Hotel Underwriting Method: Fair Share Demand, RevPAR Evidence and the Class's SBA Record

A hotel sells something that cannot be stored. A room left empty tonight earns nothing tomorrow, the price of the next night is set again by the market, and the income that has to repay a loan is rebuilt every day from two moving parts, how many rooms were sold and at what rate. That is the reason a hotel study is organized around room nights rather than around square feet or members, and the reason its revenue chapter is only as good as its count of the rooms competing for the same guests.

This article is the hotel chapter of the asset-class group of the series. The general tools are explained elsewhere and are not repeated here: fair share and the penetration index in Trade Area and Demand Methods: Capture Rates, Penetration Rates and Gravity Models, the pipeline count in supply and comparable analysis, the opening years in ramp-up and absorption and the expense budget in operating expense evidence. What follows is particular to lodging. It sets out a segmented demand model as equations, shows what a state tourism office's published series and a state's lodging tax rules can and cannot prove about occupancy and rate, lists the measures a credit officer reads in a hotel file and the text that anchors each, reads the franchise disclosure document, collects the eligibility rules that touch hotels, prints the class's SBA record from the loan cells with its floors, and closes with a short model of an 84-room hotel. The full version of a hotel study, worked through for a Business and Industry guarantee, is the Connecticut hotel worked example. Two archived posts sit near this one: a checklist on how to determine demand for a new hotel development and a market report on where ADR growth cannot outrun new supply.

Room Nights by Segment: The Demand Model Written Out

The market a new hotel enters is its competitive set, the properties a guest would weigh against it, and the quantity that market trades is the room night. A study builds demand segment by segment because each segment grows for its own reasons and each hotel wins a different slice of it. Commercial travel follows employers and their projects, leisure follows attractions, seasons and drive-to population, and group business follows meeting space, which a limited-service hotel with one small room barely has. A single blended growth rate hides all of that, and so does a single blended index. The model below is the one this article uses; every quantity in it has to be evidenced or labelled as an assumption.

1. Test-year demand in each segment equals the set's base-year room nights in that segment, grown at the segment's annual rate for the years between the base year and the year tested.
2. Recaptured demand equals the base-year total times the share the set turned away on nights it sold out, and it is counted only where the study can show those nights.
3. Rooms in the market equal the set's rooms, plus every pipeline room that opens by the test year, plus the subject's rooms; market supply is that count times 365.
4. Market occupancy equals market room nights divided by market supply.
5. The subject's fair share equals its rooms divided by the rooms in the market.
6. Subject room nights equal the fair share times the sum, across segments, of each segment's room nights weighted by the subject's penetration index in that segment and divided by 100.
7. Subject occupancy equals subject room nights divided by the subject's rooms times 365; its occupancy index is that figure over market occupancy.
8. Subject ADR equals the set's ADR grown to the test year times the subject's ADR index over 100; RevPAR equals ADR times occupancy, and the RevPAR index equals the occupancy index times the ADR index divided by 100.

The last line is what makes the model a hotel model rather than a share calculation. Occupancy and rate trade against each other: a hotel can buy occupancy with price, and a projection that sets both indexes above the set without a reason has assumed a better product, a better location or a better brand. The RevPAR index, which multiplies the two, is the figure a reviewer should compare with what the existing hotels in the set achieve.

Rounded as shown; lines may not foot by one unit in the last place.

Table 1 holds every input of this article's three model tables, sourced or assumed. Table 2 runs the model for an illustrative market of six hotels with 520 rooms and a subject of 84 rooms that opens at the start of year 1 and is tested in year 2. Before the subject opens, the set sells its rooms at 76.9%. In the base column the market reaches 156,323 room nights in the test year and holds 604 rooms, so market occupancy settles at 70.9%. The subject's fair share is 13.9%, its weighted indexes give it 20,719 room nights, and it runs at 67.6% with an occupancy index of 95.3, below 100 because its single meeting room wins little group business. The existing hotels fall to 71.4%: every room night the subject sells, other than recaptured demand, is one the set no longer sells.

