The Great Refinancing Wall
Executive summary
The United States commercial real estate refinancing problem is best understood not as a single wall in one calendar year, but as a rolling capital stack reset. S&P Global Market Intelligence estimates that roughly $946 billion of CRE mortgages matured in 2024, about $998 billion in 2025, $1.148 trillion in 2026, $1.257 trillion in 2027, and $1.138 trillion in 2028. Mortgage Bankers Association survey based estimates are lower because they measure outstanding balances after paydowns, modifications, and survey coverage, but they point in the same direction: $957 billion of outstanding commercial mortgages matured in 2025, $875 billion are scheduled to mature in 2026, and another $652 billion in 2027. The precise number varies by methodology. The strategic conclusion does not. A very large volume of debt must be refinanced, extended, recapitalized, sold, or resolved over the next several years.
Where the risk sits matters as much as how much debt sits there. At the market level, banks and thrifts held $1.844 trillion of commercial and multifamily mortgage debt in the third quarter of 2025, or 37.4 percent of the total, followed by agency and GSE portfolios and MBS at 22.5 percent, life companies at 15.9 percent, and CMBS, CDOs, and other ABS at 13.0 percent. Among domestically chartered commercial banks, smaller banks outside the top 25 held about $2.07 trillion of CRE loans in March 2026, versus about $0.89 trillion at the largest banks, meaning roughly 70 percent of domestic bank CRE exposure sat outside the top 25. The Federal Reserve also notes that regional and small institutions under $100 billion in assets account for a larger share of CRE lending than larger banks, while large bank originations have declined and smaller bank originations grew sharply into 2022 before easing.
The sector split is equally important. MBA data show that, in 2025, 35 percent of hotel and motel mortgages matured, 24 percent of office, 22 percent of industrial, 18 percent of retail, and 14 percent of multifamily mortgages excluding depository serviced multifamily. For 2026, the publicly accessible MBA summary shows 30 percent of hotel and motel loans maturing, 23 percent of industrial, 17 percent of office, and 13 percent of multifamily, again demonstrating that hospitality and office face the heaviest rollover pressure while multifamily is structurally better positioned. S&P’s loan level analysis also estimated that office accounted for about 10 percent of 2024 maturities and remains a major watch point through 2028.
Refinancing is harder because the rate and leverage math changed. As of May 21, 2026, the 10 year Treasury yielded 4.57 percent and the 5 year Treasury yielded 4.25 percent. S&P found that the average interest rate on CRE loans originated in 2024 was 6.2 percent, versus 4.3 percent on mortgages maturing in 2024, nearly a 200 basis point shock. CBRE reported that average commercial mortgage loan spreads on fixed rate seven to 10 year loans at 55 to 65 percent loan to value were 197 basis points in the fourth quarter of 2025, while lender prudence persisted with average commercial loan to value at 60.9 percent, multifamily loan to value at 66.2 percent, debt service coverage at 1.36, and debt yield at 9.8 percent.
The practical implication is straightforward. High quality, well leased, institutionally located assets can refinance. Trophy and near trophy Manhattan office proved that in 2025, with Vornado refinancing PENN 11 for $450 million and Durst closing a $1.3 billion CMBS loan on One Five One. Performing but value impaired assets are still extendable or modifiable, as shown by Ashford Hospitality’s 17 hotel Morgan Stanley pool extension and the Grand Wailea hotel loan modification. Hard cases are not being refinanced on old terms. They are being sold, marked down, or pushed through resolution. Trepp’s 2025 CMBS reappraisal cohort cleared $23 billion of collateral at a median 53 percent discount to origination, with office accounting for more than half the balance.
The thesis for owners is this: through 2028, maturities will reward preparation and punish denial. The best outcomes will not come from waiting for rates to save the deal. They will come from starting early, resetting valuation expectations, matching the asset to the right lender class, and using the full capital stack when senior proceeds alone do not clear the payoff. That is a capital markets advisory problem, not a simple debt placement problem.
Introduction and thesis
The most common mistake in current CRE discourse is to treat the refinancing wall as a temporary traffic jam that will clear if long term rates fall modestly. The data suggest something more structural. S&P’s loan level property records imply that maturities keep climbing into 2027 and remain elevated in 2028. MBA’s survey based maturity volumes show that 2025 was a peak for outstanding balances, but that 2026 and 2027 still require enormous refinancing capacity. Those two views are not contradictory. They measure different things. S&P models maturities from property records and imputes missing dates. MBA surveys servicers for unpaid principal balances at year end, which already reflect amortization, extensions, and workouts. Together, they show a market still digesting debt written in a radically different cost of capital regime.
