Meta title: Basel III Endgame 2026: The Capital Rule That Reversed Meta description: US regulators proposed a 19% capital hike in 2023. The March 2026 reproposal delivers $87.7B in relief instead. What changed — and what banks should watch. Slug: /basel-iii-endgame-capital-rule-reversal/ Last updated: August 2026
Regulators spent 2023 designing a rule to make the largest US banks hold roughly 19% more capital. In March 2026, the same three agencies — the Fed, OCC, and FDIC — published a reproposal of that same rule delivering the opposite outcome: an estimated $87.7 billion in net capital relief for the industry. Same statute, same international accord, same regulators. Almost the exact opposite number.
The sharpest detail isn’t in the risk weights. It’s in the vote. The Federal Reserve Board advanced all three proposals 6–1, and the single dissenting vote came from Michael Barr — the architect of the original 2023 rule, still sitting on the Board, voting against the rewrite of his own work. That reversal, not the technical mechanics underneath it, is the real Basel III Endgame story. It says as much about how fragile “final” bank regulation has become as it does about capital adequacy.
Key Takeaways
- The original July 2023 proposal from the Fed, OCC, and FDIC would have raised aggregate capital requirements by roughly 16% across affected firms, with the largest banks facing closer to 19%.
- Michael Barr, the Fed’s Vice Chair for Supervision and chief architect of the 2023 proposal, announced on January 6, 2025 that he would step down from that role effective February 28, 2025. He remains a Fed governor. Michelle Bowman — a public critic of the original proposal’s scope — took over the supervisory role in June 2025.
- On March 19, 2026, the agencies issued three interrelated proposals recalibrating the framework: removing the “dual stack” approach, lowering minimum risk weights, adjusting the G-SIB surcharge, and reworking the market-risk (FRTB) framework — reversing course into net capital relief instead of an increase.
- The Board advanced the package 6–1, with Barr casting the lone dissent. The FDIC board voted unanimously in favor.
- The public comment period closed on June 18, 2026. Finalization is targeted for Q4 2026, with implementation beginning in 2027.
- The EU and UK are not following the same script. The EU is phasing in temporary mitigations through 2029, while the UK’s PRA finalized its own “Basel 3.1” package for January 2027 and pushed its trading-book internal models approach back to 2028 — and is still adjusting that piece as of mid-2026. A decade after the global Basel agreement, implementation is diverging, not converging.
A rule built by one Fed Vice Chair, rewritten by his successor
It’s tempting to read the Basel III Endgame reversal as a story about capital theory — risk-weight recalibration, market-risk modeling, G-SIB surcharge mechanics. It’s really a story about how contingent “final” financial regulation turns out to be on who happens to hold one supervisory chair.
Michael Barr built the 2023 proposal around the view that the largest banks were meaningfully undercapitalized relative to their risk. By September 2024 he had already conceded ground, floating a revised framework that cut the aggregate increase from roughly 19% to roughly 9% after more than 400 comment letters. Four months later he announced he was leaving the supervisory role. Michelle Bowman, who had gone on record opposing the scope of the original proposal before she took the job, assumed it in June 2025.
The rule didn’t just get delayed between 2023 and 2026 — it changed authors, and the new author held a documented, public disagreement with the old one’s premise. The 6–1 vote in March 2026 made that explicit: the person who wrote the rule was outvoted on what it became.
That matters beyond this one rule. A capital framework that can flip from a 19% increase to an $87.7 billion relief package across a single change in supervisory leadership is not, in any meaningful sense, “final.” Banks that spent 2024 and 2025 building compliance infrastructure for the original NPR were, in effect, preparing for a rule that would be substantially rewritten by the time it mattered.
What actually softened
The March 2026 relief isn’t a single change but a stack of smaller ones.
Regulators removed the “dual stack” framework that had required banks to calculate capital under two parallel methodologies and hold the higher result, replacing it with a single Expanded Risk-Based Approach. The market-risk framework — the Fundamental Review of the Trading Book — was recalibrated to make the Internal Models Approach genuinely usable: risk-factor eligibility and P&L attribution requirements were relaxed, and the output floor that would have forced standardized calculations even on IMA desks was removed. The threshold for what counts as “significant trading activity” was also raised, pulling smaller institutions out of scope entirely. The reproposal further eliminates the requirement to deduct mortgage servicing rights from regulatory capital, a change aimed at bringing mortgage activity back onto bank balance sheets.
Minimum risk weights came down, and the buffers and surcharges layered on top were made less conservative across the board. One thing did not soften: internal models for operational risk are still gone, and the standardized approach there is still coming. For operational risk teams, this remains a methodology-transition project, not a reprieve.
Regulators frame the package as better alignment between capital and actual risk rather than simply less capital — a defensible technical position, but one that happens to arrive at a friendlier number for the industry that lobbied hardest against the original proposal.
The global standard that stopped being global
Basel III was conceived, in the aftermath of 2008, as an internationally coordinated standard specifically so capital requirements wouldn’t become a competitive lever between jurisdictions. A decade into implementation, that coordination is visibly fraying.
The US is pursuing capital relief through a comprehensive reproposal. The EU postponed FRTB implementation to January 2027 and is managing impacts through temporary Level 2 measures designed to soften the transition between 2027 and 2029. The UK’s Prudential Regulation Authority finalized its “Basel 3.1” package in PS1/26 for January 2027 but deferred the internal-models approach for trading books to January 2028, citing the need for international coordination — and in June 2026 opened a further consultation (CP9/26, closing September 18, 2026) proposing targeted adjustments to that same IMA framework, while keeping the 2028 date.
Three major jurisdictions, three timelines, three degrees of relief — and in the UK’s case, a rule that was “final” in January and back under consultation by June. For a globally active bank, that isn’t a footnote. It’s a genuinely different capital planning problem in every region it operates.
What this means for bank leadership
Net capital relief sounds like an unambiguous win, and in the near term it likely is — more capital available for lending, buybacks, or dividends instead of sitting idle against risk-weighted assets. But two things should temper the enthusiasm.
First, the rule still isn’t finalized. Comments closed on June 18, 2026, and a final rule is expected in Q4 2026 at the earliest, with phased implementation from 2027. This framework has already reversed direction once under supervisory turnover, and the specific calibrations that matter most to individual balance sheets — operational risk ILM treatment, CRE risk weights, the G-SIB surcharge formula — are exactly the provisions most likely to move between proposal and final text. Treating the March 2026 numbers as locked in is premature.
Second, capital relief is arriving at the same moment banks are absorbing new, less-tested exposures — from stablecoin-adjacent nonbank charters to shifting deposit competition dynamics. That combination is worth watching rather than simply welcoming. A thinner capital cushion is a reasonable bet when the risk environment is stable. Whether 2026’s risk environment qualifies as stable is very much the open question.
This article is editorial analysis of proposed banking regulation and does not constitute legal, regulatory, or investment advice. Proposed rules are subject to change before finalization.
Add Comment