HotelHinge Research  ·  2026 Mid-Year Edition

Hotel CMBS and Whole-Loan Debt, 2026 Mid-Year: What the Public Record Shows

Published Jul 10, 2026 · As reported in CMBS servicer filings · Free to read

Download the PDF

This is the inaugural HotelHinge mid-year hotel-debt report, free to read, with an annual edition to follow. It reads the securitized U.S. hotel mortgage market straight from the public record: CMBS servicer periodic reports, offering documents, and SEC filings. Dollar figures are deduplicated to whole loans, so a headline is never inflated by pari-passu pieces, and every operating figure is a filed or servicer-reported value, never a HotelHinge opinion of value. Sections 1 through 4 read the standing market; sections 5 through 8 add the named mid-year layer, the H1 2026 origination book, the current-tape watch list, three portfolios up close, and the exits. A constant same-store panel joins the annual edition.

1. Credit health: coverage is grinding down

Coverage is the clearest signal in the tape, and it is thinning. Average NOI debt-service coverage on tracked hotel collateral fell from 2.39x in early 2024 to 1.86x in the latest quarter, a steady, near-monotone slide. Physical occupancy has held near 68% for two years, so the compression is a debt-cost and margin story, not a demand collapse.

As reported in CMBS servicer periodic reports. DSCR averaged over property observations with a coverage figure between 0 and 10; occupancy over observations between 20% and 105%. Composition varies by period; a constant same-store panel is coming in the annual edition.

Delinquency, told honestly

The surface not-current rate is volatile and definition-dependent, swinging between 5.7% and 8.25% quarter to quarter, so a single headline number would mislead. The disciplined read is that the hard-distress core, 90-plus days delinquent plus non-performing matured balloons, is grinding up, reaching a two-year high of 3.13% of reporting loans in the latest quarter.

As reported in CMBS servicer periodic reports; CREFC payment-status codes. Rates are shares of loans carrying a non-null payment status in each quarter.

2. The maturity wall: a wave into 2029, not a 2026 cliff

Deduplicated to whole loans (the pari-passu grouping key), the near-term wall is small: only 63 whole loans ($2.47B) mature in 2026. The real crest is 2029, with 388 loans and $17.7B of balance coming due. Counting pari-passu pieces instead of whole loans would overstate the near-term count several-fold.

Whole-loan balances deduplicated by loan group (maximum balance per group), by scheduled maturity year. Balances are mortgage debt, not transaction prices. As filed in CMBS offering documents and servicer reports.

The issuance cycle that built it

The wall is the echo of an origination cycle: a 2017 to 2019 boom ($10.9B / $7.2B / $10.4B originated), a 2020 to 2021 pullback (down to $0.5B in 2021), then a 2024 to 2025 recovery. Ten-year 2019-vintage paper maturing in 2029 is exactly the crest, a clean cause and effect.

No amortization cushion

Interest-only, balloon, rate-type, and negative-amortization indicators as filed at securitization in CMBS offering documents.

3. Where the debt sits

Securitized hotel debt splits sharply between count and value. Texas leads on the number of securitized hotels (887), but New York leads on filed value by a wide margin ($52.0B across just 288 Manhattan-heavy assets).

Filed securitization appraised values per offering documents; not HotelHinge valuations and not transaction prices.

Quality-tier mix

Securitized hotels skew upper-midscale and upscale, the classic select-and-full-service profile. Only 17 luxury properties appear in the tracked pool.

Chain scale joined from the HotelHinge census on matched CMBS properties.

Marquee single loans

Filed whole-loan balances, deduplicated. Mixed-use loans (hotel plus office or retail collateral) are labeled and excluded from hotel-loan counts. Balances are debt, not prices.

Largest hotel portfolios

4. Owners and lenders

Because the hotel debt sits on the same identity spine as the property census and owner filings, it resolves to owners and originators. 371 tracked hotels are attributed to 13 public lodging REITs, and 6,025 of 6,679 securitized hotel properties (90%) resolve to a census property.

REIT attribution from SEC filings mapped onto the census. Top 8 of 13 tracked tickers shown.

Top hotel-CMBS originators

One identity spine

Debt, property, owner, and sale sit on one record. 2,193 census hotels are matched to a tracked hotel-backed CMBS loan, and 7,803 priced hotel sales ($89.59B all-time) sit alongside the debt. The HotelHinge census tracks 46,034 U.S. hotels across 51 markets, each with public-record ownership, sales, and financing attached to the property. Loan balances and transaction prices are always reported separately and never conflated.

Sales are transaction prices from the HotelHinge deals surface; debt figures are mortgage balances. The two are reported separately.

5. What is getting financed now: the H1 2026 origination book

The four sections above read the standing stock of hotel debt. The mid-year layer that follows reads the flow: what priced in the first half of 2026, where the current-tape stress actually sits, three portfolios up close, and who got out. Every loan below is deduplicated to its whole loan, and every balance is mortgage debt, never a transaction price.

The first half of 2026 priced $1.13B of new securitized hotel debt across 35 whole loans and 49 hotels, at a balance-weighted 6.8% coupon and 58% loan-to-value. Every 2026 origination in the book is registered (public) paper: none is a private 144A placement. The pool is the classic select-and-full-service Marriott / Hilton / IHG profile, five- and ten-year fixed terms, with the single largest new loan the Hilton Waterfront Beach Resort in California.

Origination terms (balance, coupon, LTV, underwritten DSCR, per-key) as filed in CMBS offering documents; balances are deduplicated to the whole loan. Balances are mortgage debt, not transaction prices. As reported in CMBS servicer filings and offering documents.

