top of page
Search

Feasibility Case Study: Underwriting the Hotel After the Recovery on Paper

  • Jul 24
  • 23 min read

July 24, 2026 By Michal Mohelsky, JD, principal of MMCG Invest, LLC



The American hotel industry has spent three years telling a true story that misleads almost everyone who hears it. Revenue per available room sits at record nominal levels. Transaction volume is growing again. The brands report record pipelines and record loyalty membership. Every one of those statements is accurate, and a credit officer who underwrites off them will buy the wrong deal at the wrong price.


Here is the version the numbers tell when you hold them still. Adjusted for inflation, US RevPAR in 2025 ran roughly 10.9 percent below 2019 [2]. Occupancy finished 2025 at 62.3 percent against 65.8 percent in 2019, a gap of three and a half points that has now persisted long enough to stop being a recovery shortfall and start being the market [1]. Every dollar of the headline RevPAR growth since 2019 came from rate, and rate carried the P&L into a cost structure that grew faster than it did: gross operating profit margins slipped from 35.1 percent in 2024 to 34.8 percent in 2025, and hotel EBITDA margins from 23.3 to 22.8 percent [1]. The industry is paying 22.1 percent more in labor compensation for 7.4 percent fewer worked hours than in 2019 [1].


RevPAR is not NOI. That sentence is the whole case study. What follows is how a feasibility study takes it apart for one representative deal: a 120-key franchised select-service hotel, eighteen years old, family-built and family-run, arriving at the moment every franchised hotel eventually reaches, the license crossroads, where the owner must renew and renovate, convert, or hand back the flag.


Tab A "The two RevPARs": nominal RevPAR 2019 $86.56 vs 2025 $100.02, real RevPAR indexed 2019=100 falling to 89.1; Tab B "The occupancy gap": 65.8 percent 2019 vs 62.3 percent 2025, 3.5 point gap; Tab C "The margin ladder": GOP 35.1 to 34.8 percent, EBITDA 23.3 to 22.8 percent, labor hours minus 7.4 percent vs compensation plus 22.1 percent since 2019. Sources 1, 2.)


1. The recovery on paper

Three structural facts define hotel underwriting in 2026, and none of them appears in a trailing twelve month statement.


First, the demand base recomposed rather than recovered. Occupancy has plateaued below 2019 for three consecutive years and drifted down, not up, in each of them [1]. The growth that exists is concentrated: luxury and upper-upscale posted gains in 2025 while economy demand contracted, and the publicly traded economy-weighted franchisors reported declining RevPAR through late 2025 and early 2026 [1]. Interstate-corridor hotels recovered best of any location type, resorts and urban cores worst, which matters directly to the composite below [1].


Second, growth stopped dropping through. HotStats measured that in 2024 the average US hotel spent $1.04 in incremental cost to capture $1.00 of incremental revenue, roughly 75 cents of it labor [31]. Occupancy-led growth now buys revenue at a loss; only rate-led growth reaches the NOI line. That single measurement should reshape every ramp assumption a lender accepts.


Third, the year 2026 supplied its own cautionary exhibit. The FIFA World Cup final was played on July 19, and the host-market data that has emerged is a rate story, not a demand story: match-day RevPAR gains in the host markets came almost entirely from rate, with occupancy roughly flat, and an AHLA survey published May 4, 2026 found approximately 80 percent of host-market hoteliers reporting performance below their own forecasts [29]. Several host markets ran double-digit occupancy shortfalls against expectations. A feasibility study that had underwritten event-driven compression into a 2026 pro forma would already be explaining a miss. Ours excludes event demand as a matter of method.



2. The two P&Ls, or the family hotel problem

The sub-$20 million hotel segment that SBA lenders actually finance is dominated by owner-operating families. AAHOA members alone own roughly 60 percent of all US hotels [26]. Those hotels produce a specific and predictable accounting artifact: books kept to minimize tax, not to present transferable cash flow. The single largest line in any hotel P&L is labor, and in a family hotel a meaningful share of it is donated.


The composite's trailing statement shows total revenue of $3,009,000, of which rooms revenue is $2,949,000 at 66.0 percent occupancy and a $102.00 average daily rate, RevPAR $67.32, plus $60,000 of other operated income. Reported expenses are $1,833,000, for a book NOI of $1,176,000 and a 39.1 percent margin. The broker's offering memorandum divides that book NOI by the $8,400,000 asking price and advertises a 14.0 percent cap rate.

