CeFPro Connect

News
Fed Stress Test Overhaul Exposes SVB Supervision Failures
Federal Reserve supervision chief Michelle Bowman has linked planned stress-testing reforms to lessons from Silicon Valley Bank, arguing that stronger scenario analysis could have identified the failed lender’s vulnerabilities years before its 2023 collapse.
Sep 22, 2026
Tags: Stress Testing Industry News
Fed Stress Test Overhaul Exposes SVB Supervision Failures
The views and opinions expressed in this content are those of the thought leader as an individual and are not attributed to CeFPro or any other organization



  • Fed stress-testing reforms are expected to be considered in the coming weeks
  • Proposals include two global market shock scenarios and temporarily freezing participating banks’ balance sheets
  • Bowman wants stress tests to identify emerging vulnerabilities rather than simply determine capital requirements
  • Scenario analysis could have exposed SVB’s capital weakness as early as 2021
  • An independent review identified supervisory delays, risk aversion and unclear accountability at the Fed 

The Federal Reserve is preparing further changes to its bank stress-testing regime as Vice Chair for Supervision Michelle Bowman argues that stronger scenario analysis could help supervisors identify vulnerabilities before they develop into crises.

Bowman said the Fed expects to consider final revisions to its stress-testing framework in the coming weeks, alongside changes to risk-based capital requirements and the surcharge imposed on global systemically important banks.

Speaking in London, Bowman highlighted proposals that could further reshape how the annual tests operate.

One would establish a date on which participating banks’ balance sheets are temporarily frozen before the Fed publishes its annual stress scenarios, preventing institutions from adjusting their positions in anticipation of the exercise.

Another proposal would introduce two global market shock scenarios, with the greater resulting loss used to calculate an institution’s stress capital buffer. Bowman said both suggestions would be considered as the Fed prepares its final rule.

The central bank also plans to seek public feedback on a revised model for assessing noninterest income in the 2027 stress test.

This would cover fee and trading revenue generated through businesses including wealth management, investment banking and market making.

“In finalizing these reforms, we will finally close the book on an opaque and unnecessarily unpredictable framework,” Bowman said.

However, Bowman argued that stress testing should extend beyond calculating regulatory capital requirements.

Greater communication between banks and supervisors could turn scenario analysis into a more effective tool for detecting emerging financial and nonfinancial vulnerabilities.

“Scenario analysis is powerful on its own,” she said. “It becomes even more powerful when it is not a one-way exercise.”

Bowman connected that argument directly to the 2023 collapse of Silicon Valley Bank, suggesting broader scenario analysis might have exposed its vulnerabilities significantly earlier.

“Had the Fed applied this type of analysis, supervisors could have measured the effect of a range of stressful scenarios on SVB’s portfolios, exposures, and capital and liquidity positions,” Bowman said.

A plausible recession scenario combining rising interest rates with investment securities measured at fair value could have shown SVB’s capital falling below minimum requirements as early as the fourth quarter of 2021, she added.

Bowman also detailed findings from an independent review of the Fed’s supervision of SVB.

She said supervisory staff knew, or should have known, about vulnerabilities at the bank by March 2022 but failed to take sufficiently prompt and decisive action to reduce its interest-rate risk and concentration of vulnerabilities.

The review also identified cultural and organizational problems within the Fed.

According to Bowman, supervisory staff operated within a long-standing culture of risk aversion in which taking no action could be perceived as safer than taking the wrong action.

Uncertainty also existed over who possessed final authority to determine appropriate corrective measures.

Bowman stressed that SVB’s failure could not be attributed to a single institution or individual, but said the episode damaged confidence in banking supervision.

“It fundamentally shook public confidence in the effectiveness of bank supervision,” she said. “That loss of confidence demands our response.”

The lessons from SVB are now feeding into a broader argument for stress testing as an active supervisory tool - one capable of identifying weaknesses before supervisors are forced to respond to them during a crisis.

Sign in to view comments
You may also like...
ad
Related insights