US CRE Office Distress 2026: The Maturity Wall, the Bifurcation, and the Bank Channel

The Maturity Wall: Size, Composition, and Rolled Balances #

The MBA Commercial Real Estate Finance survey puts total outstanding US commercial and multifamily mortgage debt at roughly 4.7 trillion dollars at year-end 2024, of which about 2.2 trillion dollars carries a non-multifamily commercial property as collateral. Banks and thrifts hold about 1.8 trillion dollars of that stock, agency and GSE portfolios hold roughly 1.0 trillion (concentrated in multifamily), life insurers hold about 730 billion, and CMBS plus other asset-backed issuance accounts for roughly 600 billion. Office collateral is approximately 740 billion dollars across all lender types, with roughly 200 billion securitized in conduit and single-asset single-borrower CMBS.

The maturity profile is front-loaded. The MBA estimated 957 billion dollars of commercial and multifamily mortgages reaching scheduled maturity in 2025, up from a downwardly revised 2024 figure near 570 billion dollars after bank extensions and CMBS A-note hope notes pushed balances forward. Across 2024 to 2026 cumulative scheduled maturity is close to 2.0 trillion dollars net of extensions, with the hardest-to-refinance slice fixed-rate office originated 2014 to 2019 at coupons near 4 percent that now must reprice into a ten-year Treasury that traded 4.0 to 4.7 percent through 2024 and 2025.

The roll math is unforgiving. A loan underwritten in 2018 at a 4.0 percent coupon and 65 percent loan-to-value, against an asset that has lost 35 percent of value and now trades at an 8.5 percent cap, cannot refinance at par. The borrower injects equity, accepts a discounted payoff, hands keys back, or the lender extends. Trepp records the office CMBS payoff-at-maturity rate falling to roughly 25 to 35 percent across 2024 and 2025 against a pre-pandemic norm above 80 percent. The remainder rolls into special servicing, which is the channel where the cycle clears.

Office Vacancy by Major MSA and the Trophy and Commodity Bifurcation #

National office vacancy is near or above 20 percent for the first time in the modern data series. Cushman & Wakefield reported national vacancy at approximately 20.4 percent in late 2024, and JLL recorded a similar 21 to 22 percent figure in early 2025 with availability running 3 to 4 percentage points higher. Major CBDs cluster above the mean. San Francisco direct vacancy ran 35 to 37 percent across 2024 and 2025, Manhattan averaged 17 to 18 percent with availability near 22 percent, Chicago CBD stood near 25 to 26 percent, and Washington DC CBD held 22 to 24 percent. Miami and Nashville recovered toward sub-15 percent vacancy, while Houston and Atlanta remained above 24 percent.

Inside each market the relevant cut is trophy versus commodity. JLL and CBRE leasing data show buildings delivered after 2015 with column-free floorplates and post-Covid HVAC retained pricing power and lease-up velocity, while pre-2000 product with smaller floor plates and limited reinvestment lost rent and occupancy at a steeper rate. Manhattan trophy asking rents held above 130 dollars per square foot at One Vanderbilt and the Hudson Yards spine, while Class B and C product traded at workout valuations 50 to 70 percent below 2019 marks. In San Francisco, Salesforce Tower and 555 California retained tenant demand while Financial District commodity stock cleared at deep discounts.

The valuation gap makes bank loss recognition slow. Green Street reports its CPPI office segment down roughly 37 percent from the March 2022 peak through 2024, and the MSCI Real Capital Analytics CPPI office index shows declines of 30 to 35 percent over the same window. Trophy is a small share of loan count but a large share of value, so blended marks understate commodity loss and overstate trophy loss. That bifurcation is why simple loan-to-value triggers misprice the cycle.

Trepp Delinquency, Special Servicing, and the Workout Pipeline #

Trepp tracks the CMBS overall delinquency rate at roughly 6.5 to 7.0 percent across 2024 and into 2025, within striking distance of the 10.3 percent peak set in July 2012. Office is the binding driver. Trepp recorded CMBS office delinquency of about 11 percent in early 2025 and tracking near or above that level in early 2026, against a 2018 to 2019 average closer to 2 percent. Multifamily CMBS delinquency rose past 4 percent in early 2025 as Sun Belt floating-rate bridge loans originated in 2021 and 2022 hit cap-rate decompression and rent flatlining. Lodging and retail held in the 5 to 6 percent range, and industrial stayed below 1 percent.