The stress columns rank the inputs by how much they matter over a short horizon. Halving every growth rate moves the subject only from 67.6% to 66.3%, and dropping the recaptured demand takes it to 65.6%. A second hotel of 100 rooms opening in the same market takes market occupancy to 60.8% and the subject to 58.0%, even though the subject's occupancy index does not change. In this illustration, over a two-year horizon, the count of rooms decides more than the growth rate does, which is a reason for a reviewer to check the pipeline first and for a study to date its pipeline search.

Recaptured demand needs particular care. It is real where the set fills on enough nights to turn guests away, and a study can show it from the set's daily occupancy: how many nights each property sold out, in which months, and in which segment. Where that evidence is missing the line should be zero. Demand induced by a new generator, such as a hospital expansion or a new plant, belongs on its own sourced line with its own start date, not inside a growth rate. The base-year segment figures come from the set's own results, which the public record does not publish property by property; the OCC's handbook treats commercially available industry performance studies as an important comparative reference that a bank should use in hotel underwriting ^[14]. Growth assumptions can at least be set against a published series: Hawaii's statewide room demand rose 2.0 percent in the nine months to September 2026, to 11,748.0 thousand room nights, while supply rose 0.1 percent ^[15]. The public series in the next section cannot replace the competitive-set work; they can check it.

A State Tourism Office Series: What Published Occupancy and Rate Can Prove

A state agency can publish hotel performance every month, and Hawaii's Department of Business, Economic Development and Tourism (DBEDT) does: its report for September 2026 gives statewide RevPAR of $226, ADR of $325 and occupancy of 69.7 percent ^[16]. The report is built from hotel survey data compiled by STR, Inc. ^[17], weighted by geography and by class of property ^[18], and for that month the survey covered 165 properties with 47,742 rooms, 83.4 percent of the islands' lodging properties of 20 rooms or more ^[19]. It splits results by chain class, from a Luxury Class ADR of $587 to a Midscale and Economy Class ADR of $200 ^[20], which matters because a study should compare its subject with the class it will compete in, not with a state average.

The definitions a study uses should be the ones a regulator uses. The OCC's handbook computes ADR as room revenue divided by occupied rooms, leaving out complimentary rooms and other occupancies that earn nothing, and RevPAR as ADR times the occupancy rate ^[21]. Those identities let a reader test any set of published figures for consistency, and Table 3 does so from DBEDT's own table of room nights and revenue ^{[1] [2]}. Statewide, September supply of 1,717.3 thousand room nights against demand of 1,196.8 thousand gives occupancy of 69.7%; revenue of 388.8 million dollars over that demand gives an ADR of $324.87, against the $324.84 the report prints ^[22], and RevPAR of $226.40 against $226.39. The small gaps come from the rounding of the totals to tenths, and they show the figures hang together. A study that quotes occupancy, ADR and RevPAR from different sources, or for different months, will not pass this test, and a reviewer should run it.

The same table carries the second lesson. On September 7, 2026, Hurricane Lowell reached Kauai and Niihau, and DBEDT links the storm to the fall in the island's hotel occupancy ^[23]. Recomputed from the published room nights, Kauai's occupancy went from 74.2% in September 2025 to 47.2% a year later, and RevPAR from $283.98 to $162.16. Revenue fell further than demand because rate fell too: the report records the island's ADR as down 10.4 percent and RevPAR as down 43.0 percent ^[24]. One event took about a third of a month's demand, and in Kauai's case the loss compounded because rate weakened along with occupancy. The statewide series also shows seasonality: the nine months to September ran at 75.6 percent occupancy ^[25], above the month itself, so a single month is not a year and an annual average is not a month.

What a series like this cannot do is just as plain. It describes a state, an island, a region or a class, never the subject's competitive set. It leaves out properties with fewer than 20 lodging units, among them small bed and breakfasts and individually rented condominiums ^[26], so a motel with fewer than 20 rooms, a property industry 721110 also holds, would sit outside its frame. And its categories move: DBEDT warns that STR reassigned some properties to new categories from January 2026, so its 2025 figures may not match earlier reports ^[27]. A study should use the series for direction, seasonality and event history, cite the month and class it took, and build the subject's own figures from the set.

