Board of Governors of the Federal Reserve System

Since the height of the financial crisis in 2009, stress testing has increased the resilience of our large banking institutions by ensuring that capital requirements are appropriately risk sensitive, based on individual firm-specific assessments, and credible to the public. Annual stress testing, which began in 2011, has allowed supervisors to communicate with banks about potential vulnerabilities well before they escalate into material risks to the financial system. Stress testing of large banking institutions undoubtedly has been a vital component in fulfilling our supervisory and financial stability mandates.

I view the Board's actions regarding the 2026 stress tests—the extension of the prior year's stress capital buffer requirements and approval of the 2026 scenarios—as being necessary to ensure that the Board's stress testing program remains a credible assessment of the resilience of the banking sector into the future.

Today's actions are part of a broader package of changes to the stress testing program that the Board is considering. I believe some of those elements may improve the stress testing process, but others may create challenges. I want to reserve judgment until we have received all of the public comments and I have had time to consider the final proposals.

I understand that putting scenarios out for public comment improves the transparency of the annual stress testing process and that public comments play an important role in democratic governance. However, when the effective date of the hypothetical stress scenarios occurs after the scenarios are released for comment, as they were this year, some banks may have real economic incentives to temporarily adjust their balance sheets to the details of the scenario, for example, by reshuffling their securities portfolios to artificially reduce stressed losses.

Put another way, the comment process may give banks not only an early preview of the test, but a window to change their answers to score better. A range of options exist for avoiding this situation to preserve the integrity and informational value of the annual stress test exercise. The decision to extend the stress capital buffer requirements from last year mitigates this concern for this year's testing cycle, and a range of options exist for remedying this concern for future cycles.

Further, as I noted last year, I hope that we will look for ways to integrate exploratory stress scenarios into our supervisory program. Rather than being used directly to determine regulatory capital requirements, exploratory scenarios can provide invaluable information for banks and supervisors by focusing on targeted sources of stress drawn from a wider range of economic conditions than those contemplated in the annual stress tests used for the stress capital buffer.