The most investable thesis is therefore not that “the wall is overhyped” or that “everything breaks at once.” It is that the market is entering a prolonged sorting cycle. Assets with strong occupancy, clear cash flow visibility, modern physical plant, and lender friendly sponsorship will refinance, though often at lower leverage and with fresh equity. Assets with decent income but impaired valuations will be extended, restructured, or recapitalized. Obsolete assets, especially weaker office and selected retail or mixed use collateral, will clear through sales, note trades, discounted resolutions, or default.
That sorting cycle is especially important because the banking system’s exposure remains broad. The Federal Reserve notes that banks hold about half of all CRE debt, and that regional and small institutions account for a larger share of that lending than banks above $100 billion in assets. The Fed’s H.8 data also show that smaller domestically chartered commercial banks held about 70 percent of domestic bank CRE loans in March and April 2026. At the same time, regional and community banks continued to show steadier CRE growth than large banks through 2025. In other words, the maturity wall is not just a borrower problem. It is also a lender portfolio management problem, which is why extensions, loan sales, and private credit capital have all become more important.
Thesis: the great refinancing wall through 2028 will not be resolved by rate cuts alone. It will be resolved deal by deal through a triage process driven by asset quality, sponsor liquidity, lender balance sheet capacity, and the borrower’s willingness to use multiple capital stack tools. Owners that start early and run a disciplined capital markets process can still preserve value. Owners that wait for the old leverage and pricing regime to return are far more likely to lose optionality.
The size and timing of the maturity wall
The first task is to frame scale honestly. S&P Global’s model and MBA’s survey both show a very large multi year problem, but not the same numeric series. That difference is useful, because it shows the range of plausible pressure rather than a single false precision estimate.
| Year | S&P Global estimated CRE mortgages maturing | MBA survey based commercial mortgage maturities | | | | | | 2024 | $946 billion | $929 billion | | 2025 | $998 billion | $957 billion | | 2026 | $1.148 trillion | $875 billion | | 2027 | $1.257 trillion | $652 billion | | 2028 | $1.138 trillion | Public summary not disclosed |
S&P’s series comes from a loan level analysis of nationwide property records and an imputation model for missing maturity dates. MBA’s series comes from its annual survey of servicers and reports unpaid principal balances as of year end, which means balances at maturity are often lower and extensions can move debt into later years. Both are credible. They simply answer slightly different questions.
The most important analytical point is timing. S&P’s series suggests the problem is not fully behind the market even after 2026. MBA’s 2026 release likewise said the maturity wall had begun to shrink from the 2025 peak, but still showed $875 billion maturing in 2026 and $652 billion in 2027. That means any owner with a 2026 to 2028 maturity is still competing for lender balance sheet, agency capacity, CMBS execution windows, and rescue capital against a very crowded field.
Where the risk sits by asset class and lender
Asset class rollover is not uniform. MBA’s public releases show that, in 2025, the heaviest maturity pressure fell on hotel and motel, then office, then industrial, then retail, with multifamily substantially lower. For 2026, hospitality and industrial remain highly exposed, office is still elevated, and multifamily again screens better. The public MBA summary for 2026 does not disclose every property type line item in the snippet that is accessible on the open web, so the table below flags what is public and what is not.
| Property type | Share of outstanding mortgages maturing in 2025 | Share of outstanding mortgages maturing in 2026 | | | | | | Multifamily | 14 percent | 13 percent | | Industrial | 22 percent | 23 percent | | Office | 24 percent | 17 percent | | Retail | 18 percent | Public summary not clearly disclosed | | Hospitality | 35 percent | 30 percent |
This table supports a simple hierarchy. Hospitality faces the sharpest rollover pressure. Office remains structurally challenged because maturity risk sits on top of vacancy and valuation pressure. Industrial still has meaningful rollover volume, but it enters the maturity cycle with stronger capital markets acceptance. Multifamily remains the best insulated of the major property types because maturities are lower as a share of outstanding balances and because agencies remain a very large capital source. Retail sits in the middle, with neighborhood center product generally stronger than enclosed mall or challenged discretionary formats.