Special servicing is the leading indicator. Trepp reports the office special servicing rate moved above 14 percent across 2024 and approached 15 percent in early 2025, the highest reading in the time series. The transfer pipeline is dominated by maturity defaults rather than monthly debt-service shortfalls: loans perform on a pay basis until maturity and then transfer because the borrower cannot refinance. That distinction matters for loss given default. A maturity-default loan with cash flow can support a discounted payoff, a modification with paydown, or A-note B-note bifurcation, all of which realize smaller losses than a deep cash-flow default.

Realized losses on resolved office CMBS loans through 2024 and 2025 ran in the 30 to 50 percent loss-given-default band on commodity assets in central business districts, with single-asset single-borrower deals on Park Avenue, Wacker Drive, and Market Street resolving inside that range. The office loss math is therefore not a tail risk in the abstract: it is a known quantity moving through a known pipeline at a known rate. The question is whether lenders can absorb the realized loss at that pace without breaking capital ratios.

Regional Bank CRE Concentration and the FDIC 300 Percent Threshold #

The FDIC and OCC interagency guidance on CRE concentrations, dating to December 2006, sets two screening thresholds. The first flags banks where construction, land development, and other land loans exceed 100 percent of total risk-based capital. The second flags banks where total non-owner-occupied CRE loans exceed 300 percent of risk-based capital and have grown more than 50 percent in the prior 36 months. Crossing the threshold does not mandate any action, it triggers heightened supervisory attention. As of mid-2024, FDIC call-report-derived counts identified roughly 1,800 US banks with non-owner-occupied CRE concentration above 300 percent of risk-based capital, concentrated in community and small regional banks under 100 billion dollars in assets.

The Federal Reserve Financial Stability Reports of October 2024 and April 2025 noted that CRE concentrations were highest in small banks and that office and multifamily delinquency had risen most at this cohort. Above 100 billion in assets through 2024 the most CRE-concentrated names included New York Community Bancorp (now Flagstar Financial), which carried CRE-to-capital ratios above 400 percent before its March 2024 capital raise led by a Liberty Strategic Capital led group, alongside Valley National, Bank OZK, and a handful of others. Mid-tier names like Webster, Independent Financial, and Pacific Premier sit in the 250 to 350 percent range depending on quarter. The supervisory question is not whether these banks will fail. With the exception of Signature Bank in March 2023, none have. It is how much net interest margin compression and provisioning the franchise can absorb before capital build slows and the board takes a sale.

Hercules tracks the 300-percent-and-above cohort weekly because it is the population most likely to appear in the merger pipeline, either as buyer for scale or as seller. The post-2023 supervisory environment under Vice Chair for Supervision Michelle Bowman in 2025 has shortened approval timelines relative to 2022 to 2024, removing a structural impediment to clearing the cohort.

Conversions, Stress Test Capital, and the Slow-Workout Base Case #

Office to multifamily and lab conversion has moved from a niche talking point to a meaningful relief valve, though small relative to vacant stock. RentCafe and Yardi Matrix tracked roughly 70,000 to 73,000 office-to-apartment units in the 2024 to 2025 pipeline nationally, up from below 25,000 in 2021. The largest market is New York City, where the City of Yes for Housing Opportunity zoning amendment passed in late 2024 and the state-enacted 467-m tax incentive for office conversions in mid-2024 cleared prior blockers. Washington DC runs a Housing in Downtown program that pays per-unit subsidy and tax abatement for qualifying conversions, and San Francisco enacted a downtown revitalization tax abatement under Mayor London Breed that the Lurie administration has retained. Boston, Chicago, and Pittsburgh added smaller programs.