No hotel in America yields 14 percent to a third-party owner. The feasibility normalization explains where the other six points went:


Normalized trailing NOI: $695,300. Margin: 23.1 percent, within half a point of the national hotel EBITDA margin CBRE measured for 2025 [1], which is precisely where an averagely run upper-midscale property should land. The normalization removed 40.9 percent of the seller's reported income, and every removed dollar was real in the sense that the seller collected it, and unreal in the sense that no financed buyer ever will.


So the same asset carries four legitimate yields at once. Fourteen percent on the seller's books. Roughly 8.3 percent on normalized trailing income against the price. Six point seven percent on normalized trailing income against the price plus the mandatory renovation, which is the honest going-in number, because the renovation is not optional. And, as Section 6 will show, 9.3 percent on stabilized income against all-in cost, which is the number the deal is actually bought on. A credit memo that does not state which of the four it is quoting has not yet said anything.


Tab A "The bridge": waterfall from book NOI 1,176.0 minus 90.3 management fee minus 195.0 market labor minus 120.4 FF&E reserve minus 22.0 insurance minus 53.0 tax step-up equals 695.3 normalized, in thousands; Tab B "Four yields, one hotel": 14.0 book cap, 8.3 normalized cap on price, 6.7 normalized on price plus PIP, 9.3 stabilized yield on all-in cost; Tab C "Why growth stopped dropping through": $1.00 incremental revenue vs $1.04 incremental cost, $0.75 labor. Sources 1, 9, 11, 27, 31.)


3. The flag: what the franchise costs and what it returns

The franchise layer is the hotel analog of a ground lease that can fire you. It is the largest operating contract in the deal, it is not assumable, and both its cost and its enforcement have hardened since 2019.


The cost first. The composite's brand stack, royalty at 5.5 percent of rooms revenue, a program and marketing fee at 3.5 percent, and loyalty, reservation and technology charges near 2.0 percent, totals 11.0 percent of rooms revenue, $324,000 on the trailing year. That sits in the middle of the 9 to 13 percent all-in range disclosed across public owner filings [34]. It is also a ratchet: CBRE measured total franchise-related fees rising 3.5 percent in 2024 against 2.7 percent rooms revenue growth, with the loyalty component growing fastest at 3.9 percent [3]. Fees are charged on revenue and grow faster than revenue, and the fastest-growing piece is the one an owner controls least.


What the flag returns is real but uneven. CBRE's brand performance work found a 26 percent cumulative RevPAR spread between the best and worst performing brand families over 2014 to 2024 [2], which means the underwriting question is never branded versus independent; it is which brand, on which terms, for how long. SBA loan-level data makes the same point from the loss side: default rates across hotel flags in the FOIA data range from around 2 percent to over 20 percent by brand [15].


Then the enforcement record, which now has names and dollar amounts. Wyndham disclosed deleting 214 properties and roughly 18,500 rooms in a single year's quality purge [20]. Marriott sued a New York franchisee for $2,603,708 after an unapproved use conversion [22]. It runs the other way too: an arbitrator ordered Choice to pay approximately $780,008 to a franchisee over vendor-program promises in March 2024 [21]. Liquidated damages provisions standardize around 24 to 36 months of royalties; on the composite's stabilized revenue that is a contingent liability of roughly $393,000 to $590,000, more than the equity many family sellers believe they have. And the comfort letter a lender receives at closing gives notice of default and a cure window. It does not give assumption of the agreement, survival of the flag through foreclosure, or forgiveness of liquidated damages. A credit file that treats the flag as collateral has misread the letter.


Two administrative dates belong in every 2026 hotel file. The SBA reinstated its Franchise Directory effective June 1, 2025, and franchisors had until June 30, 2026, three weeks before this writing, to execute the new Franchisor Certification or fall off the directory [11]. A brand whose certification lapsed is an eligibility problem, not a pricing problem. Verify it the week of closing, not the month of application.


Tab A "The stack": from $100 of rooms revenue, royalty 5.5, program and marketing 3.5, loyalty, reservation and tech 2.0, total 11.0, with channel acquisition costs band of 15 to 25 percent of guest-paid revenue per Kalibri shown as context; Tab B "The ratchet": fee growth 3.5 percent vs rooms revenue 2.7 percent, loyalty 3.9 percent, 2024; Tab C "What enforcement costs": Highmark award 780,008, Marriott v. Pride claim 2,603,708, LD formula 24 to 36 months of royalties equals 393,000 to 590,000 on the composite. Sources 2, 3, 15, 20, 21, 22, 32, 34.)


