BPI Statement Before The House Financial Services Committee, Subcommittee on Financial Institutions and Monetary Policy

Chairman Barr, Ranking Member Foster and members of the Subcommittee, thank you for giving me the opportunity to testify today. My name is Francisco Covas, and I am the Head of Research at the Bank Policy Institute, which is a research and advocacy group supported by banks with more than $100 billion in U.S. assets. As our membership comprises the full range of banks covered by the Federal Reserve’s stress tests, we welcome the opportunity to testify today.

The stress-testing regime that underpins the capital framework for large banks has consistently demonstrated that U.S. banks are well capitalized and can continue to support the economy during economic downturns. Since 2020, stress-test results have been used to set a stress-capital buffer that in effect is one element of banks’ minimum capital requirements.

As codified in the Federal Reserve’s Stress Capital Buffer (SCB) rule, the supervisory stress tests, and by extension the supervisory models and scenarios they use, play a significant role in determining banks’ capital requirements. Therefore, these models and scenarios have real-world implications for the cost and availability of credit to households and businesses, and for the role banks play in intermediation of capital markets. Firms use the results of the Fed’s stress test to allocate capital across business lines and products, to better reflect the capital requirements they face.

Banks support the use of a stress capital buffer as a component of their overall capital requirements, but the Fed’s current stress-test models and scenarios are rules that should be subject to notice-and-comment process under the Administrative Procedure Act (APA). In violation of the APA, the Fed discloses limited details on the construction of the scenarios; has failed to adopt and adhere to a standard for constructing those scenarios; fails to publicly disclose many important details of its internal models; and adopts changes to the stress tests each year without granting the public an opportunity to comment, despite the significant effects on the covered banks as well as financial markets and the broader economy.

Economic Implications of Lack of Transparency in the Stress Tests

The lack of transparency in the Federal Reserve’s stress-testing regime has significant economic costs that can adversely affect the U.S. financial system and the broader economy. Stress tests are designed to assess the resilience of large banks during severe economic downturns. The process involves Fed staff projecting loan losses and revenues for each bank under a hypothetical recession scenario. However, the opacity and inaccuracy of the Fed’s models create uncertainty among large banks regarding the level of capital they are required to hold. This uncertainty leads banks to hold excess “uncertainty buffers” of capital, which increases the costs of providing financial intermediation between borrowers and savers, ultimately reducing credit availability and market intermediation.

A seminal paper by Ben Bernanke, former Fed Chairman, was one of the first to formally analyze the impact of uncertainty on business investment.[1] In his paper, Dr. Bernanke demonstrated that increased uncertainty depresses current investment, particularly for long-lived investment projects and decisions that are economically costly to reverse. The capital requirements for large banks are partially determined by supervisory stress tests. As a result, uncertainty about banks’ post-stress regulatory capital ratios, caused by the lack of accuracy and the volatility in the projections of the Fed’s models, is likely to lead to an underinvestment in banks’ financial intermediation activities.

Banks play a crucial role in the economy by transforming short-term liabilities, such as deposits, into longer-term illiquid assets, such as loans. This maturity transformation function is inherently a long-term investment that is illiquid, making it particularly vulnerable to the uncertainty induced by the U.S. stress tests. Consequently, this uncertainty is expected to depress bank lending, leading to suboptimal outcomes for the economy. By increasing the transparency of the stress tests, the uncertainty surrounding banks’ capital requirements would be reduced. This reduction in uncertainty would boost financial intermediation more broadly, improve banking efficiency and promote a better allocation of capital, as suggested by the academic literature.

Banks also serve as critical intermediaries in capital markets, providing market-making, underwriting and liquidity provision services that facilitate efficient price discovery and capital allocation. Capital markets play a critical role in economic growth, providing about 72 percent of equity and debt financing for non-financial corporate issuers.[2] However, the Global Market Shock (GMS) component of stress tests significantly increases capital requirements for trading and capital markets activities. The GMS assumes unrealistically long periods of market illiquidity during which banks cannot hedge or rebalance portfolios, and imposes these stresses instantaneously, creating implausibly severe scenarios. As a result, banks have been forced to misallocate capital away from capital market activities, potentially reducing economic growth and market efficiency.

Several academic papers have documented the impact of stress tests on the availability of credit in the United States. The academic literature has mainly focused on the severity of the scenarios included in the stress tests. However, this is highly intertwined with the lack of transparency of the models, because both components contribute to the overall stringency of the test.

Acharya, Berger, and Roman (2018) find that banks subject to stress tests have reduced the supply of credit to borrowers with less than pristine credit scores and to cyclical firms, including small businesses.[3] Calem, Correa, and Lee (2021) show that stress tests has led to the reduction in originations of jumbo mortgage loans.[4] Chen, Hanson, and Stein (2018) demonstrate that the four largest banks significantly cut back on lending to small businesses after the Global Financial Crisis (GFC) relative to the rest of the banking sector.[5] Crucially, they show that this reduction in small business lending led to a decline in the fraction of businesses that expand employment. A cut in such lending also results in slower employment growth, a higher unemployment rate and slower wage growth in counties where the largest banks had a significant presence.

Moreover, Cortés, Demyanyk, Li, Loutskina, and Strahan (2020) showed that banks most affected by the stress tests reduced their supply of small business loans by increasing loan rates and shifting their portfolios toward safer loans.[6] Brauning and Fillat (2020) found that since the implementation of the annual stress tests in 2011, the portfolios of stress-tested banks have become more similar.[7] Banks that had poor stress-test results adjusted their portfolios to resemble those of banks that performed well. Although stress testing has led to more diversified individual bank portfolios, it has also resulted in a more concentrated overall portfolio among large banks. Furthermore, banks with worse stress-test results have reduced lending in loans most sensitive to the stress-test scenarios, leading to real economic effects for borrowers. Importantly, the act of banks shifting lending to portfolios that perform well in the stress test without transparency into the approach may lead banks to make decisions based on opaque criteria rather than sound risk assessment, potentially introducing new vulnerabilities into the financial system.