Office deserves separate treatment. S&P estimated that roughly 10 percent of CRE mortgages maturing in 2024 were office properties, and its chart shows office maturity balances remaining material through 2028. CBRE separately estimates a $131 billion office debt funding shortfall over the next four years, representing nearly a quarter of office debt originated from 2017 through 2023. Trepp reported that the office CMBS delinquency rate hit a record 12.34 percent in January 2026. That combination explains why office is no longer a conventional refinance story. It is now a market segmentation story, where tenant quality, lease duration, capital expenditure needs, and submarket relevance drive whether the answer is refinance, recap, or resolution.
Lender exposure tells the second half of the story. MBA’s third quarter 2025 debt outstanding data show a market still anchored by banks, followed by agencies, life companies, and CMBS.
| Lender type | Debt outstanding Q3 2025 | Share of total | | | | | | Banks and thrifts | $1.844 trillion | 37.4 percent | | Agency and GSE portfolios and MBS | $1.108 trillion | 22.5 percent | | Life insurance companies | $782.9 billion | 15.9 percent | | CMBS, CDO and other ABS | $642.3 billion | 13.0 percent | | All other holders combined | $553.0 billion | 11.2 percent |
The maturity burden by lender type is even more concentrated. MBA reported that in 2025 about $452 billion of depository serviced mortgages matured, versus $231 billion in CMBS, CLOs, or other ABS, $180 billion in credit companies, warehouse, or other lenders, $64 billion in life companies, and only $31 billion in agency or government backed multifamily and healthcare mortgages. For 2026, the comparable figures were $396 billion for depositories, $200 billion for CMBS, CLOs, or other ABS, $163 billion for credit companies and other lenders, $76 billion for life companies, and $39 billion for agency or government backed multifamily and healthcare.
That distribution is why regional and community banks remain central to the refinancing wall. The domestic bank data show that smaller banks outside the top 25 hold about 70 percent of domestic bank CRE loans, and the Philadelphia Fed found that CRE lending remained steady at local and community banks while shrinking at large banks during 2025. Private credit is becoming more important in this vacuum, but it should not be overstated. Public debt outstanding datasets do not isolate every debt fund cleanly. What they do show is that credit companies and other lenders face sizable maturities, and CBRE reported that alternative lenders such as debt funds and mortgage REITs accounted for 40 percent of non agency closings in the fourth quarter of 2025.
Underwriting, pricing, and what recent outcomes tell us
The refinance math is still restrictive. As of May 21, 2026, the 10 year Treasury was 4.57 percent and the 5 year Treasury was 4.25 percent. S&P’s loan level work found that the average rate on CRE loans originated in 2024 was 6.2 percent, compared with 4.3 percent on mortgages maturing in 2024, which implies roughly a 200 basis point rate shock before any decline in proceeds caused by valuation or higher debt yield standards.
CBRE’s lending surveys confirm that lenders did not simply reopen the old leverage box as volumes recovered. In the fourth quarter of 2024, average underwritten debt yield was 9.4 percent and average loan to value was 64.1 percent. In the first quarter of 2025, debt yield widened to 10.3 percent and average loan to value fell to 62.2 percent. By the fourth quarter of 2025, debt service coverage ratio rose to 1.36, debt yield was still 9.8 percent, average commercial spreads were 197 basis points for fixed rate seven to 10 year loans at 55 to 65 percent loan to value, and average commercial loan to value was only 60.9 percent. The Federal Reserve’s May 2026 Financial Stability Report said banks had eased standards through 2025, but CBRE’s underwriting data show that “eased” did not mean loose.
The GSE lane remains the clearest example of selective flexibility. Freddie Mac’s current K Deal materials show a maximum loan to value ratio of 80 percent with minimum debt service coverage of 1.25 times for fixed rate loans and 1.15 times on the maximum capped rate for floating rate loans, subject to market and product adjustments. Fannie Mae’s supplemental loan term sheet shows maximum loan to value as high as 70 percent and minimum debt service coverage as low as 1.30 times depending on asset class and use of proceeds. Those thresholds are not available to every asset. They are available to assets that fit the agencies’ mandate and quality screen.
Cap rates tell the third part of the story. CBRE’s H2 2025 Cap Rate Survey said all property cap rates were broadly steady in the second half of 2025 even as Treasury volatility persisted, and total transaction volume rose about 19 percent in 2025. But cap rate stability did not mean uniform pricing. Prime multifamily cap rate bands in late 2025 were generally in the mid 4s to low 5s in many major Sun Belt and coastal markets. Industrial class A stabilized product commonly screened around 5.0 to 6.0 percent. Neighborhood retail centers commonly screened around 5.0 to 7.0 percent. Hotels typically screened much wider, often from about 6.0 to 9.5 percent depending on format and market. Downtown office was the clear outlier, with class A stabilized ranges as low as roughly 5.5 to 7.5 percent in New York and Boston but 9.5 to 14 percent in weaker or thinner markets.
