On March 19, 2026, the Federal Reserve, the Office of the Comptroller of the Currency and the Federal Deposit Insurance Corporation did something regulators almost never do. They re-proposed a rule to give capital back.
The three agencies published a fresh Basel III endgame package that, in aggregate, would cut required capital across the US banking system by roughly $87.7 billion. Comments closed on June 18. A final rule is expected late this year, with implementation in 2027. The Federal Reserve Board advanced the package on a 6-1 vote, with Governor Michael Barr the lone dissent.
That is a reversal, not a revision. The 2023 version of this same rule, drafted after the collapse of Silicon Valley Bank, would have raised capital at the largest US banks by about 19 percent. Take two does the opposite. And it arrives at the exact moment Britain is switching on the tougher framework the Americans just walked back. The Atlantic has split on bank capital.
What the re-proposal actually does
Strip out the rulemaking machinery and the March package is a targeted reduction in Common Equity Tier 1 requirements, weighted toward the smallest banks. The agencies’ own estimates put the reduction at roughly 4.8 percent for the global systemically important banks, about 5.2 percent for large regionals, and close to 7.8 percent for smaller institutions.
The tiering matters more than the headline dollar figure. The relief skews toward smaller banks, but the direct benefit to true community banks is thin: most already sit inside the Community Bank Leverage Ratio framework, which the endgame barely touches. The 7.8 percent line describes smaller institutions in general, not the corner grocer’s bank down the road. The real winners are the mid-tier and large regional banks that run their own capital models and have spent two years lobbying against the 2023 draft.
Barr’s dissent is the tell. As the architect of the original 2023 proposal, he argued the re-proposal trades away buffer that the 2023 bank failures showed was necessary. The majority’s counter, written into the release, is that overall system capital would decrease only “modestly” and would remain far above pre-2008 levels. Both statements are true. The disagreement is about how much margin a banking system needs when the next stress arrives, and neither side can prove its number.
London keeps the framework Washington dropped
The instinct is to read Britain as the mirror image, tightening hard while America loosens. That overstates it. The Prudential Regulation Authority’s final Basel 3.1 rules, published in PS1/26 this January, take effect on January 1, 2027. But the PRA calibrated them to be close to capital-neutral: it estimates Tier 1 requirements for major UK firms will rise less than 1 percent by 2030, once transition ends. The Financial Policy Committee has actually trimmed its benchmark Tier 1 requirement from 14 percent of risk-weighted assets to 13 percent.
So the divergence is not primarily about how much capital. It is about which framework governs it. Britain is implementing the full Basel machinery: the output floor that caps how far a bank’s internal models can diverge from standardized risk weights, the revised credit and operational risk approaches, the model constraints. The US re-proposal keeps the label and softens the machinery, easing the output floor’s bite and preserving more room for internal models at the banks best equipped to game them.
That is the split worth watching. Same accord, opposite philosophies. One jurisdiction is binding its largest banks to a common standardized floor. The other is deciding, for the second time, that its largest banks can be trusted with more discretion.
The forward story is 2027 lending, not the rulebook
Freed CET1 is not idle. Every dollar of capital a bank no longer has to hold against its assets is a dollar it can lever into new lending or return to shareholders. Roughly $87.7 billion of released requirement, run through normal capital-to-assets ratios, supports several hundred billion dollars of additional balance sheet capacity. The question for 2027 is where it goes.
If it flows into credit, US regional and large-cap banks enter next year with more room to expand commercial and real estate lending than their UK counterparts, who will be absorbing a new framework at the same moment. That is a genuine competitive asymmetry, and it lands just as US banks are already reporting the strongest net interest margins in years. If instead the capital flows to buybacks, the relief becomes a shareholder transfer with little effect on the real economy, which is roughly the outcome Barr fears.
The catch is timing. Capital relief is easiest to grant at the top of a cycle, when losses are low and the buffer looks like dead weight. It is worth the most at the bottom, when it is already spent. The 2023 proposal was a response to bank failures. The 2026 re-proposal is a response to two years of industry pressure and a changed political appetite for regulation. Nothing about the underlying risk of maturity transformation changed between the two drafts. Only the willingness to hold capital against it did.
Britain made the opposite bet, and did it deliberately: keep the constraining framework, calibrate the level to neutral, and preserve the option to tighten later. Washington kept the level flexible and the framework loose. By 2027 both approaches will be live at once, on banks that compete in the same wholesale markets. The final US rule, due late this year, will set exactly how much of the $87.7 billion is real. That number, not the transatlantic symbolism, is what the largest banks are already modeling into their 2027 plans.
Discussion
Sign in to join the discussion.
No comments yet. Be the first to share your thoughts.