. The method 2 fixed coefficients—which scale GSIB systemic indicator values for size, interconnectedness, complexity, and cross-jurisdictional activity—would be subject to a one-time downward adjustment by a factor of 1.2. This adjustment reflects the observation that, since 2020, method 2 scores have grown approximately 20 percentage points more than method 1 scores (which are calibrated to global aggregate indicators) due to pandemic-era balance sheet expansion—suggesting that the recent increase in method 2 scores has been driven by macroeconomic factors rather than changes in the systemic risk profiles of individual GSIBs.

Annual indexing to nominal GDP growth. Going forward, the method 2 coefficients would be adjusted annually based on a three-year moving average of nominal U.S. GDP growth, ensuring that GSIB scores reflect changes in systemic risk relative to the broader economy rather than being mechanically inflated by economic growth and inflation.

Short-term wholesale funding score recalibration. The proposal would modify the STWF indicator by removing the risk-weighted assets denominator from the score calculation, measuring the STWF category as an absolute dollar amount, and recalibrating the resulting coefficient to target approximately 20% of aggregate method 2 scores across U.S. GSIBs—consistent with the original design intent at the framework's adoption in 2015 but which has drifted to approximately 30% due to initial data limitations.

Data averaging and cliff effect reduction. The proposal would require GSIBs to calculate certain systemic indicators—including intra-financial system assets and liabilities, securities outstanding, and cross-jurisdictional claims and liabilities—as annual averages of daily or monthly values rather than point-in-time year-end figures, reducing incentives for temporary year-end balance sheet adjustments. Method 2 surcharge increments would also be reduced from 50 basis points to 10 basis points (through narrower 20-point score band ranges, compared to the current 100-point bands) to reduce cliff effects and improve the surcharge's sensitivity to changes in firms' systemic risk profiles.

Aggregate surcharge impact. The GSIB Surcharge Proposal is estimated to reduce net aggregate GSIB CET1 requirements by approximately 3.8%, corresponding to a $23 billion (or 10%) reduction in aggregate GSIB surcharge dollar amounts. Under the proposal, method 1 surcharges would become binding for two of the eight current U.S. GSIBs, compared to none under the current framework where method 2 is universally binding.

Cumulative Capital Impact

When the ERBA Proposal and the GSIB Surcharge Proposal are considered together with the recently finalized eSLR changes and the pending stress test modifications, the cumulative projected CET1 impact for Category I and II banking organizations is as follows:

Dissenting Views

The Fed approved the proposals by a vote of 6-1, with Governor Michael S. Barr dissenting. Governor Barr argued that the combined effect amounts to a $60 billion reduction in tier 1 GSIB capital requirements—one he characterized as "unnecessary and unwise"—and expressed particular concern that the NPR incorporates over 20 material downward deviations from the Basel III standard, including the rejection of the output floor for market risk, warning that this framework could trigger a "race to the bottom" internationally.