Status and the correct comparison
On March 19, 2026 the banking agencies released a revised capital package. It includes a Basel framework for the largest banking organizations, changes to the standardized approach for other banks, and a Federal Reserve G-SIB surcharge proposal. Comments were due June 18, 2026. The materials reviewed for this September 27 article describe proposals, not a final effective capital rule. [1]
Do not carry forward a headline estimate from the 2023 proposal or a 2024 speech as though it measures the March 2026 text. Equally, do not describe all proposed capital reductions as the effect of Basel alone. The components move in different directions, and separately proposed stress-test changes alter the combined picture.
Who would face which framework?
The Basel proposal principally addresses Category I and II firms. It would replace the existing overlapping credit-risk calculations with a single expanded approach. The standardized proposal addresses other banking organizations. Significant trading activity can bring additional firms within proposed market-risk requirements; the fact sheet identifies $5 billion of trading activity or 10% of assets as relevant thresholds. [1, 2]
Recommended scoping starts with the legal entity, prudential category and activity thresholds. A regional lender’s exposure to the standardized changes is different from a large trading bank’s exposure to market-risk reform. Consolidated parent results also need reconciliation to subsidiary constraints before anyone estimates capital available for distributions or loan growth.
What changes economically
The agency fact sheet describes more differentiated treatment of mortgages and other exposures, revisions to operational and trading risk, changes in mortgage-servicing-asset treatment and a five-year AOCI transition for Category III and IV firms. The G-SIB proposal refines score bands and measurement. These are distinct mechanisms, not a uniform percentage haircut to every loan. [2]
The proposed standardized rule is the right place to test exposure eligibility, attributes and assigned weights. Portfolio classification and data quality will matter. A generic business-loan label is not sufficient evidence that an exposure receives a favorable category. [3]
Analytically, implementation can increase reporting and data costs even where required capital declines. Mortgage attributes, borrower characteristics, commitments and trading positions need traceable mappings. A lower aggregate requirement does not prove that a particular card, mortgage or warehouse portfolio benefits.
Reading the impact estimates without mixing denominators
The Federal Reserve staff memorandum estimates that, for Category I and II firms, the Basel component increases required common equity tier 1 capital by 1.4%, while the G-SIB component reduces it by 3.8%, producing a combined 2.4% reduction. Including separately proposed stress-test changes produces a 4.8% reduction. For Category III and IV firms, the standardized package including AOCI produces an estimated 3.0% reduction, or 5.2% including the stress changes. [4]
These are aggregate percentage changes in estimated required capital, not percentage-point reductions in regulatory capital ratios, individual-bank forecasts or already available cash. Different mixes of assets, AOCI and binding constraints can produce substantially different outcomes.
Scroll horizontally to see all columns.
| Comparison | Staff estimate | Interpretation |
|---|---|---|
| Category I/II: Basel component | +1.4% | Increase in estimated required CET1 |
| Category I/II: Basel plus G-SIB | −2.4% | Combined package, excluding separate stress changes |
| Category I/II: also include proposed stress changes | −4.8% | Broader combined scenario |
| Category III/IV: standardized plus AOCI | −3.0% | Aggregate; individual bank outcomes differ |
Worked example: capital efficiency is not lower credit loss
Illustrative calculation using hypothetical weights, not a classification conclusion: a $100 million exposure pool at a 100% risk weight produces $100 million of risk-weighted assets. At 90%, it produces $90 million. With an assumed 10% capital target, the associated capital falls from $10 million to $9 million. The difference is $1 million of capital capacity, not $10 million of new cash.
At an assumed 12% annual cost of equity, that $1 million represents $120,000 of annual capital-cost sensitivity. A 20-basis-point deterioration in annual losses on the $100 million pool is $200,000 and would outweigh it. Funding, expenses and taxes further affect returns. The example explains why capital relief cannot justify weaker underwriting by itself.
Nor is the $1 million necessarily distributable. Leverage requirements, stress losses, subsidiary requirements or a management buffer may be binding. A bank must identify the actual limiting constraint and rerun its capital plan before turning a risk-weight calculation into a lending or payout decision.
Why policymakers disagree
Vice Chair Bowman supports the package as better matching requirements to risk, addressing overlap with stress testing and reducing incentives for traditional lending to migrate outside banks. Her argument emphasizes calibration and credit availability. [5]
Governor Barr dissented. He argues that the combined reductions are not justified, that some departures weaken the international framework and that the interaction with leverage and stress changes matters. He supports some individual elements, including more risk-sensitive features and AOCI recognition, while opposing the package as a whole. [6]
The practical disagreement is about the resilience purchased by an additional dollar of capital versus its effect on intermediation. My assessment is that aggregate release estimates are insufficient to settle it. Evidence should examine stress losses, funding fragility, correlated exposures and whether lower requirements actually translate into durable lending capacity.
Implementation priorities and watch points
Recommended priorities are a versioned current-versus-proposed RWA engine, an exposure-data gap assessment, reconciliation to regulatory reports and stress testing at both consolidated and subsidiary levels. Keep CECL allowance assumptions, regulatory capital effects and economic expected loss distinct so the business case does not double count benefits.
Watch final rules, transition dates, changes to the standardized and trading calibrations, the G-SIB methodology and separate stress-test decisions. This article should be revised when authoritative final text changes the scenario. Evidence that would strengthen the credit-growth thesis includes bank-specific headroom under all binding constraints and an attractive risk-adjusted loan pipeline. Weak demand or worsening vintages could absorb the benefit even if the eventual rule reduces capital requirements.
Sources
- Federal Reserve and banking agencies: March 19, 2026 capital proposalsBack to text: ↑
- Agencies: March 2026 capital proposal fact sheetBack to text: ↑
- Federal Register: March 27, 2026 standardized-approach proposalBack to text: ↑
- Federal Reserve staff memorandum: scope and estimated capital effectsBack to text: ↑
- Vice Chair Bowman: supporting statement, March 19, 2026Back to text: ↑
- Governor Barr: dissenting statement, March 19, 2026Back to text: ↑