4. Supply stopped. The capital bill did not.

The cleanest tailwind in hotel underwriting today is supply. Net US supply grew 0.2, 0.6 and 0.7 percent in 2023, 2024 and 2025 against a long-run pre-2020 trend near 1.4 percent [5], and the 2026 forecasts cluster between 0.7 and 1.4 percent [1][4][8]. The national pipeline shrank 5 percent year over year in the first quarter of 2026 [8]. Financing costs, tariffed materials and labor did what no downturn quite managed: they idled the spec developer.


Construction economics explain why the pause will hold. The 2026 HVS development cost survey puts a new select-service hotel at $200,730 per key and a midscale extended-stay at $169,901 [5]. The composite's all-in basis, acquisition plus full renovation plus reserves and fees, is $91,013 per key, 45 percent of select-service replacement cost. Nobody builds against that arithmetic, which is exactly what protects the buyer who renovates instead.


But the supply pause has a shadow: the industry deferred its capital cycle through the pandemic, and the brands have resumed collecting. Industry estimates place $12 to $15 billion of deferred CapEx and property improvement plan work due over the next several years [10], a figure we flag as an industry estimate without a published methodology, against measured spending that already runs about 7 percent of revenue nationally versus the 4 to 5 percent reserve most loans carry [9]. Two regulatory items land inside that renovation wave. The refrigerant transition ended new production of R-410A equipment on January 1, 2025; the PTAC units that heat and cool nearly every economy, midscale and extended-stay guest room now transition to A2L refrigerants at roughly 10 to 15 percent higher unit cost, with no drop-in retrofit [33]. And any renovation deep enough to trigger current code brings life-safety and accessibility scope with it. The composite's PIP replaces all 120 PTAC units with A2L equipment for this reason; buying the old inventory would have bought the problem twice.


Tab A "The pause": net supply growth 1.4 percent pre-2020 average vs 0.2, 0.6, 0.7 percent 2023 to 2025 and 0.7 to 1.4 percent 2026 forecast band; Tab B "Build or buy": select-service replacement 200,730 per key, midscale extended-stay 169,901, composite all-in 91,013, 45 percent of replacement; Tab C "The capital bill": measured CapEx 7 percent of revenue and select-service 2,334 per available room vs 4 to 5 percent reserve convention, deferred backlog 12 to 15 billion flagged as industry estimate. Sources 1, 4, 5, 8, 9, 10, 33.)


5. The composite: a 120-key hotel at the license crossroads

The subject is a 120-key, four-story, interior-corridor select-service hotel under a national upper-midscale flag, built in 2008 by the selling family on an I-75 commercial corridor in a north central Florida county seat, a metro of roughly 200,000 people fed by interstate through-traffic, two regional hospitals, a university satellite campus and a distribution belt. Trailing occupancy of 66 percent against a 62.3 percent national year is what an interstate location is supposed to do; interstate hotels are the only location type operating at or near their 2019 occupancy [1].


The franchise agreement, signed at opening, expires in 2028. The brand has offered a new long-term license conditioned on a change-of-ownership property improvement plan. That put three options in front of the buyer, and the feasibility study priced all three.


Convert to extended-stay. The segment's occupancy premium of roughly 12 points over the total industry [25] makes conversion the fashionable answer, and it is the wrong one here. Economy extended-stay supply grew 10.3 percent year over year in early 2025, almost entirely through conversions [25], which is a crowd, not an opening. Extended-stay CMBS delinquency rose from 1.38 percent in 2022 to 7.47 percent by mid-2025 [17], the distress signature of exactly those late conversions. Our research found no named select-service-to-extended-stay conversion with a published cost per key, meaning any such budget is unbenchmarked, and the physical scope, kitchens in 120 rooms, is a gut renovation. Finally, the SBA eligibility screen cuts against it: a hotel must derive more than half its revenue from guests staying 30 days or less [14], and a successful extended-stay strategy walks the asset toward that line. Declined.


Deflag and run independent. The 11 percent fee stack is a real number, but an interstate transient hotel lives on the brand's reservation and loyalty engine precisely because its guests are strangers passing through. There is no defensible US measurement of an independent premium net of fees for this product type, and the brand-family performance spread [2] argues the fee conversation belongs inside the directory, not outside it. Declined.


Renew and renovate. The chosen path: a $1,980,000 PIP at $16,500 per key, building and site scope of $870,000 and FF&E scope of $1,110,000, against a new long-term license. That per-key figure sits inside the published $10,000 to $25,000 change-of-ownership range for the product type and reflects the 30-plus percent post-pandemic escalation in renovation cost [10].


