Board of Governors of the Federal Reserve System

I'd like to thank staff for their hard work in developing the proposals before the Board today. The package includes many provisions that I have supported in the past, as part of the Board's July 2023 capital proposals and in my September 2024 speech discussing the potential re-proposal of these rules, and which I continue to support.

However, I cannot support any of the three proposals as presented today. Bank capital rules help to ensure that banks fund themselves with capital commensurate with the risks of their activities and the risks that they pose to the U.S. financial system. Today's proposals, when combined with the impact of proposed changes to the Board's stress tests, would lower the common equity tier 1 capital requirements of the largest banks by 4.8 percent. When those changes are considered in combination with the recent changes to the enhanced supplementary leverage ratio (eSLR), the impact on global systemically important banks (G-SIBs) is even larger: a decrease in tier 1 capital requirements of 6.0 percent, which is $60 billion.

These significant reductions in capital requirements are unnecessary and unwise. The capital surcharge for G-SIBs could be refined and the Basel III reforms could be adopted in the United States without materially weakening the capital framework. Today's proposals, if adopted, would harm the resilience of banks and the U.S. financial system.

G-SIB surcharge
Starting with the proposal to revise the calculation of the G-SIB surcharge, I support certain aspects of the proposal that would better align a firm's surcharge with its systemic risk, such as calculating the surcharge based on annual averages rather than year-end values and assigning surcharges based on narrower scoring bands. However, the proposal would make certain modifications to the G-SIB surcharge that are not adequately supported, and that would lead to unwarranted capital requirement reductions for the very largest banks. These changes could have a negative impact on financial stability and the competitive landscape.

For example, the proposal would re-calibrate the G-SIB framework so that the short-term wholesale funding component would account for 20 percent of G-SIBs' aggregate method 2 scores. Banks' reliance on short-term wholesale funding can leave firms vulnerable to runs that undermine financial stability. When the Board adopted its G-SIB framework, it stated that it intended for the short-term wholesale funding component to comprise 20 percent of method 2 scores for G-SIBs. As it turns out, however, G-SIBs are even more reliant on short-term wholesale funding than initially thought, meaning that the component has made up approximately 30 percent of firms' method 2 scores. The proposal reasons that the appropriate response to this would be to reduce the impact of short-term wholesale funding on the method 2 score calculation, but there is a different (and more consequential) conclusion: if this type of funding is so much more prevalent than the Board initially suspected, that just means the problem the Board was trying to deal with was bigger than initially believed. Further, banks' reliance on short-term wholesale funding, relative to their risk-weighted assets, has grown from approximately 30 percent to 40 percent between 2016 and 2024. Given these facts, there is little to be said in favor of reducing the impact of this type of funding on firms' G-SIB surcharges.

As another example, the proposal would make a one-time downward adjustment to the factors that determine the surcharge, otherwise known as the method 2 systemic indicator coefficients, to offset an observed divergence in growth rates between the G-SIBs' method 1 and method 2 G-SIB scores since 2019. While technical in nature, the proposed adjustment to method 2 coefficients is significant because it replaces principled, risk-based calibration with arbitrary alignment to a different, and less stringent way to calculate the surcharge: the method 1 approach. This has the harmful effect of eroding the purpose of having distinct measures of systemic risk and weakening the foundation of capital rules. In the context of a package increasing the resilience of the banking system, I could support an approach updating scores in a principled way, but I cannot see a basis for the type and extent of adjustment included in this proposal. Method 1 and method 2 are intended to measure systemic risk differently—that is the point of having two methods—and there is no reason to expect them to necessarily move in tandem, or to make adjustments when they don't.

The G-SIB surcharge proposal would cut common equity tier 1 capital requirements for the largest banks by 3.8 percent, or $33 billion. I cannot see a justification for such a reduction.

Large bank proposal
I appreciate the progress in moving forward on implementing the Basel III reforms in the United States. Adopting those standards would be a crucial step to making the U.S. and global financial systems safer and ensuring that U.S. banks are seen by their foreign regulators as being subject to sufficiently strong standards. Many aspects of today's proposal are consistent with the Basel III agreements, and our prior proposals, and I support those components of the rule. However, the proposal includes far too many downward deviations from the international accord that would result in a much weaker capital framework. I fear that, if this much weaker version of Basel III is adopted in the U.S., it could trigger a "race to the bottom" on standards, harming the global financial system. There are over 20 material downward deviations from the Basel III standard.

I won't comment on all these downward adjustments but will focus on a few. One goal of the Basel III reforms was to address the undercapitalization of market risk—the risk resulting from exposure to price movements caused by changes in market conditions, market events, and issuer events that affect asset prices. In particular, the reforms aimed at capturing the risks associated with financial instruments such as collateralized debt obligations (CDO), credit default swaps, and mortgage-related securitizations, which became apparent during the global financial crisis. The market risk framework agreed to in the Basel III accord is meant to ensure that banks have sufficient capital to guard against the risks associated with trading these types of instruments.