Finally, uncertainty around capital requirements also poses significant challenges for banks as an investment asset class. U.S. banks have experienced a decline in their market value relative to book value over the past decade, partly due to increased regulation.[8] This trend makes banks less attractive to investors, as evidenced by their low price-to-tangible-book-value ratios compared to the broader market. This regulatory uncertainty makes it difficult for investors to accurately value banks and assess their risk-return profiles. The result is a higher cost of capital for banks, as investors demand higher returns to compensate for the increased risk and unpredictability. These higher costs are ultimately passed on to customers and businesses in the form of higher interest rates or fees. Moreover, during periods of severe financial stress, this diminished investor appetite could significantly impair banks’ ability to raise new capital when it is most needed, exacerbating economic downturns and limiting banks’ capacity to support lending and economic growth.

In conclusion, the lack of transparency in the Fed’s stress-testing regime imposes significant economic costs, including reduced credit availability (particularly for small businesses), slower employment growth, diminished market liquidity, and less efficient capital allocation across the banking sector. While reduced credit for riskier borrowers and lower market liquidity could be optimal if stress scenarios were reasonable and models accurate, we believe neither is currently the case. To address these issues, policymakers should establish a cogent standard for stress scenarios, allow public comment on them and provide full transparency on the models used. By reducing the opacity of supervisory models and subjecting scenarios to notice and comment, policymakers can decrease uncertainty and foster a more efficient financial system that better serves the needs of the U.S. economy.

On the Importance of Notice and Comment on the Models Used to Calculate the Stress Capital Buffer

The current SCB framework faces several critical issues that hinder the ability of banks and other stakeholders to effectively assess and manage capital requirements:

  • One major issue is the excessive volatility of the SCB, which can fluctuate significantly from year to year, often without reflecting actual changes in a bank’s risk profile. This volatility makes it difficult for banks to plan and manage their capital efficiently.
  • The lack of transparency in the SCB framework makes it challenging for banks, investors and other stakeholders to understand how the capital buffer requirement is determined and to propose improvements.
  • The stress test also produces inaccurate and counterintuitive results inconsistent with more granular bank models and recent market experience, such as trading and counterparty losses during the COVID-19 pandemic.
  • The current reconsideration process for the SCB does not give banks a meaningful opportunity to appeal the stress-test results.

The remainder of this section presents concrete examples of the inaccuracy and volatility of supervisory stress-testing models used to calculate the stress capital buffer. If the Fed made the models transparent, there would likely be substantial review by experts, academics, and banks could then make suggestions to the Fed to update the models to be more accurate and avoid some of the problems that will be discussed next.

SCB Changes Significantly Year Over Year and Often Does Not Align with Changes in a Bank’s Risk Profile

The first significant issue with the current stress capital buffer framework is its excessive volatility. Although some volatility in the SCB is necessary and expected as the scenarios change each year to reflect evolving economic conditions, emerging risks and changes to bank portfolios, the current level of volatility appears excessive and disconnected from actual changes in banks’ risk profiles. This unpredictability makes it challenging for banks to effectively plan and manage their capital requirements, since there is only one quarter between receiving the stress-test results and the new requirement becoming effective. By contrast, banks have two years to comply with changes to the global systemically important banks (GSIB) capital surcharge.

To read the full statement, please click here, or click on the download button below.

[1] See Bernanke, Ben S., “Irreversibility, Uncertainty, and Cyclical Investment,” Quarterly Journal of Economics, February 1983, Vol. 98, No.1, pp. 85–106.

[2] SIFMA, US Capital Markets Are the Largest in the World: 2023 Capital Markets Fact Book (2023), available at https://www.sifma.org/wp-content/uploads/2022/07/2023-SIFMA-Capital-Markets-Factbook.pdf

[3] Acharya, Viral, Allen Berger, and Raluca Roman, “Lending Implications of U.S. Bank Stress Tests: Costs or Benefits?,” Journal of Financial Intermediation, April 2018, Vol. 34, pp. 58–90.

[4] Calem, Paul, Ricardo Correia, and Seung Jung Lee, “Prudential Policies and their Impact on Credit in the United States,” Journal of Financial Intermediation, April 2020, Volume 42, pp.  1008–26.

[5] Chen, Brian, Samuel Hanson, and Jeremy Stein, “The Decline of Big-Bank Lending to Small Business: Dynamic Impacts on Local Credit and Labor Markets,” NBER Working Paper No. 23843.

[6] Cortés, Kristle, Yuliya Demyanyk, Li Lei, Elena Loutskina, and Philip Strahan, “Stress Tests and Small Business Lending,” Journal of Financial Economics, April 2020, Vol. 136, Issue 1, pp. 260–279.

[7] Brauning, Falk and Jose L. Fillat, “The Impact of Regulatory Stress Tests on Bank Lending and its Macroeconomic Consequences,” FRB of Boston Working Paper No. 20–12.

[8] See Sarin, Natasha and Lawrence H. Summers, “Understanding Bank Risk Through Market Measures,” Brookings Papers on Economic Activity, Fall 2016. Available at https://www.brookings.edu/wp-content/uploads/2017/02/sarintextfall16bpea.pdf