The purchase price is $8,400,000, $70,000 per key. The going-concern appraisal, mandatory for a special-purpose property and performed by a Certified General appraiser who has completed at least four going-concern appraisals of equivalent property in the past 36 months [11], allocated $6,870,000 to real property, $570,000 to furniture, fixtures and equipment, and $960,000 to intangible and business value, and concluded $8,520,000, above the contract price. Had it concluded below roughly 95 percent of project cost, the shortfall would have landed on the borrower as additional equity, a trap that surprises hotel buyers more often than any other single SOP provision [11].


One more screen, because hotels are operating businesses inside the SBA's eyes: the transient test. The trailing guest ledger shows about 92 percent of revenue from stays of 30 days or less, with one project-crew account making up most of the balance. The property passes with room to spare, and the loan file carries an annual length-of-stay reporting covenant so that it keeps passing on the record, not on memory [14].


6. Sources, uses, and the leverage inversion

Readers of our self-storage case study will remember its central finding: at 2026 pricing, an SBA 7(a) loan constant near 10.3 percent sat above every honest storage cap rate, so every storage deal was a projection deal and leverage was negative at the door. Hotels invert that finding, and the inversion is the most useful thing this article has to say.


Hotels are a named special-purpose property under SOP 50 10 8 [11]. That costs the borrower a 15 percent minimum contribution on the 504 side, 20 percent for a new operator, plus the appraisal regime above. What it buys is access to the one SBA structure whose math works: the 504. The 25-year debenture priced at an effective rate near 5.85 percent in early 2026 [30], a loan constant of about 7.6 percent. The conventional first at 7.50 percent carries a constant of 8.87 percent. Blend them at 50/35 and the 504 side of the stack costs 8.36 percent annually against hotel going-in cap rates that HVS places at 8.0 to 8.5 percent for stabilized assets, with first-quarter 2026 transactions averaging 8.6 percent [4]. Neutral to positive at the property line. The 7(a) at Prime 6.75 plus 2.50, 9.25 percent, amortizing over 10 years for its non-real-estate uses, carries a constant of 15.4 percent, which is not leverage at all; it is amortization on a schedule. So the structure below puts everything durable on the 504 and keeps the 7(a) small and short.


Uses, totaling $10,921,500. Real property $6,870,000; FF&E in place $570,000; intangible and business value $960,000; PIP building and site scope $870,000; PIP FF&E scope $1,110,000; working capital and a ramp-and-seasonality reserve $325,000; third-party and closing costs, including the feasibility study, going-concern appraisal, environmental work and PIP architecture, $185,000; and the FY2026 SBA guaranty fee of $31,500, which is 3.5 percent of the $900,000 guaranteed portion of the 7(a) [13].


Sources, totaling $10,921,500. A conventional 504 first mortgage of $4,425,000, 50 percent of the $8,850,000 fixed-asset project (real property plus the entire PIP), at 7.50 percent over 25 years, annual service $392,400. A CDC debenture of $3,097,500, 35 percent, at a 5.85 percent effective rate over 25 years, annual service $236,100. A 7(a) loan of $1,200,000 at 9.25 percent over 10 years, annual service $184,400, funding the guaranty fee, working capital and reserve, third-party costs, the in-place FF&E and $88,500 of the intangible allocation. And cash equity of $2,199,000, 20.1 percent of the project, which covers the remaining $871,500 of intangible value and the $1,327,500 borrower contribution the 504 requires at 15 percent of its project. No seller note was used; had one been, the rules are worth reciting, because they changed: a seller note counts toward the injection only on full standby for the entire loan term and only up to half the required injection, and a seller retaining even 1 percent of equity guarantees the loan personally for two years [11].


Notice what did the sizing. The combined SBA exposure, debenture plus 7(a), is $4,297,500, comfortably inside even the old $5 million aggregate cap. The July 4, 2026 policy change that decoupled the programs and doubled the combined ceiling to $10 million [12] was not the binding constraint here and will not be the binding constraint on most deals this size. Coverage was. Total stabilized debt service is $812,900, a blended constant of 9.32 percent on $8,722,500 of debt, and the structure holds 22 percent equity not because a rule demanded it but because the stabilized numbers do: NOI of $1,015,000 over debt of $8,722,500 is an 11.6 percent debt yield, just clear of the 10 to 12 percent floor hotel lenders now apply before LTV ever enters the conversation [17], and NOI over debt service is 1.25x, sized to the covenant with nothing to spare. The floor binds before the ceiling. That is the sentence to remember from 2026 hotel finance. Where the new $10 million ceiling does bind is one size class up: a $14 million flag-and-PIP acquisition structured as a $10 million 504 project (a $5 million first, a $3.5 million debenture, $1.5 million down) beside a $3 million 7(a) carries $6.5 million of combined SBA exposure, a structure that was simply illegal on July 3, 2026. Nobody has underwritten that class of hotel through SBA before. Someone is about to.


