Coverage: weekly note · Information to 29 June 2026
TGB Premium Intelligence · Subscriber Edition
This report is part of TGB Premium Intelligence and is available by request to qualified institutions. It will not enter the open archive for at least twelve months.
TGB Weekly Research Note · 29 June 2026
The Federal Reserve Board's 2026 stress test produced a 1.6 percentage point aggregate CET1 decline, smaller than in recent years, but the results will not feed into large-bank capital requirements, which stay in place until 2027.
The desk view
All 32 banks in the Federal Reserve Board's 2026 stress test stayed above their minimum CET1 requirements, and the aggregate ratio fell from 12.8% to a projected low of 11.2%. The Board said the 1.6 percentage point decline was smaller than in recent years. The more consequential feature of this cycle is institutional rather than numerical: in February 2026 the Board voted to keep current stress capital buffer (SCB) requirements until 2027, and on 24 June 2026 it confirmed that this year's results will not affect capital requirements. The June 2026 requirements document repeats the SCBs disclosed in 2025. The link that normally carries stress-test results into binding buffers has therefore been suspended while the Board reviews public feedback on its models. Our reading is that the 2026 test is best treated as information on resilience rather than as a capital event.
A smaller drawdown under a differently shaped scenario
The Board released the results on 24 June 2026, as it had announced on 9 June. It said large banks are well positioned to weather a severe recession, and all 32 banks remained above their minimum CET1 requirements in a hypothetical recession it described as similar in severity to the prior test. The aggregate CET1 ratio falls from an actual 12.8% in the fourth quarter of 2025 to a projected minimum of 11.2%, before recovering to 12.7% by the first quarter of 2028. That 1.6 percentage point decline compares with 1.8 percentage points in the 2025 test, a difference of 0.2 percentage points.
The scenario was not uniformly milder. Relative to 2025, the peak-to-trough fall in real GDP was 4.6% against 7.8% and the fall in house prices 30% against 33%, but commercial real estate (CRE) prices fell 39% against 30%. The 2026 path also took the unemployment rate to a peak of 10%. Balance sheets moved against the banks as well: loan balances grew by around 10% in 2025, concentrated in credit cards and wholesale loans, and the Board said this growth, together with more severe corporate bond spread and CRE price paths, raised stressed loan losses year over year.
The comparison with 2025 needs care. The Board said the 2025 decline was itself smaller than in recent years and in part reflected unintended volatility in the models used for the stress tests. For 2026 the supervisory models remained largely unchanged from the prior test because they are out for public comment. This suggests that two consecutive smaller drawdowns, both produced by models the Board has put out for comment, should not be read as a clean signal of improved resilience.
| Measure | 2025 test | 2026 test |
|---|---|---|
| Real GDP, peak-to-trough fall | 7.8% | 4.6% |
| House prices, fall | 33% | 30% |
| Commercial real estate prices, fall | 30% | 39% |
| Aggregate CET1 decline | 1.8pp | 1.6pp |
Source: Federal Reserve Board, 2026 Federal Reserve Stress Test Results, June 2026, 24 June 2026; Federal Reserve Board, 2025 stress test results press release, 27 June 2025.