Khan Capitals | June 2026
Key Takeaways
- All 32 banks passed the 2026 Fed stress test, released on 24 June, remaining above minimum common equity tier 1 requirements under a scenario that assumed unemployment peaking at 10%, a 30% fall in house prices and a 39% collapse in commercial real estate values, according to the Federal Reserve.
- Projected losses exceeded $708 billion, led by roughly $200 billion on credit cards, around $160 billion on commercial loans and about $75 billion on commercial real estate; yet aggregate CET1 ratios declined by just 1.6 percentage points.
- The pass clears the way for buybacks and dividend increases, which banks typically announce in the days after results. Capital return, not capital adequacy, is now the live question for the sector.
- The test is getting friendlier by design. The Fed confirmed the 2026 results will not affect large-bank capital requirements before 2027 while it revises the framework, part of a broader softening of the post-2008 supervisory architecture.
- The unmeasured risk sits outside the tested perimeter. A decade of migration of leveraged lending into private credit means the stress test says progressively less about where credit risk in the economy actually lives.
The 2026 Fed Stress Test: A Paper Apocalypse, Comfortably Absorbed
The 2026 Fed stress test asked America’s largest banks to survive an imagined catastrophe: unemployment rising to 10%, house prices down 30%, commercial property down 39%, and markets in disarray. On 24 June the Federal Reserve published the answer. All 32 tested institutions absorbed more than $708 billion of hypothetical losses and stayed above their minimum capital requirements, with the aggregate common equity tier 1 ratio falling by only 1.6 percentage points along the way.
On the surface this is a clean bill of health for the American banking system, and the sector will treat it as one: the pass triggers the annual ritual in which banks refresh their stress capital buffers and announce dividend increases and buyback programmes in the days that follow. But the more interesting reading of the 2026 Fed stress test is as a document about the test itself. An exercise designed after 2008 to make banks prove resilience under duress is becoming, year by year, a gate that everyone clears with room to spare. Whether that reflects genuinely fortified banks, a softening examination, or risk quietly relocating beyond the examiner’s reach is the question that matters for investors.
What the Scenario Threw at the Banks
The severely adverse scenario in the 2026 exercise was, on paper, a genuine depression. Unemployment more than doubles from current levels to a 10% peak. Residential property loses almost a third of its value. Commercial real estate, already the most scrutinised asset class in American finance, falls 39%. Equity markets and corporate credit deteriorate in sympathy. The scenario is deliberately harsher than any single historical episode; it is closer to 2008 layered on top of a commercial property collapse.
Run through the banks’ books, that world produces more than $708 billion of projected losses over the nine-quarter horizon. The composition tells its own story about where American household and corporate risk now sits. Credit cards contribute roughly $200 billion, the single largest bucket, a reminder that the marginal dollar of consumer stress lands on unsecured revolving credit. Commercial loans add around $160 billion. Commercial real estate, for all its notoriety, contributes about $75 billion, painful but manageable across 32 balance sheets that have spent three years provisioning for exactly this exposure.
| 2026 severely adverse scenario | Assumption | Projected outcome |
|---|---|---|
| Unemployment | Peak of 10% | ~$200bn credit card losses |
| House prices | Decline of 30% | Mortgage losses within buffers |
| Commercial real estate | Decline of 39% | ~$75bn CRE losses |
| Commercial lending | Severe corporate downturn | ~$160bn commercial loan losses |
| Aggregate impact | $708bn+ total losses | CET1 down just 1.6pp; all 32 banks pass |
A 1.6 Point Dent From a Depression
The headline that deserves the most attention is not the pass rate but the shallowness of the capital drawdown. An aggregate CET1 decline of 1.6 percentage points, against starting ratios comfortably in the double digits for most large banks, means the system emerged from a simulated depression with capital to spare. Fifteen years of retained earnings, issuance discipline and regulatory pressure have built a capital base that dwarfs the one that entered 2008.
Two other forces flatter the arithmetic. First, bank earnings power is itself a buffer: pre-provision revenue over nine quarters absorbs a large share of stress losses before capital is touched at all, and today’s wider net interest margins, a product of the higher-for-longer rate regime, make that cushion thicker than in the low-rate decade. Second, the composition of bank balance sheets has changed: a decade of regulation pushed the riskiest lending out of the banking system, so the assets being stressed are simply safer than the 2007 vintage. Both effects are real. Only one of them should comfort an investor thinking about the financial system as a whole, a point we return to below.