Lodging Tax Collections: What a Room Tax Base Measures

Where no tourism office publishes performance, a study may be tempted to work backwards from lodging tax collections: divide the tax by the rate and call the result room revenue. The arithmetic is simple, but the tax base is not the hotel market, and Texas shows why. The state levies 6 percent of the cost of a room ^[28], collected by owners, operators or managers from guests renting a room or space costing $15 or more a day ^[29]. Local layers sit on top, generally at up to 7 percent for a city or county ^[30], and in Houston a guest pays a state tax, a city tax, a county tax and a sports authority tax on the same night ^[31].

Table 4 sets out how each rule moves an estimate built from tax receipts away from the revenue of a competitive set.

Read together, the rules make a tax-based figure a measure of a jurisdiction's taxable lodging, not of a hotel market: useful for a long trend, for the size of a county's lodging economy and for the direction of a recovery, and unsuitable as a competitive set's revenue. A study that uses one should state the jurisdiction, the rate it divided by, the exemptions it could not remove and the months covered, and it should reconcile the result with a count of rooms. Two more cautions apply. Texas returns fall due on the 20th of the month after the month taxed, or after the quarter for a quarterly filer ^[38], so collections booked in one month describe stays in an earlier one. And the tax is collected from guests ^[29] and passed to the state, so it is not hotel income; the Connecticut worked example carries its state's room tax as a memo line for that reason.

The Measures a Credit Officer Reads in a Hotel File

The OCC's handbook says successful hotel lending requires specialized knowledge and an adequate understanding of the hospitality business ^[39], and it gives reasons. Hospitality properties have a history of considerable volatility in income and value, and a hotel's operations can be complex, with a sizable component that is not real estate ^[40]. Full-service hotels earn much of their income outside the rooms, from banquets and food and beverage, while limited-service hotels and motels offer little or no food service and limited meeting space ^[41]. A flag brings a central reservation service and guest loyalty programs ^[42], and keeping it can require rigorous maintenance and upkeep that the handbook says should be factored into operating expenses ^[43]. Among the property factors it lists are the franchisor's reputation, the franchise agreement's duration and termination rights, the property's age and condition, the state of the furniture, fixtures and equipment (FF&E) and revenue seasonality ^[44].

Coverage is where these measures meet the loan. The handbook says properties with volatile cash flows, hotels among them, may warrant a higher debt service coverage ratio ^[45], and it treats an amortization of generally not more than 20 years as the hotel end of its range for income property ^[9]. The Business and Industry rule computes its coverage ratio on EBITDA net of expected replacement capital expenditures ^[46], which in a hotel is the FF&E reserve. For 504 lending, 1.15x is the minimum the site's SBA program page states, on historical results or on projections two years out ^[13].

Conventions. This article follows the conventions of the Loan Analytics worked examples. Coverage is tested against the program floor the site's program page states, where the program has one, and against a stated 1.25x lender assumption, and the higher of the two binds. Equity is measured at the loan the cash flow supports, with fees recomputed on that loan. An acquisition is tested on the seller's historical earnings, not on projections.

Table 5 lists the measures, how each is computed and what anchors it.

Two items in that list change how a study is read. The first is the collateral line. A hotel appraisal may carry values for FF&E and for intangibles such as goodwill beside the market value of the real estate ^[49], and the Business and Industry rule makes a lender deduct any business or going concern value from the reconciled market value before it discounts the collateral ^[50]. A feasibility study is not an appraisal and does not value the hotel, but it should tell the lender which parts of the price its cash flow rests on: the flag, the reservation system and the management are not real estate. The second is classification. In the handbook's classification of commercial real estate, hotels and motels count as non-owner-occupied ^[51], so the same hotel can be an operating business in the borrower's hands and non-owner-occupied real estate in the bank's classification, and a study written for a bank should not lean on owner-occupied treatment.

The Franchise Disclosure Document: Fees, Territory and the Brand's Own Record

A franchised hotel's cost line starts in a federal disclosure document. The Franchise Rule makes the franchisor state its initial fees and any conditions under which they are refundable in Item 5 ^[52], and tabulate in Item 6 each further fee the franchisee owes the franchisor or an affiliate, along with charges either of them imposes or collects for a third party ^[3], with fee types that run from royalties to remodeling, advertising, advertising cooperatives, audits, transfers and renewals ^[53]. Where a fee can rise, the document must give the formula for the rise or its maximum ^[54]. A study should take royalty, marketing, reservation and loyalty charges from that table, state the revenue base each is charged on, and carry any scheduled step-ups into the projection years.