The stabilized year supporting those ratios: 71.0 percent occupancy at a $115.00 rate, RevPAR $81.65, total revenue $3,656,000, a 41.0 percent gross operating profit of $1,499,000, less a 3 percent management fee, $125,000 of stepped-up property tax, $103,000 of insurance and a 4 percent reserve, for NOI of $1,015,000 at a 27.8 percent margin. A 21 percent RevPAR lift over trailing is the renovation dividend, and it is conservative against the one public post-renovation benchmark in SEC filings, Extended Stay America's measured 28.2 percent RevPAR gain and 7.5-point index gain across a 42-hotel program [24].


The coverage path: 0.84x in year one, 1.00x in year two, 1.25x at stabilization in year three, 1.31x in year four. Year one is not covered by operations and was never underwritten to be; the interim facility runs interest-only while the PIP is in construction, the debenture funds at completion, and the $325,000 reserve exists to absorb the roughly $107,000 year-one gap plus the trough months inside it. And in year ten the 7(a) retires, debt service steps down to $628,500, and the structure deleverages itself: the toll tranche, once paid, converts into a coverage cushion no refinancing was needed to buy.


At an 8.25 percent exit capitalization rate, a quarter point wide of today's average, stabilized NOI supports a value near $12,300,000 against $10,921,500 of cost, 70.9 percent leverage through the stack. We present that as a feasibility indication, not an opinion of value; the opinion of value belongs to the appraiser.


Tab A "Sources and uses": stacked bars both totaling 10,921,500; uses RE 6,870.0, FF&E 570.0, intangibles 960.0, PIP building 870.0, PIP FF&E 1,110.0, WC and reserve 325.0, third-party 185.0, guaranty fee 31.5; sources bank first 4,425.0, debenture 3,097.5, 7(a) 1,200.0, equity 2,199.0, in thousands; Tab B "The inversion": constants vs yields bars, debenture 7.62, 504 blended 8.36, bank first 8.87, blended stack 9.32, 7(a) 15.37, against going-in band 8.0 to 8.5 dashed, stabilized yield on cost 9.29, debt yield 11.6, with self-storage 7(a) 10.28 constant shown as the cross-study reference line; Tab C "The path": DSCR 0.84, 1.00, 1.25, 1.31 with 1.25x covenant dashed and year-10 step-down to service of 628.5 annotated. Sources 4, 11, 12, 13, 17, 30.)


7. Ramp, seasonality, and the trough month

Every number in the last section leans on a ramp, so the ramp gets its own discipline.

The pre-pandemic evidence said new hotels reached a 100 percent RevPAR index in about 17 months, per an STR presentation at the 2019 Hotel Data Conference, publicly reported and flagged accordingly, and Cornell's large-sample academic work found occupancy parity in about 1.75 years [23]. Both figures were measured against a rising market. Today's entrant chases an occupancy index that has been flat to falling for three years [1], so we underwrite renovated and repositioned hotels to 30 to 36 months, with year-one occupancy at 65 to 75 percent of stabilized, and we treat anything under 24 months as an aggressive assumption requiring named, documented demand. The composite's own year one embeds renovation displacement of roughly 300 basis points, at the top of the 100-to-330-basis-point range public REITs have disclosed for active renovation programs [24], phased in three periods: pre-renovation baseline, disruption plus ramp, and a post-renovation year measured from the month after completion.


Seasonality gets the same treatment. A 25-year amortization schedule bills the same amount in the September trough as in the March peak, and an interstate Florida market swings hard between them. The feasibility model therefore runs monthly, not annually, states a trough-month coverage figure, and sizes the reserve to the consecutive uncovered months rather than to a rule of thumb. Underwrite the trough month, not the annual mean; an annual 1.25x that conceals three sub-1.0x months is a covenant default with better manners.


