Banking on Stability?

The Swift Centre’s world leading professional forecasters were asked to estimate the likelihood of three important events to US financial stability:

Will a major U.S. bank experience a liquidity crisis by 31st December 2027;
Will U.S. commercial office values drop by 35% before the 31st December 2027; and
If commercial office real estate does drop by 35%, how will that impact the likelihood of a major US bank experiencing a liquidity crisis.

This article summarises the forecasts, integrates recent market data and research, and draws out implications for risk managers and investors.

Key Takeaways

There is just over 1 in 10 (11%) chance that a major U.S. bank experiences a liquidity crisis by the end of 2027. Given the base rate of a major bank crisis is modest and there are currently strong buffers and regulatory oversight, there is a low likelihood we see another Silicon Valley Bank type crisis over the next 18 months.
There is a slightly higher, 13% probability, that before the end of 2027 we see a 35% drop in U.S. commercial office values. Though the Swift Centre forecasters view such a deep decline as unlikely (thanks to stabilizing valuations and reduced exposure at large banks) it remains high enough that financial firms exposed to commercial office real estate should take notice.
If we do see a 35% drop in US commercial office real estate value, then this more than doubles (to 25%) the likelihood that a major U.S. bank will experience a liquidity crisis. The predictors of a significant drop in commercial office value would likely point to a systemic risk of a severe real-estate or economic downturn. Therefore, while a systemic banking crisis is unlikely under current conditions, shocks in commercial real estate could amplify vulnerabilities, making contingency planning essential.

Unlikely, but Plausible and Severe Scenarios

https://viz.swiftcentre.org/results/8Xz4MAg06nk/1756302363012?showQuestions=iaEdu8qAzs4

Each Swift Centre forecaster approached the questions differently, using history, regulation, macro trends and idiosyncratic risks to calibrate their numbers. For the risk of a major U.S. bank liquidity crisis, individual forecasts ranged from 8% to 18%, averaging around 11%.

Numerous forecasters started by reflecting on how often top‑20 banks have run into liquidity trouble in the past. As one forecaster observed: “since 1972 (chosen simply because it was just after the end of the Bretton Woods system), there appear to have been four years in which this has happened: 1974, 1984, 2008 and 2023… this in turn implies that the probability of a major US bank experiencing a liquidity crisis in the next 2 years and 4 months is approximately 16-17%”. Another felt that changes since the 2008 crash lowered the risk of such crises occurring and therefore we should expect such a crisis to be less likely than the base rate since 1970 would suggest.

Other forecasters emphasised the resilience of large banks. One wrote that they were “putting this at a fairly small probability, given no obvious economic weaknesses, or excesses in credit extensions.” They noted that liquidity standards have been significantly strengthened since the global financial crisis and that top‑tier banks have access to abundant resources, leading them to assign only 4% likelihood of such a crisis. Another started with a base rate of one failure every ten years (a 2.5% probability over the 2.5‑year window) and added a few percentage points for poor management decisions and another 3% for potential recessions or interest‑rate shocks, arriving near 8.5%.

Going beyond historical patterns and bank resilience, several forecasters took into account emerging macro and political risks. One forecaster highlighted rising populism, erratic trade policy, growing political pressure on the Federal Reserve, immigration restrictions and the unknown impacts of artificial intelligence on labour markets. In their words, the bar for a top‑20 bank to fail is high, “but extraordinary external pressures cannot be dismissed”. Others echoed this view, adding percentage points for the ballooning federal deficit, potential credit‑rating downgrades and regulatory rollbacks, yet still capped their estimates below 20 % because of the heavy scrutiny and diversified funding bases of the largest banks.

https://viz.swiftcentre.org/results/8Xz4MAg06nk/1756453813201?showQuestions=8qZECRE_MvY

For the likelihood of a 35% drop in U.S. commercial office value, forecasters’ estimates ranged from 5% to 24%, with an average of 13%. Most viewed such a sharp decline as unlikely, pointing to signs of stabilization across the sector. They noted that many firms are bringing employees back to the office, while cheaper rents are drawing new tenants, particularly AI-focused startups, into downtown spaces. Several also observed that crime rates in major metropolitan areas have declined, easing one of the key deterrents to office occupancy. Taken together, these factors suggest that the market may already have absorbed the worst of the recent corrections.

While many saw signs of stabilization, some forecasters cautioned that risks remain: a severe recession, policy-driven labor shortages, or tighter credit conditions could still drive prices down. Others highlighted that although automation and hybrid work may affect office demand over time, these effects are unlikely to materialize at scale within the forecast window. The consensus was that while a 35% collapse is not impossible, the sector’s improving fundamentals make such an extreme outcome a low-probability event in the next two years.

To provide a predictor of whether a liquidity crisis was more or less likely to occur, the Swift Centre forecasters also looked specifically at US commercial office space value’s impact on the likely liquidity of banks. This was chosen as a potential indicator to determine if commercial real estate is a key channel of stress for major banks, even if direct exposure is limited. This is due to such a drop in the forecast period (until end of 2027) potentially being correlated with other economic or societal impacts such as tightening credit, eroding confidence, broader economic decline, or significant transitions in work from AI automation.

Collectively, the Swift Centre panel put the conditional probability of a major bank liquidity crisis at 25% - more than double the baseline estimate. When analysing this question forecasters concluded that a 35% drop in office values would almost certainly signal a severe recession, sharp tightening of credit, or a confidence-driven funding squeeze that could stress even the largest institutions.