Recent market outcomes line up with that pricing map.
A clear refinance case is Vornado’s PENN 11. In July 2025 the company completed a $450 million refinancing on the 1.2 million square foot Manhattan office building. The new loan is five year, interest only, matures in August 2030, and carries a fixed rate of 6.35 percent. Vornado paid down $50 million versus the prior loan. That is what refinance looks like in the current market: prime location, institutional sponsorship, still financeable, but usually with some paydown and tighter structure.
A second successful office example is Durst’s One Five One in Times Square. Reuters reported that Durst closed a $1.3 billion CMBS loan in August 2025 at a 5.865 percent rate with a 56.5 percent loan to value ratio based on a $2.3 billion property valuation. That is another reminder that the securitized market is open for strong collateral, but not for everything.
A restructure case is Ashford Hospitality Trust. In April 2025 Ashford announced that it extended the Morgan Stanley Pool mortgage loan secured by 17 hotels beyond its November 2024 final maturity date. In a separate 2025 annual report disclosure, Ashford said it refinanced the Renaissance Hotel in Nashville in September 2025 with a new $218.1 million non recourse mortgage carrying a two year term and three one year extension options. Hospitality can still access capital, but extensions, shorter terms, and option based structures are common where lenders want more time and more control.
A modification case from the securitized market is the Grand Wailea hotel loan. Trepp’s May 2025 extended CMBS delinquency analysis said the $510.5 million loan received a 63 month modification extension through August 2025 with additional borrower controlled extension options, supported by a 2024 debt service coverage ratio of 1.27. The same Trepp analysis noted that the $1.1 billion MHC 2021 MHC loan secured by 63 mobile home properties was extended 36 months to March 2026 despite a debt service coverage ratio below 1.0. That is the current workout playbook for many cash flowing but imperfect loans: extend, reserve, and buy time.
A balance sheet transfer case is Atlantic Union’s sale of nearly $2 billion of CRE loans to Blackstone in June 2025. Reuters reported that the transaction was executed at a slight discount and gave Atlantic Union room to reduce exposure and redeploy capital, while Blackstone expanded its strategy of buying CRE loans from banks retreating from the sector. In a market like this, one owner’s refinancing shortfall is often another capital provider’s origination opportunity.
The harder edge of the cycle is visible in office defaults and discounted resolutions. Fitch, via Reuters, said a $180 million loan on 261 Fifth Avenue in Manhattan defaulted at maturity in September 2025 and helped push office CMBS late payments higher. Trepp then reported that office CMBS delinquency reached a record 12.34 percent in January 2026. More broadly, Trepp’s 2025 CMBS reappraisal cohort showed $23 billion of collateral clearing at a median 53 percent discount to origination, with office alone accounting for more than half the reappraised balance. That is what happens when extension, fresh equity, and sale execution all fail to bridge the debt and value gap.
A practical action plan for sponsors
The right playbook depends on when the loan matures, how much of the gap is rate driven versus value driven, and which lender class best fits the asset. But the sequencing is consistent.
If the maturity is inside 12 months, the owner should already be underwriting the refinance with today’s debt yield and loan to value assumptions, not with wishful rate assumptions. That means a lender ready package that includes trailing 12 month property level reporting, a forward leasing and capital expenditure narrative, updated rent roll analytics, reserve status, tenant rollover analysis, insurance and tax trend detail, and a sponsor liquidity picture. Current lender behavior supports this caution. Even as standards improved modestly in late 2025, commercial underwriting still centered on near 10 percent debt yields, low 60 percent average loan to value, and materially higher debt service coverage than the pre 2022 market.
Lender selection should follow asset logic, not habit. Stabilized multifamily should usually test agency capital first because agencies and GSE related vehicles hold about half of multifamily debt outstanding and only a small portion of that book matures in any one year. Core industrial and grocery anchored retail can often fit life company or bank permanent money if leverage needs are disciplined. Transitional assets, lease up stories, or assets needing future capital often require debt funds, mortgage REITs, or structured credit. Office should be screened brutally: if the asset is not already proving tenant stickiness, lease durability, and competitive relevance, the senior refinance should be underwritten as an extension or recap conversation, not as plain vanilla permanent debt.