The Payout Cycle Begins
Mechanically, the results feed each bank’s stress capital buffer, the add-on to minimum requirements that calibrates how much capital it must carry. A clean pass with a shallow drawdown translates into stable or lower buffers, and therefore into headroom. Banks usually disclose their updated requirements and shareholder-return plans in the days after the release; the sequence of dividend increases and enlarged buyback authorisations is as close to a calendar event as equity markets have.
The scale of the headroom matters for the equity story. With requirements unchanged before 2027 and earnings robust, the large banks enter the second half of 2026 with more distributable capital than at any point in the post-crisis era, at the same time as financials have been among the market’s leading sectors. For a market that has been narrowly led by technology, a capital-return cycle in financials offers something the AI trade cannot: yield, buybacks and valuation support that does not depend on a capex narrative, a contrast that has become sharper as the semiconductor complex has turned volatile.
There is a cyclical judgement embedded in every payout decision, and it is worth naming. Returning capital at scale is rational when organic lending opportunities are scarce and shares trade near book value; it is also, historically, what banking systems do near the top of credit cycles, when losses look smallest in the rear-view mirror. The 2026 cohort is far better capitalised than any predecessor that made this mistake, and buybacks are not imprudent at these capital levels. But the speed with which surplus capital is now being converted into distributions rather than lending capacity is itself a statement about where the banks believe returns, and risks, sit in this economy.
A Test That Is Being Rebuilt Around the Banks
The 2026 exercise also confirmed a quieter shift: the Fed stated that these results will not affect large-bank capital requirements before 2027, as the central bank continues revising the programme, including proposals to average results over two years and to disclose more of the models. Each change is defensible on technical grounds; volatility in year-to-year results has long been a legitimate bank complaint. Taken together, they mark a supervisory cycle turning from adversarial to accommodative.
History suggests both readings of that turn deserve respect. The friendlier framework reduces the risk that arbitrary model changes force sound banks to hoard capital. It also reduces the informational value of the exercise itself: a test that everyone passes, on a methodology increasingly negotiated with the tested, drifts towards ritual. The stress test was born because the system’s own risk models had failed catastrophically; its credibility has always rested on being harder than the banks would like. Investors should treat the 2026 pass as evidence of strong bank capital, not as evidence that the examination remains demanding.
Where the Risk Went Instead
The deeper caveat is about perimeter. The stress test examines 32 banks. It does not examine the private credit funds, business development companies and direct lenders that now hold a large and growing share of exactly the leveraged corporate lending that once sat on bank balance sheets. That migration is why bank books look cleaner than in 2007, and it is also why the stress test’s reassurance is narrower than it appears: the same regulation that fortified the tested institutions pushed the riskiest credit into vehicles nobody stress-tests publicly at all.
That perimeter question is not hypothetical in 2026. Redemption pressure across parts of the non-traded BDC and private credit complex has been building all year, and banks are connected to that world through subscription lines, fund leverage and counterparty exposure that the public disclosures capture only partially. The honest summary is that the stress test proves the core of the system can absorb a depression on its own books; it says much less about how the core behaves if the periphery, which now holds the sharpest credit risk, comes under simultaneous strain.
| Question | What the 2026 test answers | What it leaves open |
|---|---|---|
| Bank solvency under stress | Strong: 1.6pp CET1 drawdown against a depression scenario | Model softening as the framework is revised |
| Shareholder returns | Green light for dividends and buybacks into 2027 | Whether payouts prove pro-cyclical at the top |
| System-wide credit risk | Banks’ own books are demonstrably cleaner than 2007 | Untested private credit exposure and its links back to banks |
Investor Implications
Equities. The mechanical consequence of the pass is a visible pipeline of dividend increases and buyback authorisations across the large banks, arriving while the sector already trades with momentum. The capital-return case for financials is now about as clean as it gets; the discipline worth keeping is to distinguish banks returning surplus capital from banks returning capital because growth is scarce. Net interest margins in a higher-for-longer world, fee recovery in capital markets businesses, and credit costs on cards, where the stress test itself located the biggest losses, are the variables that separate the two.