Several forecasters highlighted that while the Federal Reserve’s stress tests assume a 30% decline in commercial real estate and still show large banks meeting minimum capital requirements, a real-world crisis rarely unfolds in isolation. One forecaster emphasised correlation over causality, arguing that a collapse in office values would likely coincide with other systemic shocks, making liquidity strains more probable even if direct office exposure remains limited.

Conclusion

The Swift Centre’s aggregated forecasts suggest that a major bank liquidity crisis before the end of 2027 is unlikely but not impossible - a strong contender for a severe but plausible event to monitor. Strong capital buffers, diversified portfolios and improved risk management underpin the low base rate, but idiosyncratic failures exacerbated by policy shocks or unexpected asset declines remain a threat. Commercial office real‑estate values are showing signs of stabilisation, and the probability of a 35 % drop is viewed as modest. Nevertheless, a severe downturn in the office sector could more than double the risk of a bank crisis due to economic contagion and confidence shocks - a reflection that such a precipitous decline would not be an isolated event but a canary in a broader recession or financial shock.

Swift Centre Benefits
Rather than just providing binary predictions, the Swift Centre’s forecast provides a unique probabilistic lens through which to evaluate emerging risks. As outlined above, our forecasters provide their judgments along with a library of rationales - from base‑rate calculations, to comparisons of historical crises, and macro‑risk considerations - to be reviewed, scrutinised, and integrated by decision makers. By weaving these forecasts and explanations together, the Swift Centre creates an asset for organisations and individuals who want to stress test their assumptions and prepare for low‑probability, high‑impact events.

Background Research and Analysis

Baseline probability of a bank liquidity crisis

Stronger capital and liquidity buffers. The Federal Reserve’s 2025 stress‑test results show that large banks remain well capitalised. All 22 banks in the 2025 test stayed above minimum common‑equity tier 1 capital requirements under a severe recession scenario[1]. The test assumed a 30% decline in commercial real‑estate prices and a 33 % fall in house prices[2]. Even under these conditions, the aggregate CET1 ratio declined only 1.8 percentage points[3]. Banking Dive’s analysis notes that banks could absorb more than $550 billion in losses and still maintain regulatory capital[4].
Limited commercial real‑estate exposure at big banks. A Visual Capitalist compilation based on UBS data shows that JPMorgan, Bank of America, Citi and other large banks have 4-21% of their loan portfolios in commercial real estate[5]. On average, big banks have about 11% of their loans in CRE, compared with 21.6% for small banks[6]. Even Wells Fargo, often cited for its office exposure, has around 21% of its loans in CRE[7].
Lessons from 2023 bank failures. The collapse of Silicon Valley Bank in March 2023 was a case of asset‑liability duration mismatch combined with rapidly rising interest rates. The bank invested heavily in long‑dated government bonds while funding itself with demand deposits; when rates rose, the bond portfolio lost value, triggering a run[8]. This idiosyncratic mismanagement (and similar issues at Signature and First Republic) does not necessarily apply to the largest banks, which are subject to stricter supervision and stress tests. Nevertheless, these events remind risk managers that poor hedging and concentration risk can quickly erode confidence.

Factors related to a 35% drop in office value

Stabilisation in commercial real‑estate portfolios. A mid‑2025 review by LightBox finds that office losses at major banks levelled off in the second quarter of 2025. Wells Fargo reported improved performance in previously stressed office assets and reduced its allowance for credit losses on office loans by $105 million, citing more stable valuations[9]. Several banks offloaded risk to private credit buyers or selectively resumed lending, indicating a shift away from worst‑case assumptions[10].
Return‑to‑office momentum. Colliers reports that 2025 is becoming the year when companies act on their long‑term office plans. Many large employers are tightening in‑office mandates, recalibrating space needs and bringing workers back[11]. Although office attendance remains below pre‑pandemic levels, Manhattan’s leasing volume in January 2025 was 36% above its 10‑year monthly average[12], and San Francisco posted its first positive net absorption since 2019[13]. This suggests a floor may be forming under office demand.
Exposure is concentrated in a few institutions. Visual Capitalist’s data show that some mid‑sized banks have high CRE exposure, such as New York Community Bancorp has 57% of its loans in CRE[14] and M&T Bank 40%[15]. However, the largest banks maintain more diversified portfolios[6]. These concentrations are relevant for credit investors but do not pose systemic risk on their own.

Relationship between office value and liquidity risk

Economic contagion. The Fed’s stress tests show that banks can withstand a 30% decline in CRE prices[2], but the scenario includes other shocks (unemployment rising to 10%)[2]. If valuations fell even further, lenders might face broader asset write‑downs and funding pressures.
Confidence shocks and runs. Even if losses are manageable, a crisis of confidence (similar to the runs triggered by SVB’s attempt to raise capital[8]) could push a bank into illiquidity. Under stress, markets often react not to fundamentals but to uncertainty about hidden exposures. Should office valuations plummet unexpectedly, investors and depositors could reassess banks’ real‑estate books, forcing institutions with idiosyncratic weaknesses to seek emergency funding.

[1] [2] [3] [16] Federal Reserve Board - Federal Reserve Board's annual bank stress test showed that large banks are well positioned to weather a severe recession, while staying above minimum capital requirements and continuing to lend to households and businesses

https://www.federalreserve.gov/newsevents/pressreleases/bcreg20250627b.htm

[4] Largest banks sail through Fed’s stress test | Banking Dive

https://www.bankingdive.com/news/largest-banks-sail-through-feds-stress-test/751987/

[5] [6] [7] [14] [15] [17] [18] [19] [20] [21] The U.S. Banks With the Most Commercial Real Estate Exposure