U.S. Regulators should raise -- not lower -- big bank capital requirements
The following post – co-authored with our friend and colleague, Richard Berner (NYU Stern School of Business) – was submitted in August 2025 as a comment to the U.S. bank regulatory agencies regarding their proposed modifications to the leverage ratio standards for U.S. global systemically important banks.
To Whom It May Concern:
We write to oppose the agencies’ proposal to alter the enhanced Supplementary Leverage Ratio (eSLR).
In our view, the proposal substantially weakens leverage, total loss-absorbing capacity, and long-term debt requirements for global systemically important banks (GSIBs). As a result, it would reduce key safeguards implemented in response to the 2008 financial crisis and add to the risks of serious financial instability and taxpayer-funded GSIB bailouts that Congress and the agencies sought to eliminate.
For the following reasons, we urge you to withdraw the proposal and prioritize implementing the Basel Endgame risk-based capital rules in a way that preserves the strength of existing capital requirements.
When policymakers propose a major change in the bank capital regime, they should conduct and publish in advance a comprehensive quantitative impact study (QIS) of its financial and economic impact (see, for example, here). Opponents of the 2023 Basel Endgame proposal criticized the lack of such a comprehensive QIS when regulators proposed to raise capital requirements. The same criticism applies to this proposal to lower capital requirements. To be sure, there is no legal requirement for a QIS. Nevertheless, a careful and independent QIS would illuminate the economic and financial trade-offs associated with changes in bank capital requirements. It would focus on the entire capital regime (including risk-weighted requirements and stress tests), rather than narrowly on the leverage requirement. We expect that such a study would foster public discussion and serve as a guide (and a goal) for policy action.
The goal of making the leverage ratio a backstop is misplaced. The leverage ratio should be complementary to the risk-based capital ratio, not a mere backstop. Together with stress tests, these ratios establish the prevailing capital regime for the large banks, so that it is inappropriate to change one without carefully accounting for the impact on the regime as a whole. For example, lowering the leverage ratio sufficiently so that it almost never binds encourages arbitrage: in response, banks likely will dynamically adjust their assets to reduce their risk-based capital requirements. Such adjustments may well proceed until the leverage ratio is again binding, but with overall capital levels lower than before. The agencies’ static economic analysis ignores this dynamic element.
Even if the leverage ratio were to be used as a backstop, policymakers should raise the risk-adjusted capital requirement, rather than lower the leverage ratio. The GSIBs’ average leverage ratio peaked in 2016 (see Figure 1). It declined markedly even before the pandemic, so that the decline cannot be attributed primarily to quantitative easing or the post-2019 wave of Treasury issuance. Rather, the timing suggests that (following the Dodd-Frank Act of 2010 and the Basel III implementation of 2013) it took several years for GSIBs to learn how to increase risk-taking without concomitant capital increases. To address such regulatory arbitrage, and to restore the banking system’s safety to the 2016 level, it would be better to raise risk-adjusted capital requirements, rather than lower the leverage requirement. This change also may encourage GSIBs to shift from riskier activities to Treasury market intermediation, in line with the current proposal’s purported aims. (For other reasons why policymakers should raise (rather than lower) capital requirements on GSIBs, see here.)
Figure 1. U.S. GSIB Average Leverage Ratio (Quarterly or Semi-annually),
June 2002-December 2024
Notes: The chart shows the Tier 1 leverage ratio for the eight US banks designated as global systemically important banks. Shaded areas denote NBER-dated recessions. Source: Federal Reserve Bank of Kansas City Bank Capital Analysis (4Q 2024, Chart 1).
The proposal will substantially erode bank capital. The agencies’ analysis reveals that the proposed rule would cut tier 1 capital requirements for the GSIBs’ depository institution subsidiaries by 27 percent, or by $213 billion. To put this in perspective, banks heavily criticized the 2023 Basel Endgame proposal for increasing capital requirements of 38 large banks by $170 billion, yet the leverage ratio proposal would reduce requirements by an even larger amount for a subset of these banks.