Fixed income. For bank creditors, the test reinforces what spreads already imply: senior bank paper at the top of the system is a rates product again, with idiosyncratic risk concentrated further down the capital stack and in smaller institutions outside the tested 32. The more interesting relative-value question sits between tested bank credit and the untested private credit complex, where yield premia compensate for exactly the opacity the stress test does not reach.
Cross-asset. A well-capitalised banking core changes the character of the next stress event: 2008-style bank-solvency contagion is less likely, while liquidity events that originate in funds and markets, in the style of March 2020, become relatively more important. Portfolio hedges built for a banking crisis are hedging yesterday’s transmission channel.
What to Watch
- Late June to early July: individual banks’ capital-return announcements, the dividend increases and buyback programmes that follow the results, and the updated stress capital buffers disclosed alongside them.
- 1 October: new stress capital buffers take effect for the tested banks, setting binding requirements for the coming year.
- Into 2027: the Fed’s finalisation of the revised stress-testing framework, including result averaging and model disclosure, which will determine how demanding the exercise remains.
- Q3 earnings season: card delinquency and charge-off trends, the loss category the 2026 scenario identified as the system’s largest, against a labour market that has been cooling at the margin.
Conclusion
The 2026 stress test delivered the least surprising headline in American finance: every bank passed. The capital base built since 2008 is real, the payout cycle it unlocks is substantial, and the equity case for the large banks rests on foundations the test genuinely validates. The caution is about what the ritual now measures. A test that no longer moves requirements, run on models being renegotiated in the banks’ favour, examining a perimeter that excludes the fastest-growing pool of credit risk in the economy, is a weaker signal than it was a decade ago even if every number in it is accurate. The system the test was built to protect has moved partly outside it; investors should enjoy the buybacks while remembering that the next stress event is unlikely to announce itself inside the tested perimeter.
Frequently Asked Questions
What is the Fed stress test?
The annual Dodd-Frank Act stress test models how the largest US banks would perform through a hypothetical severe recession over nine quarters. It projects losses, revenue and capital ratios under the adverse scenario, and the results feed each bank’s stress capital buffer, which sets how much capital the bank must hold and constrains dividends and buybacks.
What did the 2026 stress test scenario assume?
The 2026 severely adverse scenario assumed unemployment peaking at 10%, a 30% decline in house prices and a 39% decline in commercial real estate values. Under those conditions the 32 tested banks were projected to lose more than $708 billion, including roughly $200 billion on credit cards, while aggregate CET1 capital ratios fell by only 1.6 percentage points.
What does passing mean for bank dividends and buybacks?
A pass allows banks to plan shareholder distributions above their capital requirements, and banks typically announce dividend increases and buyback authorisations in the days after results. With the Fed confirming that 2026 results will not raise requirements before 2027, the large banks enter the period with unusually large distributable capital.
Does passing the stress test mean the financial system is safe?
It means the tested banks hold enough capital to absorb a severe modelled recession. It does not cover private credit funds, direct lenders and other non-bank vehicles that now hold much of the economy’s riskiest lending, nor does it fully capture banks’ indirect links to them. Strong bank capital reduces one class of risk while leaving the untested perimeter open.
Sources: Federal Reserve, 2026 stress test press release, 24 June 2026; Federal Reserve, 2026 Dodd-Frank Act Stress Test Results; Bloomberg, Fed Stress Test: Big Banks Pass, Paving Way for Payouts; InsiderFinance, Fed Stress Test 2026 Leaves Capital Rules Unchanged; Seeking Alpha, 2026 Fed Stress Test: Banks Got Their Green Light.
Related Reading: The rate regime that underpins bank margins is covered in From Cuts to Hikes: The Fed Rate Hike Repricing and The Hawkish Hold: Inside Warsh’s First FOMC. The untested perimeter is examined in BDC Redemptions and the Private Credit Squeeze, and the case study in non-bank transmission remains the March 2020 liquidity crisis. For the fundamentals, start with our explainer on bank and company balance sheets. The payouts those buffers permitted duly arrived in Wall Street’s record quarter. See also how companies return cash.
Leave a Reply