Sponsor concentration in the new book

6. The watch list: where the current-tape stress sits

A watch list is only honest if it reads the current tape. A loan whose last servicer report is from 2021 or 2022, still carrying an old delinquency flag, has exited the active panel; a stale flag is not current stress. Restricting to loans reporting on the current (April 2026 or later) servicer tape leaves 41 whole loans genuinely under stress: 34 in hard distress (90-plus days delinquent, a non-performing matured balloon, or flagged non-recoverable) and 7 in the softer 30-to-89-day band, together carrying $906.4M of current balance.

The marquee stressed asset is the JW Marriott Chicago: a $90.1M whole loan, $79.3M current balance, matured in August 2022 and carried as a non-performing matured balloon, modified. It also sits in the write-down table below.

CREFC payment-status codes: 0 current, A grace, B under 30 days, 1 is 30-59, 2 is 60-89, 3 is 90-plus, 4 performing matured balloon, 5 non-performing matured balloon. Current balances and servicer advances as reported in CMBS servicer periodic reports; balances are mortgage debt, not prices.

Filed re-appraisal write-downs

Where a special servicer orders a new appraisal, the filed value can fall hard. The current tapes carry 12 filed re-appraisal observations across 10 distinct assets (some re-appraised on more than one tape), each cutting the value stamped at securitization by 45% or more. The deepest is the Radisson Oakland at -69.0%; the largest by dollars is the JW Marriott Chicago, marked from its $370.4M securitization value to $186.4M, down 49.7%.

7. Three portfolios up close

Aggregate trends hide the range of individual stories. Three tracked hotel portfolios show how differently the same market treated different books. Read these as portfolio-level trajectories: the operating figures (NOI, revenue, occupancy) are portfolio-wide across all collateral in each deal, while the loan balance is the specific note this trust tracks. Portfolio composition can shift period to period, so the operating overlay carries the reporting hotel count; a strict constant-hotel panel joins the annual edition.

Starwood Capital Group Hotel Portfolio: a deep COVID crater, a partial climb back

Operating figures are portfolio-wide across all 65 hotels; the note is this trust's tracked piece. As reported in CMBS servicer periodic reports. 2026 partial-year figures are omitted from the trajectory.

Wyndham National Hotel Portfolio: the stress contrast

Portfolio-wide operating figures and this trust's servicer P&I advances, as reported in CMBS servicer periodic reports.

Northwest Hotel Portfolio: the steady third

Not every book is a stress story. This portfolio behind a $112.9M whole loan (eight hotels report in the servicer panel) stayed current throughout, its note amortizing steadily while NOI dipped in 2020, recovered above $36M in 2022, and held in the mid-$20M range since, with occupancy near 71%.

Portfolio-wide operating figures and note balance, as reported in CMBS servicer periodic reports.

8. The exits: who got out, and how few

The quietest section is itself a finding. Across all of H1 2026, only 4 whole loans left through a full payoff or a yield-maintenance refinance, together retiring $130.9M of balance, and only 2 portfolio loans saw any new defeasance. Thin exits are exactly what the maturity wall predicts: with the crest in 2027 through 2029, there is little 2026 paper to pay off. Nothing here is padded.

Payoff and refinance activity as reported in CMBS servicer periodic reports; balance paid off is the beginning balance retired. Balances are mortgage debt, not prices.

New H1 2026 defeasances

Two portfolio loans began defeasing collateral in the first half, one hotel each, substituting government securities for a released property rather than paying the loan off.

Defeasance activity as reported in CMBS servicer periodic reports.

Methodology and honest limits

This report is built from the HotelHinge hotel-debt database: 367 hotel-bearing securitizations, 2,334 loan pieces collapsing to 1,588 whole loans, 6,679 decomposed properties, and monthly performance panels spanning 471 servicer periods.

• Sources. CMBS servicer periodic reports and offering documents (Annex A), plus SEC filings for REIT attribution. Every figure is a filed or servicer-reported value; nothing is a HotelHinge opinion of value.
• One-quarter reporting lag. The latest fully populated quarter is the newest data; servicer reporting lags the calendar by roughly a quarter, always.
• Dollar figures are deduplicated by whole loan. Whole-loan balances attach to every pari-passu piece, so naive summation multi-counts. A naive sum reads $254.7B against the correct $58.1B deduplicated universe. All balances use one balance per loan group.
• Trend charts are period averages. The property set changes every period, so a raw average is read as a direction, not a precise level. A constant same-store panel (2,942 hotels matched across Q2 2024 and the latest quarter) is in preparation for the annual edition. Where mean and median diverge, the robust median is reported.
• Filed figures only. Occupancy, DSCR, appraised values, and loan terms are published only where filed in servicer reports or offering documents, and are attributed as reported in CMBS servicer filings. HotelHinge does not publish its own RevPAR, ADR, occupancy, or cap-rate estimates.
• The mid-year named layer. Sections 5 through 8 are drawn from H1 2026 offering documents (new originations and sponsor detail) and from the current servicer tapes (watch-list status, servicer advances, filed re-appraisals, payoffs, and defeasances). The watch list is restricted to loans reporting on the April 2026 or later tape, so a stale flag from an exited loan is never counted as current stress.
• Debt is not price. Loan balances, outstanding balances, and filed appraised values are debt and value figures, never transaction prices. Portfolio-loan balances stay at the portfolio level and are never allocated to a single hotel.