Given these goals, it is important that the market-risk related aspects of the Basel III reforms are implemented as fully as possible in the United States. Instead, the proposal meaningfully deviates downward from those standards. For example, the internationally agreed upon Basel III reforms include both a standardized and an internal models approach for computing market risk capital requirements, but include an "output floor," based on the standardized approach, to ensure that a firm's use of its own internal models does not result in unwarranted reductions in capital requirements. The U.S. proposal before us today rejects that output floor entirely. This will encourage other jurisdictions to do the same, undermining a key reform and cornerstone of the Basel III agreement. Going even further, the proposal, instead, would use the output of the standardized approach as a cap on a firm's market risk capital requirements, rather than a floor.

As another example, the proposal would reduce, relative to Basel, the capital penalty for a bank's use of market risk factors with insufficient market price data ("non-modellable risk factors"). A different part of the Basel III framework seeks to deal with some of the risks inherent in securitized assets (such as CDOs). Under the Basel III standards, only securitizations that meet certain structural and transparency standards are eligible for a preferential "p-factor," a parameter that scales up or down the amount of capital a bank needs to hold with regard to a particular securitization. Today's proposal would extend that preferential treatment to all securitizations.

Each individual deviation from Basel III is perhaps able to be rationalized on its own, but in the aggregate, they result in a significantly weaker capital framework. Worse, when combined with the proposed changes to the stress test framework, changes to the market risk capital requirements would decrease overall capital requirements by 5.8 percent for the largest banks. The Basel III accord was intended to strengthen banks' ability to withstand risks from their trading activities, to protect against another crisis. Instead, capital requirements for those activities are being significantly relaxed.

The capital impact associated with today's proposals is much more significant as a result of recent changes to the eSLR. Proponents of the changes to the eSLR argued that those changes would have minimal effects, given the strength of the risk-based capital framework. As I discussed in my dissent to that final rule, the effects of that rule would be much starker if the risk-based framework were subsequently relaxed, as is being proposed today. For the largest bank holding companies, today's proposals result in aggregate tier 1 capital requirement reductions that are 60 basis points greater than they would have been without the recent eSLR changes; for the largest depository institutions, the difference is 230 basis points.

The impact of today's proposal is also exacerbated by the fact that the largest banks would be required to compute their risk-weighted assets under only one approach, rather than two, as under the current rule. This "single stack" approach reduces complexity, but only slightly; the largest banks could easily continue to calculate their capital requirements under the relatively simple standardized approach. Doing so would be valuable, because the different approaches have differing ways of assessing risk, which can be important to ensuring that banks are adequately capitalized. And a dual stack also ensures compliance with the U.S. Collins Amendment.

Standardized approach proposal
Another of today's proposals would modify the standardized approach in the Board's capital rule, which would generally apply to all banks except for the very largest. As part of this proposal, category III and IV banking organizations would be required to include elements of accumulated other comprehensive income (AOCI) in common equity tier 1 capital, consistent with the treatment already applicable to larger firms. I support this change, which was part of our earlier package, as it would help to ensure that the capital requirements for these banks accurately reflect their capacity to absorb losses, such as from interest rate risk.

However, I cannot support other aspects of the standardized approach proposal. As a whole, the proposal would weaken capital requirements for all banks. For Category III and IV firms—large banks whose failure could impose significant externalities on the U.S. banking and financial system—the proposal would reduce capital requirements by 3 percent, or 5.2 percent when combined with the effects of the stress test changes. For banks under $100 billion, the proposal would decrease capital requirements by almost 8 percent. Overall, this is not a proposal targeted to help community banks, but rather a proposal that helps some of the biggest banks in the country.

Generally speaking, these reductions in capital requirements are not justified. The standardized approach proposal incorporates aspects of the large-bank proposal that lower risk-weights but largely does not incorporate those aspects of the large-bank proposal that raise risk-weights. This does little to make the standardized approach more risk-sensitive or a more accurate reflection of the riskiness of a bank's assets; it mostly just makes the capital requirements lower.

Conclusion
Today's proposals add to a growing list of actions that materially weaken the Board's capital requirements, particularly for the largest banks. The stress test proposal, the eSLR final rule, and today's proposals collectively would leave the U.S. banking and financial system in a more vulnerable position.

Capital requirements are being significantly reduced. There are suggestions that liquidity requirements could also be reduced. Additionally, Federal Reserve supervisory staff have been cut by over 30 percent, and supervisory practices have been weakened.¹ Banking is built on trust. I worry greatly that these actions are rapidly eroding that trust. I dissent.

1. See Michael S. Barr, "The Case for Strong, Effective Banking Supervision" (speech at the Kogod School of Business, American University, Washington, DC, November 18, 2025). Return to text

i. On March 19, 2026, an erroneous reference regarding capital requirements of $23 billion was corrected to $33 billion.