Banks ace stress test, but no capital changes at stake

Michelle Bowman
Federal Reserve Vice Chair for Supervision Michelle Bowman.
Bloomberg News

Processing Content
  • Key insight: Banks performed well in this year's stress test despite a more severe worst case scenario, but the results will have no impact on capital requirements.
  • Expert quote: "Today's results underscore the strength of the banking system. As we work to increase the transparency and accountability of the stress test, public feedback will help us continue to improve and instill greater confidence in the stress test and its results." — Federal Reserve Vice Chair for Supervision Michelle Bowman.
  • Forward Look: The Fed is weighing several changes to the stress testing process that could go into effect as soon as next year.

Megabanks and large regionals passed another round of Federal Reserve stress testing with flying colors — but, unlike recent years, their strong performance will not earn them a capital reprieve.

Collectively, the 32 banks that were examined in this year's stress test saw their aggregate common equity Tier 1 capital ratios fall by just 1.6% under the exam's severely adverse scenario, the lowest level on record since the process was reformed in 2020. No single bank came with 2 percentage points of the mandatory minimum ratio of 4.5%.

Unlike previous years, in which a bank's performance one year would have implications for the following year's stress capital buffer, capital requirements will remain unchanged this year as the Fed weighs several changes to its stress testing process. 

"Today's results underscore the strength of the banking system," Fed Vice Chair for Supervision Michelle Bowman said in a statement issued alongside the exam findings. "As we work to increase the transparency and accountability of the stress test, public feedback will help us continue to improve and instill greater confidence in the stress test and its results." 

Unlike last year, when only the 22 largest banks — those deemed Category I, II or III — were examined, this year's stress test included 10 Category IV banks that are tested every other year. The last time this cohort of banks was assessed in 2024, they registered an aggregate maximum capital decline of 2.8%, the largest of the post-2020 era. 

This year's field included two first-timers: Synchrony Financial, which posted a maximum CET1 decline of just 0.1 percentage points, and First Citizens, which saw its capital level fall by 4.5 percentage points, the second largest decline among tested banks.

Deutsche Bank posted the largest decline in its CET1 capital level, with an 8.2 percentage point decline. Despite those hypothetical losses, the bank still maintained a minimum capital level of 12.7% during the test, well above the required 4.5% minimum threshold. 

Six of the nation's largest banks performed slightly worse this year than they did in 2025, though some of those outcomes are more indicative of how they performed last year. 

Three of the premier custodial banks in the country — Bank of New York-Mellon, Northern Trust and State Street — saw modest increases in capital declines in 2026 after seeing their capital positions increase slightly under the 2025 test. Goldman Sachs saw its CET1 capital level decline 0.2 percentage points more this year, while JPMorganChase increased half a percentage point and Wells Fargo inched up 0.4 of a percentage point.

This year's adverse scenario was in some ways more severe than last year's, owing to the countercyclical design of the test — meaning it intensifies after years of strong economic growth and gets easier when the economy is already under stress. Commercial real estate prices fell by 39% instead of 30%, credit spreads increased by 4.7 percentage points instead of 3.9, and equity prices fell 58% instead of 50%.

As a result, banks saw higher levels of loan defaults in this year's test, accounting for $625 billion of projected losses, which contributed to an additional 0.2 percentage points of aggregate capital decline. Credit cards were the biggest source of losses, accounting for $203 billion, or 29% of the projected total, followed by commercial and industrial loans at $158 billion, or 22% of the total.

Another loss driver in this year's exam was changes in available for sale securities, which increased less because the scenario included a smaller decline in interest rates. This also increased estimated capital declines by 0.2 percentage points. 

Higher interest rates in this year's model led to a projected increase in net interest income for banks, a factor that more than offset the impacts of loan losses and unrealized gains, reducing capital declines by a half percentage point. 

In April 2025, the Fed proposed a change to the stress test that would average the results over two years to determine individual bank stress capital buffers, with the intent of reducing year-to-year volatility. The proposal would also adjust the timing of when new stress capital standards would go into effect.

The Fed followed that proposal up with another set of reforms in October, which would open stress testing models, framework and scenario design to public comment. The central bank has collected comments on both proposals, but has not yet finalized either set of changes. 

In February, the board of governors voted to keep the stress capital buffer unchanged this year as it weighs these potential reforms. 

"Waiting to calculate new stress capital buffer requirements until we receive public feedback will give us the opportunity to correct any deficiencies in our supervisory models based on that feedback," Bowman said in a statement at the time. "This should further improve the transparency, effectiveness, and fairness of our models and improve our accountability to the public."


For reprint and licensing requests for this article, click here.
Federal Reserve Compliance Risk
MORE FROM AMERICAN BANKER
Load More