When senior proceeds do not clear the payoff, sponsors need to decide quickly whether the gap is bridgeable. In the current market, four tools recur. The first is sponsor paydown, which is often the cheapest capital if the asset is still strategically important. The second is preferred equity or mezzanine capital, which can work where the property has durable cash flow but the senior lender will not advance enough proceeds. The third is an extension or restructure with reserves, cash management, leasing tests, or partial amortization. The fourth is a sale, note sale, or joint venture recap if rescue capital would overcapitalize a weak asset. Recent outcomes show that markets still reward realism. PENN 11 refinanced because the sponsor accepted a paydown and market terms. Ashford extended and refinanced because it matched the structure to the asset. Distressed CMBS resolutions cleared at large discounts because waiting did not restore value.
Timing matters almost as much as structure. Banks and depositories still control the largest maturity bucket, and smaller banks hold the majority of domestic bank CRE exposure. That means relationship management still matters, especially for extensions and modifications. But it also means borrowers should not rely on a single incumbent lender. If the relationship bank is overexposed to the property type or the market, the borrower should run an external process early enough to create leverage with life companies, CMBS, agencies, or debt funds before the maturity date becomes a workout date.
The decision tree below captures the practical sequence.
The core insight for sponsors is that maturity management is now a portfolio strategy discipline. It requires early valuation work, lender mapping, gap sizing, and parallel exit planning. That is where capital markets advisory earns its keep. The owners who treat the process as strategic tend to keep optionality. The owners who treat it as a last month debt request tend to discover too late that the market already priced the asset for them.
Conclusion
The great refinancing wall is real, but it is not a single number and it is not a single date. Depending on the dataset, it is either a peak in 2025 that remains very large through 2027, or a rolling maturity wave that keeps building into 2027 and stays elevated in 2028. Either way, the market is still in the middle of it. Banks retain the largest exposure, smaller and regional institutions matter more than many headline discussions suggest, and sector outcomes are diverging sharply. Multifamily, industrial, and strong neighborhood retail still have financing lanes. Hospitality can transact, but often with more structure and shorter certainty. Office is financeable only where leasing, tenancy, and location overcome both rate shock and valuation impairment.
For sponsors, the winning posture through 2028 is not optimism for its own sake. It is disciplined preparation. Underwrite proceeds conservatively. Start the process early. Match the asset to the right lender class. Be honest about whether the gap should be funded, restructured, or sold. In the current cycle, preserving optionality is the highest form of value creation. For a firm like Sterling Asset Group, that is the capital markets advisory mandate that matters most: not just finding debt, but designing the decision before maturity makes the decision for the owner.
Open questions and limitations
Publicly accessible summaries do not disclose every property type line item in the 2026 MBA maturity release, so the asset class table above includes the figures visible in public MBA snippets and explicitly flags the missing retail line. That does not change the broader conclusion that hotel and office carry the heaviest rollover stress and multifamily the least among the major sectors, but it does limit precision for retail in 2026.
Private credit is also harder to isolate cleanly in public outstanding debt data than banks, agencies, CMBS, or life companies. MBA’s debt outstanding tables break out finance companies and an “other lenders” style maturity bucket, while CBRE’s market commentary shows debt funds and mortgage REITs are taking a much larger share of new non agency closings. Readers should therefore treat public “private credit” estimates as directional rather than fully census like.
Institutional Thinking. Disciplined Execution. Strategic Real Estate Capital.
Built for sponsors, investors, and counterparties seeking a platform that combines market intelligence, capital structuring, and long-horizon real estate perspective.
Disclaimer
This research essay is provided for informational purposes only. It does not constitute investment advice, investment research for the purposes of any securities laws, a recommendation, or an offer or solicitation to buy or sell any security, strategy, or financial product. The views expressed are as of the date of publication and are subject to change without notice.
The analysis contains forward-looking statements and scenario estimates based on the author’s judgment and stated assumptions. Actual outcomes may differ materially due to changes in geopolitical conditions, policy responses, market liquidity, and other factors. Any quantitative projections are illustrative stress-testing ranges, not predictions.
Information is drawn from sources believed to be reliable (including official agencies and international organizations), but accuracy and completeness are not guaranteed. Past performance and historical relationships are not indicative of future results. Investing involves risk, including the possible loss of principal.