A methods note lenders increasingly ask about: the competitive penetration analysis in our studies, occupancy index, ADR index, RevPAR index against a named comp set, is built without STR subscription data, which our sourcing policy excludes. The accepted substitutes are folio-level data from Kalibri Labs, whose dataset draws on roughly 35,000 US hotels and already powers a national appraisal practice [32], CBRE's market and submarket series [1], forward pace from Amadeus where group matters, and a transparent fair-share build-up from the comp set's room counts and segmentation. Appraisal reviewers and SBA lenders accept a penetration analysis whose comp set is named, whose source is disclosed and whose capture logic is shown. The vendor was never the method.


8. Stress, tripwires, and what it has cost others

The bear case holds the renovation budget and the structure and misses the market: stabilized occupancy of 68 percent instead of 71, a $110.00 rate instead of $115.00. Revenue lands at $3,348,000, and after the same cost structure with variable expenses flexing down, NOI comes to $860,000. Coverage: 1.06x. The deal survives; the 1.25x covenant does not, and that is the correct design. A structure whose bear case dips below covenant but stays above cash breakeven, with 22 percent equity and a funded reserve behind it, fails gracefully into a workout conversation rather than a default. The credit decision is whether the institution can live with that conversation.


The tripwires we would write into monitoring, each with a named basis:

  • RevPAR index stalling below 85 percent of the comp set past month 18, the empirical marker of a broken ramp [23].

  • ADR more than 10 percent under pro forma at any checkpoint. Rate misses do not self-correct the way occupancy misses can, and the HotStats incremental-cost finding means an occupancy save at a discounted rate is a margin loss wearing a demand costume [31].

  • Length-of-stay drift toward the 30-day line, tested annually against the transient rule [14]. The mitigation for a soft transient year is longer stays, and the eligibility screen sits directly on that path.

  • PIP schedule or budget overrun beyond the contingency, which reopens both displacement and the brand's cure clock [10].

  • Franchise certification status, checked against the SBA directory after the June 30, 2026 deadline, and the comfort letter's cure window read, not assumed [11].

  • Insurance and deductible drift. Nineteen states and the District allow percentage-based named-storm deductibles [27]; on this asset a 5 percent deductible is roughly $430,000 of borrower liquidity the policy assumes exists. The current soft market, minus 16 percent on catastrophe-exposed property [27], is the moment to buy that deductible down, not to bank the savings.

  • Pricing-display compliance. The FTC's all-in pricing rule for short-term lodging took effect May 12, 2025, following state analogs and a state attorney general settlement with a major brand [28]. Drip pricing is now a litigation exposure, and franchisee booking flows inherit it.


For calibration, the loss record. Hotels and motels, NAICS 721110, show a cumulative 7(a) charge-off rate near 1.8 percent across $44.2 billion of loans since 1995 [15], among the best real-estate-secured cohorts SBA has, and an order of magnitude below the 23 to 28 percent lifetime figure often quoted for the blended food-and-accommodation sector code, which is a restaurant statistic wearing a hotel costume. That is the number to put in front of a credit committee. The number to put beside it: the overall 7(a) trailing default rate reached 4.8 percent in March 2026, a twelve-year high [16], and lodging CMBS delinquency touched 7.31 percent the same month before easing to 6.01 percent by May [17], with $48 billion of hotel CMBS maturing across 2025 and 2026 and roughly 30 percent of all hotel loans maturing in 2026 alone [18]. The asset class is sound; the vintage demands respect. And because the agencies sit this sector out entirely, Fannie Mae's guide excludes hotels by name [19], the refinancing of that wall runs through banks, CMBS and the SBA, which is precisely why the maturing cohort will feed the acquisition pipeline of buyers like the composite's for the next two years.


Tab A "The bear bridge": waterfall 1,015.0 minus 96.0 rate miss minus 103.0 occupancy miss plus 44.0 variable-cost flex equals 860.0, coverage 1.06x vs 1.25x covenant dashed and 1.00x breakeven line, in thousands; Tab B "What it has cost others": Highmark award 780.0, Marriott v. Pride claim 2,603.7, LD exposure band 393.0 to 590.0, named-storm 5 percent deductible 430.0; Tab C "The two default rates": NAICS 721110 cumulative charge-off 1.8 percent vs NAICS 72 sector 23 to 28 percent vs 7(a) overall TTM 4.8 percent, with basis-difference note. Sources 14, 15, 16, 17, 18, 21, 22, 27, 28, 31.)


