Bank stress tests after the 2008 crisis: how supervisors test for a crash

What is a bank stress test after the 2008 crisis?

A bank stress test models a hypothetical shock, such as rising unemployment, falling equity markets, declining GDP, rising interest rates, or rising oil prices, and measures the loss a bank would take (Wikipedia, Stress test (financial)). The purpose is to assess whether large banks can absorb losses while continuing to lend. Tests can also combine events or replay historical scenarios.

Regulators became interested in their own stress tests beginning in 2007, and the tests have been routinely performed since. The 2008 crash explains why: some balance sheets that looked sound in calm years proved fragile. A test asks the question before the crash and records the answer for the public.

What did the first US bank stress test of May 2009 find?

The Federal Reserve, the Office of the Comptroller of the Currency, and the FDIC announced the Supervisory Capital Assessment Program results on 7 May 2009 at 5 p.m. Eastern time. The program covered the 19 largest US bank holding companies, which one source describes as the core of the US banking system, and it measured capital on a Tier 1 common basis (Federal Reserve press release; Wikipedia, Supervisory Capital Assessment Program).

Ten banks needed additional capital: Bank of America $33.9 billion, Wells Fargo $13.7 billion, GMAC $11.5 billion, Citigroup $5.5 billion, and 6 others with smaller amounts. The 10 amounts listed sum to $74.6 billion, and the other 9 banks were rated adequate. The result sent a clear message to the market about which banks held enough capital.

Which scenario did the 2009 bank stress test assume?

The more adverse scenario assumed real GDP growth of -3.3% in 2009 and 0.5% in 2010, an unemployment rate averaging 8.9% in 2009 and 10.3% in 2010, and house prices down 22% in 2009 and 7% in 2010 on the Case-Shiller 10-City Composite (Wikipedia, Supervisory Capital Assessment Program).

History allows a comparison. US real GDP fell 4.3% from peak to trough, unemployment peaked at 10.0% in October 2009, and home prices fell by over a fifth from early 2007 to mid-2011 (Wikipedia, Great Recession; Federal Reserve History). The scenario set a 2010 unemployment average above the later peak, which shows a scenario can be severe enough to cover the path that followed.

How did CCAR and Dodd-Frank make stress tests a yearly routine?

The Federal Reserve introduced the Comprehensive Capital Analysis and Review in 2009 as an extension of the tests used during the 2008 crisis (Wikipedia, Comprehensive Capital Analysis and Review). It runs annually alongside the Dodd-Frank Act supervisory stress test. The Federal Reserve publishes 3 scenarios each year, baseline, adverse, and severely adverse, and projects 9 quarters ahead.

The Federal Reserve may order a bank to suspend planned dividends and buybacks until its target capital balance is restored. In October 2012 regulators required the largest banks to test twice per year, once internally and once by the regulators, and from 2014 midsized firms with $10 billion to $50 billion in assets joined (Wikipedia, Stress test (financial)).

How did the European bank stress tests of 2009 to 2014 compare?

The EU-wide tests of 2009 and 2010 assumed a benign macroeconomic environment, with -0.6% euro-area growth even in their adverse scenarios. By 2011 the adverse scenario had been adjusted to a growth path of -4.0% (Wikipedia, Stress test (financial)).

In 2014, 25 banks failed a stress test conducted by the European Banking Authority. The contrast with the earlier tests teaches one lesson: a mild scenario produces mild results, and a test gains credibility when its scenario matches the risks that exist. Readers of any stress-test result can start by reading the scenario before the headline.

What makes a bank stress test credible after 2008?

A financial stress test is only as good as the scenarios on which it is based (Wikipedia, Stress test (financial)). 3 features help: scenarios that combine several shocks, scenarios that replay historical episodes, and published results at bank level. The Federal Reserve conducts its supervisory test annually with at least 2 scenarios and publishes bank-level results.

Post-stress capital levels across large banking institutions have generally increased since the Federal Reserve began conducting Dodd-Frank stress tests, and the results help set each firm’s stress capital buffer. Tests therefore feed back into capital, because the result sets part of the buffer a bank must carry.

How can readers use bank stress tests to read crash risk today?

A stress test result is a statement about one bank and one scenario. For crash risk, read 3 things: the scenario, the size of the capital buffer after stress, and the date of the test. As of October 2026, comparing the scenario with the shocks you worry about shows whether a published result covers your case. This page is general information, not investment advice.

The EQOS Asia method page explains how a structural score covers the whole system instead of 1 bank, and the Λ simulator estimates a first score in about 10 minutes. A 30-minute pre-diagnostic then applies the approach to your exposure in Singapore or across Asia.

Frequently asked questions

About the author

Stéphane Jambu

President, EQOS Forecast Asia

Stéphane Jambu leads EQOS Forecast Asia and works as a consultant in e-commerce, search visibility, and AI. LinkedIn profile