This post has been updated to clarify several scenario explanations and to correct the jump-off point to peak/trough calculations for the CRE, VIX and Dow Jones Index scenarios in Table 1.
On Oct. 24, 2025, the Federal Reserve proposed its scenarios for the 2026 supervisory stress test. For the first time, the Fed is seeking public comment on the scenarios by Dec. 1 before finalizing them by Feb. 15 (link). The Fed also released a new Scenario Design Policy Statement which describes, in more detail, how the Fed calibrates the paths for scenario variables through “guides” (link). To generate scenario paths for variables not calibrated using the guides, the Fed published a small-scale macro model (link).
The prominent themes of the proposed 2026 severely adverse scenario are a sharp decline in commercial real estate (CRE) prices and investor aversion to long-term assets. Concern about credit risk is reflected by a sharp increase in the BBB spread and a higher-for-longer path for equity market volatility measured by VIX. Scenario variables based on the proposed guides are generally toward the more severe half of possible values stipulated in those documents, reflecting current concerns about the health of the economy.
Each bank’s performance under the severely adverse scenario will become part of their annual stress capital buffer, which is one component of their minimum capital requirement.
The Proposed Scenario Design Policy Statement
Since 2019, the Fed has provided guides describing how it calibrates its severely adverse scenario paths for the unemployment rate and house price growth. The proposed 2025 Scenario Design Policy Statement adds guides for more variables, including: commercial real estate prices, equity prices, the VIX, five- and 10-year Treasury yields, the BBB corporate spread, mortgage spreads and certain international series.[1] Meanwhile, scenarios for GDP, disposable personal income per capita, the three-month Treasury bill yield and inflation are generated by the Fed’s “macro model for stress testing.”
Table 1 compares the proposed 2026 stress test scenarios for certain variables against the possible ranges outlined in the new guides. The five- and 10-year Treasury yield scenarios are at the midpoint of their ranges while the scenarios for the BBB corporate spread, CRE prices, stock market prices, VIX and mortgage rate spread are toward the more severe half of their possible range. These scenarios may change between now and February 2026 as the Fed incorporates public feedback and updates the scenario’s starting values as more up-to-date data becomes available.[2]
Table 1: Comparing 2026 Scenario to Proposed Fed Guides
(a). Scenario values can fall outside the proposed guide ranges under certain conditions. For example, the 5-year Treasury yield guide says yields will fall between 1.5 and 3.5 percent, subject to a lower bound of 0.3 percent or a decline of 0.3 percentage points from the jump-off level, whichever is lower.
Comparing the Proposed 2026 Severely Adverse Scenario to Previous Years
The Fed’s guide for the unemployment rate provides for an increase between 3 and 5 percentage points or to a minimum level of 10 percent. In the proposed 2026 scenario, the unemployment rate rises 5.5 percentage points to 10 percent by the third quarter of 2027. While 10 percent is often the peak value for the unemployment rate, the starting value for unemployment is higher in the 2026 scenario than in 2021 to 2025 scenarios, so the increase is less (Exhibit 1).
The Scenario Design Policy Statement requires the ratio of the house price index to disposable personal income per capita to decline 25 percent or enough to bring the ratio down to its Great Recession trough, whichever is greater. House prices are already relatively low relative to disposable income, so the 2026 scenario has them decreasing less than in the past (Exhibit 2). As a result, credit losses for residential mortgages should be less pronounced than in some prior stress tests.
Commercial property prices decline 40 percent in the proposed 2026 severely adverse scenario — comparable to the 2021 to 2024 stress tests (Exhibit 3). The 2021 scenario described significant risks to property types affected by adjustments related to the COVID-19 event while the 2022 scenario reflected concerns over the prolonged continuation of remote work (link and link). The severe decline in CRE prices in the 2026 scenario could reflect continuing concern over high vacancy rates, slow rent growth or heightened refinancing risk for CRE borrowers. The proposed path for CRE prices may result in higher projected credit losses as compared to last year’s exercise.
The severe CRE price decline in the 2022 scenario was accompanied by similarly large increases in the VIX and the BBB spread. In that year’s scenario, the Fed reasoned that large declines in residential and CRE prices had potential to spill over into asset markets and investor sentiment as captured by sharper increases in equity market volatility and corporate bond spreads (link).
The proposed 2026 scenario includes a unique path for the VIX — peaking around the same level as past years but remaining higher for longer (Exhibit 4). Equity market volatility matters most for banks’ market risk and operational risk losses, although pre-provision net revenue calculations are also affected in various ways.
Like 2022, the proposed path for the BBB corporate spread increases 4.4 percentage points from its starting value — the second most severe observed increase (Exhibit 5). A sharp increase in the BBB corporate spread will result in higher credit losses, as it is used a summary metric for default risk in the Fed’s corporate model (link). The BBB spread also appears in the Fed’s models for pre-provision net revenue, securities and operational risk.
One distinct feature of this year’s scenario is the steep yield curve — measured by the difference between the 10-year and three-month Treasury yields — which peaks higher than in any prior stress test and remains elevated throughout the scenario (Exhibit 6). The peak is similar to the 2018 scenario where the Fed described investor aversion to long-term fixed-income assets as keeping 10-year Treasury yields elevated (link). Higher long-term yields in the 2018 scenario reduced unrealized gains on U.S. Treasury and Agency securities but resulted in higher net interest income.
The Federal Reserve’s proposed 2026 severely adverse scenario is broadly consistent with prior stress test exercises. The scenario does feature a noteworthy decline in CRE prices, a significant widening of credit spreads, a higher-for-longer equity market volatility path and a steep yield curve.
[1] Guides for the unemployment rate and house prices in the new 2025 statement were unchanged from the 2019 statement.
[2] Currently estimates by the Fed are used in place of unavailable data points. The estimates will be replaced in the final 2026 scenario with published values after they become available. With new data, the jump-off values for the scenarios will change as will the scenario paths based on these new jump-off values.
