Ladies and Gentlemen:
The Bank Policy Institute[1] and the U.S. Chamber of Commerce[2] submit this letter in response to the Board of Governors of the Federal Reserve System’s notice of proposed rulemaking relating to the Federal Reserve’s stress capital buffer requirement.[3]
In light of the current flaws in the stress testing framework, which result in considerable volatility in the SCB and overall capital requirements, we urge the Federal Reserve to expeditiously adopt a final rule on the averaging proposal. As described in Section II.B, given the significant uncertainty introduced by the proposal’s comment period ending immediately before firms receive their 2025 stress tests results and—in accordance with well-established market conventions—announce their indicative SCBs and planned capital distributions, we also urge the Federal Reserve to announce (prior to firms receiving their results of the 2025 stress tests) that firms will be permitted to operate under the existing SCB framework through September 30, 2026. Doing so would eliminate considerable uncertainty in firms’ capital planning processes and avoid the introduction of additional volatility in SCBs. We also urge the Federal Reserve to include asymmetric averaging in the final rule and allow, but not require, a firm to opt in to the revised SCB framework prior to October 1, 2026. Because of the existing deficiencies in the stress testing models and scenario development, discussed in Section I below, it is essential that the Federal Reserve implement the averaging proposal with our proposed changes to mitigate the volatility resulting from these deficiencies for the 2025 stress testing cycle.
The proposal represents a constructive first step in the Federal Reserve’s efforts to improve its overall stress testing framework and bring the framework into compliance with the law, but there is more work to be done. Accordingly, in response to Question 1 of the proposal and the Federal Reserve’s stated intention to engage in additional rulemakings related to the stress testing framework, this letter begins with non-exhaustive general comments on the stress testing framework and then provides comments on the specific revisions to the framework and focused questions in the proposal.
I. Comments Related to the Overall Stress Testing Framework
Although the proposal would reduce some of the volatility in the SCB, it does not resolve significant deficiencies in the current stress testing framework or the legal flaws in the current framework that are the subject of litigation in the Southern District of Ohio.[4] We welcome the Federal Reserve’s recent commitment to issue proposed rules by October 2025 that would disclose and seek comment on the stress test models and scenarios to be used in the 2026 stress tests; ensure notice and comment annually on the stress test scenarios and on any material changes to the models; and seek comment on a scenario design framework that includes objective standards for the annual development of the scenarios.[5] Below are some recommendations for improving the overall stress testing framework and addressing the root causes of volatility in the SCB. However, due to the information asymmetry that exists today, we are inherently limited in the feedback we can provide. We therefore look forward to providing additional feedback through future notice-and-comment proceedings.
A. Stress test models, scenarios and results should be fully transparent.
First, stress test models and scenarios should be subject to full notice and comment, as required by the Administrative Procedure Act, including full disclosure of the models. This process will allow the public the opportunity to provide feedback on models and scenarios, enable the Federal Reserve to improve the models and scenarios based on the public’s views and practical experience, and provide regulated parties adequate notice as to how their legal obligations will be determined. Below, we make several suggestions for how the stress testing models and scenarios could be improved, but we are inherently limited in our ability to provide helpful feedback by the current lack of transparency concerning the Federal Reserve’s models and scenario design processes. Among other problems, this lack of transparency has led to unpredictability and volatility in SCBs from year-to-year, which the Federal Reserve seeks to reduce through averaging. The ultimate answer to these deficiencies is a fully transparent model and scenario design process that notifies the public of the standards the Federal Reserve is applying and allows the public to provide input to help refine the stress testing models and scenarios, including the global market shock and largest counterparty default components.
