Executive Summary 

• The Federal Reserve has finalized reforms to its bank stress-testing framework that substantially follow the proposals introduced in 2025, including requirements for public input on annual stress-test scenarios and material model changes as well as  changes intended to make resulting capital requirements more predictable. 

• The final framework pairs greater transparency in scenario and model design with a two-year averaging mechanism for the stress capital buffer beginning in 2028, a combination the Federal Reserve estimates will reduce year-over-year volatility in capital requirements by approximately 50 percent without materially affecting aggregate capital requirements.  

• The reforms mark a significant evolution in stress testing from a largely supervisory exercise toward a more transparent regulatory process, while leaving unresolved the broader question of how much supervisory discretion should remain in a framework that directly affects banks’ binding capital requirements. 

Introduction 

The Federal Reserve has finalized changes to its annual bank stress-testing framework that will, for the first time, establish a recurring process for public input on the scenarios and material model changes used in the supervisory stress test. The final rules, approved on September 30, 2026, are largely similar to proposals the Federal Reserve issued in 2025, meaning that the significance of this week’s action lies less in the introduction of an entirely new approach than marking the transition to a new supervisory philosophy.  

Stress testing has become one of the defining features of the post-2008 regulatory landscape. What began as an emergency confidence-building measure during the financial crisis has matured into a recurring supervisory exercise that influences capital planning, investor expectations, and broader perceptions of systemic resilience. The finalization of the Federal Reserve’s reforms now adds another dimension to that evolution: The assumptions underlying the test will be subject to greater public scrutiny, while the effect of any single year’s stress-test results on capital requirements will be moderated through a two-year averaging mechanism beginning in 2028. 

Taken together, these changes address two longstanding characteristics of stress testing that have often appeared in tension. The first is the opacity of the supervisory process, particularly the models and assumptions used to translate a hypothetical economic scenario into projected bank losses. The second is the volatility that can result when changes in those assumptions produce substantial changes in a bank’s required capital buffer from one year to the next. The final framework attempts to address both problems simultaneously, increasing transparency while reducing the extent to which a single annual exercise determines a bank’s capital requirement. 

From Crisis Response to Institutional Norm 

The origins of contemporary U.S. stress testing lie in the 2009 Supervisory Capital Assessment Program, which was designed during the financial crisis as a standardized means of assessing whether large financial institutions had sufficient capital to withstand severe economic conditions. Its importance extended beyond the mechanics of the exercise: By publishing information about the resilience of individual institutions, the program provided markets with information at a moment when uncertainty about bank solvency was itself a source of financial instability. 

Dodd-Frank subsequently established annual stress-testing requirements, and the Federal Reserve’s Comprehensive Capital Analysis and Review developed into a recurring component of supervision for the largest banking organizations. Over time, stress testing evolved from an extraordinary crisis-response mechanism into a permanent feature of prudential regulation. Although subsequent tailoring reforms narrowed the population subject to the most intensive requirements, the tests remain central to the supervision of the largest banking organizations. 

That evolution matters because stress testing now does more than assess resilience. The Federal Reserve uses the projected decline in a bank’s common equity tier 1 (CET1) capital ratio under stress to determine its stress capital buffer, which is added to other applicable capital requirements. Consequently, the assumptions embedded in the supervisory exercise can affect the amount of capital a bank must hold and therefore influence decisions concerning lending, capital distributions, acquisitions, and balance-sheet management. under stress to determine its stress capital buffer, which is added to other applicable capital requirements. Consequently, the assumptions embedded in the supervisory exercise can affect the amount of capital a bank must hold and therefore influence decisions concerning lending, capital distributions, acquisitions, and balance-sheet management. 

Why Stress Tests Matter to Regulators and Banks 

Stress testing provides supervisors with a forward-looking assessment of capital resilience that static regulatory ratios cannot. Rather than asking whether an institution satisfies a capital requirement under current conditions, the exercise asks how its capital position would evolve under a severe hypothetical scenario. 

That distinction is important because the stress-test scenario necessarily embodies judgments about which risks deserve attention and how those risks might interact. A recession accompanied by declining asset prices, higher unemployment, deteriorating commercial real estate conditions, or financial-market disruption can produce very different projected losses depending on the assumptions incorporated into the scenario and the models used to translate those assumptions into firm-specific results. 

The stress capital buffer then gives these projections regulatory significance. Under the previous framework, the maximum projected decline in CET1 during the stress period was a principal determinant of the institution’s stress capital buffer. The result was a direct relationship between the outcome of the supervisory exercise and the capital requirement applicable to the institution. 

The Federal Reserve’s 2025 proposal to average stress-test results over two years was designed to reduce the volatility associated with that relationship. The final rule adopts that approach, requiring the Federal Reserve beginning in 2028 to use the results of the two most recent annual supervisory stress tests when calculating the stress capital buffer for firms subject to the test in both years. The Federal Reserve estimates that this change, together with the other reforms, will reduce year-over-year volatility in capital requirements by approximately 50 percent.  

The Significance of Public Comment 

The most visible change in the final framework is the formalization of public input into stress-test design. The Federal Reserve will annually invite public comment on the stress-test scenarios and on material changes to the models used to conduct the exercise, while also publishing additional documentation concerning the framework. The agency has adopted the models that will be used for the 2027 stress test under this revised framework.  