Even at 70,000 units the relief is modest against an estimated 800 million to 1.0 billion square feet of structurally vacant office. JLL and Cushman estimate that 10 to 15 percent of US office stock is structurally suitable for residential conversion by floor plate, light, and core-to-window depth, putting the addressable population at perhaps 100,000 to 150,000 units over a five-year horizon. Lab conversion is more constrained by floor-to-floor height and mechanical loads, and the 2024 downturn in life sciences leasing has reduced lab demand. Conversion is part of the answer for specific assets, not a system-level solution.

The Federal Reserve 2025 Dodd-Frank Act stress test severely adverse scenario assumed a 30 percent peak-to-trough decline in CRE prices, an unemployment peak above 10 percent, and large declines in equity and house prices. The 23 tested banks held aggregate stressed CET1 ratios above 9 percent at the trough, well above the 4.5 percent minimum, and the Fed concluded the largest banks could continue lending under that scenario. The relevant tail is therefore not the G-SIBs and Category III names that take the test, it is the Category IV and below cohort with concentrated CRE books that does not, where supervisory cycles, OCC matters requiring attention, and FDIC informal actions do the work the stress test does not capture.

Implications: Bank M&A, Life Insurers, REITs, GSEs, and the Fed Path #

The base case through 2026 and into 2027 is a slow workout, not a crisis. Bank earnings absorb the realized office loss at a 25 to 50 basis point cumulative provision lift over 2025 and 2026, concentrated in the cohort with CRE above 300 percent of risk-based capital and CBD office above the national mean. The merger pipeline accelerates inside that cohort: two to four announced deals above 25 billion dollars, six to twelve transactions in the 5 to 25 billion range, and continued attrition below 10 billion in assets define the 2026 to 2028 cadence given the Capital One and Discover precedent and the Bowman supervisory posture.

Life insurer exposure runs through general-account mortgage portfolios at roughly 730 billion dollars per the MBA tally. Insurers underwrote at lower loan-to-value than banks (typically 55 to 65 percent), held longer-dated paper, and mark through statutory accounting rather than current fair value, which damps quarterly volatility. The NAIC and state regulators require risk-based capital charges that escalate with delinquency, and realized life-industry losses should run materially below bank rates. Mortgage REITs that hold CMBS and originate floating-rate transitional loans took most of the visible mark in 2023 and 2024, with several large names trading below tangible book through 2025 and resolving the legacy office tail through sales.

The GSE channel is concentrated in multifamily through Fannie Mae, Freddie Mac, and Ginnie Mae, with limited office exposure. Multifamily delinquency at the GSEs has risen modestly but stays below bank and CMBS multifamily readings, and the FHFA loan limit and underwriting framework provides a stabilizing floor on the largest segment of CRE debt. Federal Reserve policy runs the other way: a federal funds rate held at 4.0 to 4.5 percent through 2025 and 2026 keeps refinancing math hard, and a pivot to 3.0 to 3.5 percent across 2026 and 2027 would compress maturity-default loss-given-default by perhaps 5 to 10 percentage points on commodity office. The cycle prices off the FOMC dot plot as much as off vacancy.

Sources #

• Mortgage Bankers Association Commercial Real Estate Finance
• MBA 2024 Commercial Real Estate / Multifamily Finance Loan Maturity Volumes
• Trepp CMBS Delinquency Reports
• Trepp Special Servicing Reports
• Federal Reserve H.8 Assets and Liabilities of Commercial Banks
• FDIC Quarterly Banking Profile
• FDIC and OCC Interagency Guidance on Concentrations in Commercial Real Estate Lending, December 2006
• Federal Reserve Financial Stability Report, April 2025
• Federal Reserve 2025 Dodd-Frank Act Stress Test Results
• Federal Reserve Senior Loan Officer Opinion Survey on Bank Lending Practices
• Cushman & Wakefield US Office MarketBeat
• JLL US Office Market Outlook
• CBRE US Office Figures
• Green Street Commercial Property Price Index
• MSCI Real Capital Analytics
• RentCafe Office to Apartment Conversion Reports
• NYC Department of City Planning, City of Yes for Housing Opportunity
• OCC Semiannual Risk Perspective