Second, the Federal Reserve has publicly and consistently stated that it does not depart from the results of the stress tests in calculating a firm’s SCB.[6] The Federal Reserve’s Stress Testing Policy statement provides, unequivocally, that “[t]he Federal Reserve does not make firm-specific overlays to model results used in the supervisory stress test. This policy ensures that the supervisory stress test results are determined solely by the industry-level supervisory models and by firm-specific input data.”[7] However, the Federal Reserve recently represented in litigation that it does not always base a firm’s SCB solely on the models and current-year data.[8] If the Federal Reserve believes that it needs the flexibility to apply either a firm-specific overlay (i.e., when the firm’s results would not be based solely on the public models and current-year data) or an industry-wide overlay (i.e., a common adjustment to every firm’s results), it must issue for notice and comment and subsequently adopt a rule that (i) establishes the criteria for the application of an overlay, (ii) outlines the process by which the Federal Reserve determines whether to apply an overlay, and (iii) provides affected firms with adequate notice of the application of an overlay and an opportunity to appeal the Federal Reserve’s decision.
B. Improvements to the pre-provision net revenue (“PPNR”) model components are necessary to correct errors in the models.
As noted above, the Federal Reserve has committed to proposing for public comment the models used in stress testing by October 2025. We expect to provide detailed technical comments on the PPNR models once published and we are able to review the models in full. However, at a minimum, the Federal Reserve should modify its PPNR model components to reduce volatility and improve their accuracy.[9] The Federal Reserve should also revise PPNR modeling to eliminate any losses that are double counted in both PPNR and the GMS component. Further, the risk characteristics of financial products and markets have shifted since the enactment of post-crisis regulatory reforms, most notably the elimination of proprietary trading and subprime securitizations. The Federal Reserve should revise PPNR modeling to ensure it does not produce inaccurate projections under the erroneous assumption that the risk characteristics of products and markets have remained static since the financial crisis.
Read the full letter here.
[1] The Bank Policy Institute is a nonpartisan public policy, research and advocacy group that represents universal banks, regional banks, and the major foreign banks doing business in the United States. The Institute produces academic research and analysis on regulatory and monetary policy topics, analyzes and comments on proposed regulations, and represents the financial services industry with respect to cybersecurity, fraud, and other information security issues.
[2] The U.S. Chamber of Commerce is the world’s largest business federation. It represents approximately 300,000 direct members and indirectly represents the interests of more than three million businesses and professional organizations of every size, in every industry sector, and from every region of the country.
[3] See Federal Reserve, Modifications to the Capital Plan Rule and Stress Capital Buffer Requirement, 90 Fed. Reg. 16,843 (Apr. 22, 2025).
[4] Those legal vulnerabilities have been explained at length in the plaintiffs’ complaint and brief in Bank Policy Institute et al. v. Board of Governors of the Federal Reserve Systems, Case No. 2:24-cv-04300, (S.D. Ohio), which has been temporarily stayed in light of the Federal Reserve’s commitments to implement reforms to the stress tests in the near term. Plaintiffs’ complaint and brief are incorporated by reference in this letter.
[5] Joint Motion to Stay Proceedings, Bank Policy Institute et al. v. Board of Governors of the Federal Reserve Systems, Case No. 2:24-cv-04300, ¶ 3 (S.D. Ohio) (May 23, 2025).
[6] See e.g., August 23, 2024 Federal Reserve Letter to the Goldman Sachs Group at 4 n.17 (the Board “does not implement firm-specific overlays in the supervisory stress test” and has a “policy of not using additional input data” absent comparable data from “all firms . . . in a given area”).
[7] 12 C.F.R. Part 252, Appendix B, § 2.8 (emphasis added).
[8] See Defendant’s Cross-Motion for Summary Judgement, Bank Policy Institute et al. v. Board of Governors of the Federal Reserve Systems, Case No. 2:24-cv-04300 (S.D. Ohio) (Apr. 29, 2025) (“[T]he Board occasionally exercises case-specific discretion in determining whether to apply overlays to firm-specific data prior to application of the stress test models to produce the firms’ final results.”).
[9] In particular, the Federal Reserve should, among other measures: (i) floor PPNR at $0, including through the trough quarter relevant to the SCB; (ii) reflect the benefit of higher client activity in trading revenue during periods of market volatility; (iii) reduce the sensitivity of noninterest expense projections to total assets (to avoid the effect of higher Federal Reserve bank balances and other high-quality assets on expenses); (iv) address any asymmetry in the treatment of structured notes and their hedges; (v) reflect interest rate hedges when projecting PPNR; and (vi) normalize for impact of non-recurring expenses.