9. What the credit file should contain

A lender-grade hotel feasibility file in 2026, in the order a reviewer will look for it:


The two P&Ls, side by side: the seller's books and the normalized statement, with every adjustment stated and sourced, and the four yields labeled. The going-concern appraisal by a qualified hotel appraiser with the mandatory land, building, FF&E and intangible allocation, reconciled to the sources-and-uses [11]. The independent feasibility study itself, which for a special-purpose property is expected practice rather than a courtesy [11]. A monthly, not annual, cash flow with a stated trough-month coverage and a reserve sized to it. The PIP as a priced, escrowed capital project with a displacement estimate, never as an operating expense. The franchise file: current FDD, directory and certification status, comfort letter with its cure window highlighted, and the liquidated-damages exposure computed in dollars. The transient-test worksheet and its annual covenant. The insurance program with deductibles expressed in dollars against borrower liquidity. And the structure memo stating what sized the loan, because "the debt yield floor, not the program ceiling" is a sentence that should appear in writing before someone asks why the proceeds are smaller than the borrower hoped.


Outlook

Three forces set the next two years. Supply stays historically low through at least 2027; the pipeline shrank again in early 2026 and nothing in the cost data reverses it [5][8]. The maturity wall converts, month by month, into the acquisition pipeline for well-capitalized small operators, with the SBA stack now able to follow them up to $10 million of combined exposure [12][18]. And the gap between nominal and real performance keeps disciplining anyone who underwrites the headline: until real RevPAR recrosses its 2019 line, every hotel pro forma is a market-share argument, and market-share arguments are exactly what feasibility studies exist to test.


The composite closes with 22 percent equity, a 1.25x stabilized coverage sized to the floor, an all-in basis at 45 percent of replacement cost, and a fifteen-year flag on a renovated asset in the one location type that never really lost its guest. That is not a story about a recovery. It is a story about arithmetic, which is the only story a credit file should ever tell.



Author: Michal Mohelsky, J.D., Principal, MMCG Invest, LLC




Michal Mohelsky, J.D. | Principal | mmcginvest.com 

Phone: (628) 225-1125




Disclaimer: This report is provided for informational purposes only and does not constitute investment, legal, or tax advice. Data presented herein is derived from proprietary MMCG databases and third-party sources believed to be reliable; however, MMCG Invest makes no representation as to the accuracy or completeness of such information. Figures from third-party industry databases have been independently verified and, where appropriate, adjusted to reflect MMCG's proprietary analytical methodology. Statutory and regulatory references are provided for context and must be verified with counsel before reliance. Past performance is not indicative of future results.



Sources

  1. CBRE Hotels Research, Trends in the Hotel Industry, March 2025 edition (approximately 2,600 US hotels) and related CBRE insights briefs, 2024 and 2025 data; CBRE US Hotel Figures, Q3 2025 and Q1 2026.

  2. CBRE Hotels Research, Hotel Brand Performance 2025 (inflation-adjusted RevPAR; brand-family RevPAR spread, 2014 to 2024).

  3. Robert Mandelbaum and Andrew Hartley, CBRE Hotels Research, "The Cost of Franchising, Core and Soft," August 28, 2025.

  4. HVS, U.S. Market Pulse, March 2026 and April 2026 editions (capitalization rates, discount rates, transaction evidence via MSCI Real Capital Analytics).

  5. HVS, U.S. Hotel Development Cost Survey, 2026 edition, July 2026 (cost per key; net supply growth).

  6. Katy Black, MAI, HVS, "Extended-Stay Profitability Face-Off: Economy Vs. Midscale," April 3, 2025.

  7. JLL, 2025 U.S. Hotel Investment Trends Report, February 5, 2026 ($24 billion volume, up 17.5 percent).

  8. Lodging Econometrics, U.S. Construction Pipeline Trend Report, Q1 2026 (6,020 projects; 705,825 rooms; down 5 percent year over year).

  9. ISHC and HAMA, CapEx 2023 study, sixth edition, November 28, 2023 (national $5,147 per available room, 9 percent of revenue; select-service $2,334, 7.3 percent).

  10. Getzler Henrich, "The Looming Crisis" (ISHC-linked deferred CapEx and PIP estimate of $12 to $15 billion); Matthews and Nehmer/HVS renovation cost escalation reporting.

  11. US Small Business Administration, SOP 50 10 8, effective June 1, 2025 (special-purpose property list; equity injection; going-concern appraisal requirements; seller-note standby rules; Franchise Directory and Franchisor Certification).

