On the ninth of March 2026, Michelle Bowman, the Federal Reserve's Vice Chair for Supervision, announced a re-proposal of the Basel III Endgame that the American banking industry had sought, lobbied for, and publicly demanded since the summer of 2023 [1]. The original proposal, published jointly by the Fed, the Office of the Comptroller of the Currency, and the Federal Deposit Insurance Corporation in July 2023, would have raised aggregate capital requirements for the largest American banks by approximately nineteen per cent [2]. The new proposal targets an aggregate impact of zero to five per cent. Three years of intense lobbying, two administrations, one Federal Reserve chair confirmation, and an industry spending campaign that one commentator estimated at several hundred million dollars in direct advocacy have produced, in the end, a regulatory framework that will leave American bank capital requirements substantially where they currently stand.

The announcement was greeted with undisguised satisfaction by the major American banking institutions, whose Common Equity Tier 1 ratios, the primary measure of loss-absorbing capacity, have risen substantially since the post-2008 period of reconstruction and now sit, for most systemically important institutions, well above the minimum requirements that even the original 2023 proposal would have imposed [3]. On one reading of the situation, the capital-neutral outcome is a reasonable acknowledgement of this reality: banks that have already rebuilt their capital bases to levels that provide meaningful buffers against stress do not require regulatory mandates to accumulate further quantities of capital that would otherwise be available for lending. On another reading, it is the product of a political moment in which the calculus of financial stability has been subordinated to the calculus of economic stimulation, and the error, if it is one, will become visible only when the next severely adverse scenario arrives.

The Long Arc of the Basel Endgame

The history of Basel III's American implementation is, in some respects, a history of regulatory ambition colliding with the practical difficulty of translating international standards into a domestic political economy that has never been reliably hospitable to the project of coordinated bank supervision. The Basel III framework itself was agreed in September 2010, in the aftermath of a financial crisis whose central lesson, as interpreted by the Financial Stability Board and the Basel Committee on Banking Supervision, was that the major banks of the world had been insufficiently capitalised to absorb the losses that the combination of complex structured products and institutional interconnectedness was capable of generating [4]. The framework's implementation in the United States was gradual, contested, and repeatedly modified, a pattern that the Endgame re-proposal brings to its logical conclusion.

The specific provisions at issue in the 2023 Endgame proposal were primarily those relating to market risk, operational risk, and the treatment of trading book assets under the Fundamental Review of the Trading Book, the component of the international standard that would have most significantly affected the capital requirements of the large trading and investment banking operations at Goldman Sachs, JPMorgan, Morgan Stanley, and their equivalents [5]. The industry's objection was not primarily to the principle of higher capital but to the specific risk-weighting methodologies, which it characterised as failing to reflect the actual risk characteristics of the relevant instruments and as constituting what both industry representatives and some sympathetic academics described as "gold-plating": the application of international standards in a form more onerous than those standards required or than other major jurisdictions were applying. This was not an entirely disingenuous argument. European and British regulators were, by 2025, implementing their own Basel III Endgame frameworks with modifications that differed meaningfully from the American 2023 proposal, and the competitive implications of an American framework substantially more burdensome than its international equivalents were real and documentable [6].

What Remains in the Re-Proposal

The capital-neutral framing should not obscure the fact that the March 2026 re-proposal is not a complete withdrawal from the Basel III Endgame project. The framework retains elements of the Fundamental Review of the Trading Book, in modified form, and preserves the requirement for enhanced operational risk capital treatment for the largest and most complex institutions [1]. What it removes, or substantially reduces, are the components that would have had the largest aggregate impact on capital levels: the expanded scope of standardised approaches, the recalibrated risk weights for residential mortgage exposures, and the most burdensome treatment of equity and credit valuation adjustments. The aggregate impact, as Vice Chair Bowman described it, is zero to five per cent, which, for institutions whose current capital ratios already provide substantial buffers above the effective minimum, is operationally close to zero. The industry's conclusion that it has won is not unreasonable, and Major banks are expected to reinvest accordingly in their trading and mortgage servicing operations, reversing the scaling back that had been occurring in anticipation of higher requirements [7].

The implementation timeline is equally significant. The Fed expects to finalise the technical rules by the fourth quarter of 2026, with implementation beginning in 2027 [1]. This means that the stress tests scheduled for later in 2026 will occur under the existing framework, not the new one, and the results of those tests will constitute the first public evaluation of the thesis that current CET1 levels are adequate to absorb a severely adverse scenario without requiring the additional buffers that the original Endgame proposal was designed to provide. That is, to say the least, an interesting sequencing.