This change responds to a fundamental tension in stress testing. Supervisors require sufficient flexibility to identify risks that may not be apparent from historical data, while banks and other interested parties have an interest in understanding the assumptions that determine their regulatory obligations. Historically, the Federal Reserve maintained considerable discretion over the construction of scenarios and models, in part because excessive predictability could encourage institutions to manage specifically to the test rather than to the broader range of risks that the test is intended to represent. 

Opening the process to public comment does not eliminate that tension, but it changes the institutional balance. Academic researchers, market participants, banks, and other stakeholders will have an established opportunity to identify potential deficiencies in scenarios or models before those assumptions are incorporated into the supervisory exercise. The process may therefore improve the information available to supervisors while also making the regulatory basis for the resulting capital requirements more transparent. 

At the same time, transparency creates a potential tradeoff. The more information institutions have about the construction of a stress test, the greater their ability to anticipate the conditions against which they will be evaluated. The question is therefore not whether transparency is beneficial in isolation, but whether the information disclosed is sufficient to improve accountability without making the exercise so predictable that it loses some of its diagnostic value. 

The final framework is notable because the Federal Reserve has now accepted this tradeoff as an enduring feature of the stress-testing process rather than a temporary experiment. Public comment is no longer merely a proposed procedural reform; it is part of the framework through which future scenarios and material model changes will be developed. 

Reducing the Consequences of Annual Volatility 

The final rules also make clear that greater transparency is being introduced alongside an effort to make the capital consequences of stress testing more stable. This is an important development because the two reforms address different stages of the same process. 

Public comment concerns the construction of the test: What risks should be included, how severe should the scenario be, and how should projected losses and revenues be modeled? Two-year averaging concerns what happens after the test has been conducted: How much weight should a single year’s result have in determining a bank’s capital requirement? 

The answer under the final framework is that a single year will no longer determine the stress capital buffer beginning in 2028 for firms tested in consecutive years. Instead, the Federal Reserve will average the relevant results from the two most recent annual exercises. The agency expects this to reduce volatility without materially changing aggregate capital requirements.  

This distinction is consequential. A stress test remains a hypothetical exercise, and its results necessarily depend on assumptions that can change from year to year. If those changes translate directly into large movements in binding capital requirements, institutions may face difficulty distinguishing changes in their underlying risk from changes caused by the supervisory scenario itself. Averaging provides a degree of insulation from that problem without severing the connection between stress-test performance and capital requirements. 

The tradeoff is that a two-year average will necessarily make the capital requirement somewhat less responsive to any single year’s results. A sharp deterioration or improvement in a bank’s projected resilience will therefore be incorporated into the capital requirement more gradually than under a purely annual approach. The broader supervisory framework consequently remains important because supervisors retain other tools for responding to emerging risks that may not yet be fully reflected in an averaged stress capital buffer. 

The Global Market Shock and the 2027 Transition 

The final rule also modifies the treatment of the global market shock for firms with significant trading activities. Such firms will now be tested against two global market shock components each year, with the shock producing the largest losses for an individual firm used in calculating that firm’s stress-test results. 

This change illustrates that the final framework is not simply a procedural reform. The Federal Reserve is also modifying substantive elements of the stress-testing methodology in an effort to make the exercise more risk-sensitive. For institutions with large trading books, where exposures can vary substantially across asset classes and market conditions, the use of multiple shocks is intended to provide a broader assessment of potential market losses while retaining a firm-specific element in determining which shock is binding. 

The timing of the reforms is also significant. The Federal Reserve will conduct the 2027 stress test using models adopted as part of the new framework, while the two-year averaging mechanism for stress capital buffers will begin in 2028 so that the averaging process relies only on stress-test results generated under models that have been subject to the new transparency and public-input process.  

What Stress Testing Is—and What It Should Be 

The final framework makes the broader policy question surrounding stress testing more concrete. At its core, stress testing is not simply a technical exercise for estimating hypothetical losses. It is a mechanism through which supervisors communicate which risks they consider important, impose forward-looking discipline on banks, and translate judgments about resilience into regulatory capital requirements. 

The decision to solicit public comment changes the institutional character of that mechanism. The Federal Reserve will continue to determine the scenarios and models, but the process will now incorporate an established opportunity for outside participants to challenge assumptions and identify risks. This may improve the quality of the exercise, but it also places greater emphasis on how the Federal Reserve evaluates and responds to those comments. 

The simultaneous adoption of two-year averaging addresses a different concern. If the purpose of stress testing is to measure resilience rather than produce highly volatile annual capital requirements, then reducing the influence of any single exercise may improve the relationship between the test and the regulatory objective. The final framework thus moves stress testing toward a model in which transparency, supervisory judgment, and capital stability are intended to operate together rather than treating them as competing objectives. 

Conclusion 

The Federal Reserve’s final reforms largely turn its 2025 proposals into policy, pairing greater transparency around stress-test scenarios and models with measures intended to make capital requirements less volatile from one year to the next. That is a meaningful improvement in predictability, particularly for institutions making capital-allocation decisions under a framework in which supervisory assumptions can have direct consequences for lending and distributions. Yet it does not resolve the more fundamental problem of relying on a highly discretionary hypothetical exercise to determine how much capital banks must hold. The reforms may make that process more transparent and somewhat less volatile, but the coming stress cycles will ultimately determine whether greater procedural openness translates into a more disciplined regulatory framework or simply a more visible – and gameable – exercise of supervisory discretion.