  12. SBA Policy Notice 5000-879058, May 18, 2026, effective July 4, 2026 (combined 7(a) and 504 limit of $10 million).

  13. SBA Information Notice 5000-872051, August 28, 2025 (FY2026 fee schedule).

  14. SBA Form 2234, Part C (eligibility test: more than 50 percent of gross annual income from guests staying 30 days or less).

  15. SBA 7(a) and 504 FOIA loan-level datasets, data.sba.gov; NAICS 721110 analysis by sbalenders.com (Darren King), March 10, 2025 (approximately 1.8 percent cumulative charge-off; 25,447 loans; $44.2 billion, 1995 to 2024).

  16. Lumos Data, SBA 7(a) loan-level default analysis, July 17, 2026 (trailing-twelve-month default rate 4.8 percent as of March 31, 2026).

  17. Trepp CMBS surveillance as reported by trade press, March through June 2026 (lodging delinquency 7.31 percent March 2026, 6.01 percent May 2026; extended-stay delinquency 1.38 percent 2022 to 7.47 percent July 2025).

  18. CRED iQ via Commercial Observer, February 2025 ($48 billion hotel CMBS maturing 2025 to 2026); Mortgage Bankers Association loan-maturity survey, February 21, 2025 (share of hotel loans maturing 2026).

  19. Fannie Mae Selling Guide, B2-3-01 and B4-2.1-03 (hotel ineligibility for agency execution).

  20. Wyndham Hotels & Resorts, Form 10-K for FY2022, filed February 16, 2023 (deletion of 214 properties, approximately 18,500 rooms).

  21. Highmark Lodging v. Choice Hotels International, AAA Case No. 01-21-0004-5554, award reported by Law360, March 6, 2024 (approximately $780,008).

  22. Trade press reporting of Marriott International's 2024 complaint against Pride Hotel LLC, Queens, New York ($2,603,708 claimed).

  23. Cathy Enz and Linda Canina, Cornell University, "How Fast Do New Hotels Ramp Up Performance?" (3,494 US hotels); STR presentation, 2019 Hotel Data Conference, as publicly reported by trade press (17-month RevPAR index ramp; pre-pandemic vintage, flagged).

  24. Extended Stay America, Inc., Form S-1/A, 2013 (post-renovation RevPAR up 28.2 percent; RevPAR index up 7.5 points, 42 hotels); Pebblebrook Hotel Trust disclosures, 2024 and 2025 (renovation displacement of 100 to 330 basis points).

  25. The Highland Group, U.S. Extended-Stay Hotels Bulletin, March 2025 and subsequent 2025 editions (occupancy premium; conversion-led supply growth of 10.3 percent).

  26. AAHOA and Oxford Economics ownership study, 2021 (60 percent of US hotels; 34,260 properties); AHLA and Hireology, Front Desk Feedback survey, February 20, 2025 (65 percent of hotels understaffed; housekeeping the top gap at 38 percent).

  27. Marsh, Global Insurance Market Index, Q1 2026 (US property rates down 10 percent; catastrophe-exposed down 16 percent); NAIC, "Hurricane Deductibles," June 2025 (19 states plus DC; percentage deductibles).

  28. Federal Trade Commission, Rule on Unfair or Deceptive Fees, 16 CFR Part 464, effective May 12, 2025; California SB 478, effective July 1, 2024; Colorado Attorney General settlement with Marriott, February 7, 2024.

  29. AHLA host-market survey, May 4, 2026 (approximately 80 percent of host-market hoteliers below forecast), and published host-market World Cup performance reporting, June and July 2026 (rate-led match-day gains; flat occupancy).

  30. SomerCor and Eagle Compliance LLC, SBA 504 debenture pricing (25-year effective rate 5.85 percent, January 2026; debenture rates 4.12 to 4.16 percent, February and March 2026).

  31. HotStats, incremental revenue and cost analysis, 2024 ($1.04 of cost per $1.00 of incremental revenue; $0.75 labor); HotStats break-even occupancy analysis, 2020 vintage.

  32. Kalibri Labs, The Kalibri Report and cost-of-acquisition research (channel acquisition costs of 15 to 25 percent of guest-paid revenue; folio-level dataset of approximately 35,000 US hotels).

  33. US EPA, AIM Act Technology Transitions rule (new R-410A equipment prohibited from January 1, 2025; A2L transition cost and retrofit constraints per manufacturer and trade guidance).

  34. RLJ Lodging Trust, Form 10-K (franchise royalty and other fee ranges); published franchise disclosure document summaries for major upper-midscale brands, 2023 to 2025.

 
 
 

Comments

bottom of page