The Divergence From Europe

What has received less attention in the American discussion than it deserves is the extent to which the capital-neutral outcome widens the regulatory gap between the United States and its major international counterparties. The European Union began implementing its version of the Basel III Endgame, the Capital Requirements Regulation III (CRR3), on the first of January 2025 [6]. The CRR3 incorporates a phased introduction of the output floor, the provision that limits banks' ability to use their own internal risk models to reduce capital requirements below a standardised floor, rising from fifty per cent of the standardised approach in 2025 to seventy-two and a half per cent by 2030. The European Banking Authority's own impact assessment estimated a capital shortfall of 124.8 billion euros for European institutions required to build toward the full floor [8]. European banks are, in other words, building capital at the same time as American banks are being released from the obligation to build it.

The competitive implications of this divergence are not straightforward. American banks will carry lower capital requirements relative to risk-weighted assets than their European counterparts for at least the next several years, which provides a structural advantage in the pricing of risk and the generation of return on equity. This may attract business from European to American institutions in markets where the two compete directly, most obviously in complex structured products, leveraged finance, and the trading book activities that the Fundamental Review of the Trading Book was specifically designed to address. The Atlantic Council's analysis of Basel fragmentation noted that regulatory arbitrage of this kind does not necessarily reduce systemic risk; it relocates it, concentrating activity in the jurisdiction with the lighter regulatory touch and thereby increasing the potential consequences of a failure in that jurisdiction [9].

The Case for the Capital-Neutral Position

The strongest argument for the Bowman re-proposal is the one that the Vice Chair herself advanced, which is that the existing capital levels of the American G-SIBs are, by any historical and international standard, substantial. The Common Equity Tier 1 ratios of the largest American banks currently range from approximately eleven to fifteen per cent, well above the seven per cent minimum (including the capital conservation buffer) that Basel III established as the bedrock requirement [3]. Even allowing for the G-SIB surcharge that applies to the most systemically important institutions and the countercyclical buffer that regulators can activate during periods of excess credit growth, the existing buffers represent a dramatically different position from the five to eight per cent ratios that characterised the same institutions in 2008. The capital rebuilding of the post-crisis period, achieved through a combination of regulatory pressure and the organic accumulation of retained earnings during a decade of profitability, was real and substantial.

Furthermore, the relationship between regulatory capital requirements and bank behaviour is not linear in the direction that simple analysis suggests. There is credible evidence, developed in the empirical literature on bank capital and credit supply, that capital requirements above a certain threshold do not increase safety by a commensurately meaningful amount, but do reduce credit availability in ways that affect the real economy [10]. If American banks' current capital ratios already provide genuine resilience against severe stress scenarios, as the annual stress test results have consistently suggested, then requiring additional capital beyond those levels imposes costs on credit supply and economic activity without generating proportionate safety benefits. This is a legitimate analytical position, not a rationalisation, and it has a respectable literature behind it.

The Question the Stress Tests Will Answer

The difficulty is that the adequacy of current capital levels is a thesis that rests heavily on the assumption that the stress scenarios used to calibrate it are sufficiently severe. The Federal Reserve's stress tests have, since 2009, used scenarios designed to capture "severely adverse" macroeconomic conditions, including very large increases in unemployment, significant declines in asset prices, and substantial contractions in economic activity [11]. Whether those scenarios adequately capture the specific failure modes that the Basel Endgame rules were designed to address, including the concentrated losses that can accumulate in complex trading books under conditions of market liquidity stress, is a question that has divided analysts along predictable lines, with the industry arguing that internal models capture these risks adequately and regulators arguing that they do not.

The March 2026 re-proposal essentially accepts the industry's position. The 2026 stress tests will be the first major empirical test of whether that acceptance was warranted. They will not, of course, generate a crisis to test the thesis; they will generate a model-based assessment of resilience under hypothetical adverse conditions. But the results will matter, both as a public signal of whether regulators are comfortable with the capital levels that the new framework will permit and as a political fact that will shape the next iteration of the argument, which, given the cyclical history of Basel implementation in the United States, will eventually come.

Walter Bagehot's observation that a well-run bank should be one whose solvency is not in question under any plausible adverse scenario was written in an era when a bank's capital consisted of the personal fortunes of its partners and its resilience was therefore a matter of straightforward accountability. The capital adequacy framework that replaced that accountability with quantitative requirements has never fully resolved the question of what level of requirement is genuinely prudent rather than merely conventional. The American decision to hold capital requirements at current levels, and to test that decision against whatever conditions 2026 and the years beyond it bring, will generate a great deal of data. Some of it will be reassuring. The rest will depend on what happens.