The proposal significantly underestimates likely capital depletion. While the agencies acknowledge the proposal would reduce capital requirements for GSIBs’ bank subsidiaries, they claim holding company risk-based capital rules will prevent GSIBs from paying out substantial dividends and buybacks to shareholders. The agencies estimate BHC-level capital requirements would drop by only 1.4 percent ($13 billion), at which point risk-based capital requirements become the binding constraint, limiting how much capital GSIBs could distribute. However, this analysis is seriously flawed for two key reasons. First, as noted above, holding companies will likely dynamically adjust their balance sheets, optimizing their risk-weighted assets to the point that the leverage ratio again becomes binding. Second, the analysis ignores the interaction with other developments, such as the Fed’s stress test changes, that have begun to allow BHCs to distribute more capital than before. Finally, the proposal counts on capital support in a period of stress from the holding companies to their depository subsidiaries. Experience, however, suggests that support in a crisis flows in the opposite direction. One reason is the holding companies’ need to support their broker/dealer subsidiaries.
GSIB behavior shows they are not capital constrained. Following the announcement of the Federal Reserve stress test results in June, the GSIBs increased their dividends by an average of 12.3 percent and announced a total of $70 billion of share buybacks (see here). Rather than use this capital to boost purchases of Treasuries, the GSIBs chose to distribute it to shareholders.
The proposal would undermine a key feature of the leverage ratio: its risk-insensitivity. Unlike the risk-based capital requirements, the leverage ratio does not currently rely on risk assessments. Consequently, it is less vulnerable to risk misjudgments by banks and supervisors. However, the proposal would tie each firm’s eSLR to its risk-based GSIB surcharge, effectively undermining the leverage ratio’s insensitivity to risk misjudgments.
If banks respond to the proposal by acquiring more long-duration Treasuries, it could lead to systemically dangerous risk concentration. The current risk-weighting scheme is insensitive to interest-rate risk: banks need no more capital to hold a 30-year Treasury bond than to hold a Treasury bill or a reserve at the central bank. Thus, if reducing the eSLR motivates GSIBs to acquire long-dated Treasurys (rather than central bank reserves or short-dated Treasurys), it could lead to a common exposure that threatens financial stability. The spillovers from Silicon Valley Bank’s 2023 collapse, largely as a result of excessive interest rate risk, foreshadow this danger.
In contrast to its stated aim, the proposal may not boost activity in the Treasury market. There is little evidence that reducing leverage requirements will create incentives for banks to increase their Treasury market and Treasury repo activity. As Federal Reserve Governor Michael Barr noted in his dissent, banks could easily use their freed-up balance sheet capacity for other low-risk activities that offer higher returns than Treasury trading and repo. Moreover, as Barr explained, if banks are bound by capital requirements in good times, they will not have unused balance sheet capacity when markets turn volatile.
Other regulatory tools can foster smoother Treasury intermediation while improving risk-management incentives. There are much better tools available to improve Treasury market functioning, such as those identified in recommendations stemming from the work of the Interagency Working Group on Treasury Market Reform. These include expanding central clearing of Treasury collateral and repo, improving data quality and availability, enhancing trading venue transparency and oversight, and examining the effects of risk management practices.
As a further alternative to lowering the eSLR, regulators could exclude bank reserves and short-dated Treasury instruments from its denominator, subject to an offsetting increase in the required eSLR ratio. The agencies’ economic analysis suggests that a similar policy alternative (which excludes reserves and all Treasury securities from the denominator) would increase GSIBs’ intermediation capacity by more than the agencies’ actual proposal, while nevertheless reducing the incentive to boost leverage. Consistent with the systematic approach to the bank capital regime that we favor, if the agencies were to exclude reserves and short-dated Treasuries from the ratio denominator, they should also adjust – i.e., increase – the applicable leverage ratio requirement, much as the Bank of England did in 2017, to compensate for the reduction of assets in the denominator. Importantly, we oppose excluding medium- and long-dated Treasury securities from the leverage ratio denominator, because that would encourage GSIBs to expand interest-rate-risk taking in a potentially systemic fashion (as previously described).
Acknowledgement: We thank Jeremy Kress for very helpful comments and suggestions. Professor Kress’ comment letter is available here.