The handbook then sets a test: franchise fees underwritten at the higher of actual or 4 to 6 percent of total revenue ^[4]. The two halves point in different directions. Where a brand's own schedule is lighter than the band, the band governs; where it is heavier, as the illustrative brand in Table 7 is at 9.6% of total revenue, the actual terms govern and the band is no comfort.

The rest of the document bears on supply and on risk. Item 12 tells the franchisee whether it receives an exclusive territory, and a franchisor that grants none must say so in the words the rule prescribes, including that the franchisee "may face competition from other franchisees, from outlets that we own, or from other channels of distribution or competitive brands that we control" ^[55]. A fair share model that counts the subject's rooms but not the brand's right to open another flag nearby has missed a source of supply. Item 17 must say what renewal means, and where it applies that a renewing franchisee may be asked to sign a contract on materially different terms ^[56], so a loan that outlasts the franchise term carries renewal risk. Item 19 allows financial performance figures only where the franchisor has a reasonable basis and the figures are in the document ^[57]. Outside Item 19 the rule leaves two openings: a franchisor may hand over the actual records of an existing outlet the buyer is weighing, or add to its Item 19 figures ^[58]. A hotel purchase is the first case, and those records are the seller's history rather than a brand average. Either way a study can use franchisor figures as a check on its projection, never as the projection. Item 20 reports the franchised outlets terminated, not renewed, reacquired by the franchisor or closed for other reasons ^[59], which gives a study the brand's own account of how its outlets have fared.

SBA adds its own gate. It keeps a directory of the franchise and other brands it has reviewed and judged eligible for its financial assistance ^[60], brands that meet the FTC definition of a franchise must be in it to obtain SBA financing ^[61], and SBA warns that a place in it is no seal of approval and guarantees nothing about how the business will do ^[62]. In a 504 project the money for the flag cannot come from the 504 loan: ancillary business expenses, franchise fees, working capital and advertising among them, are not payable from its proceeds ^[63]. The franchise application fee and any opening marketing charge therefore need a source other than the 504 loan, the borrower's cash for example, and the sources and uses statement should show which; the series' sources and uses article builds that statement in full.

The Eligibility Rules That Bear on a Hotel

Industry 721110 holds establishments "primarily engaged in providing short-term lodging in facilities known as hotels, motor hotels, resort hotels, and motels", which may also run food and beverage, recreation, conference and convention services ^[64]. Casino hotels, with a casino on the premises, have their own industry, 721120 ^[65], while a hotel or motel with limited gambling such as slot machines and no casino stays in 721110 ^[66]. The line matters for SBA because 13 CFR 120.110(g) lists among ineligible businesses those "deriving more than one-third of gross annual revenue from legal gambling activities" ^[67]. A study for a hotel with gaming machines should report gaming revenue as its own line and its share of the total.

The passive-business rule is the second test. SBA business loans exclude landlords and developers whose loan would buy or improve assets they do not themselves use or occupy, Eligible Passive Companies excepted ^[68]. Where a hotel is held in two companies, one owning the real estate and one running the hotel, the exception can cover the structure, on conditions that include these. The property company may spend proceeds only on property it rents to the operating company, or companies, that run the business, or on a change of ownership among its own owners ^[69]; the lease term, including renewal options that belong solely to the operating company, has to reach the loan's maturity ^[70]; the operating company joins the loan, either as a co-borrower or by guaranteeing it ^[71], and it must be a co-borrower when a 7(a) loan also pays for its working capital or other assets ^[72]; and the rent may cover no more than debt service and the property company's direct costs of ownership ^[73]. A study should name which company operates the hotel, which employs its staff and which holds the franchise, because the cash flow it projects belongs to the operator.

The occupancy rules come third, where part of the building is let to someone else, a restaurant tenant or a retail bay for example. In a new building the borrower may lease up to 20 percent permanently if it occupies and uses at least 60 percent ^[74], with plans to take over part of the remaining space inside three years and all of it within a decade ^[75]; where the building already stands, permanent tenants may take up to 49 percent, provided the borrower itself occupies and uses at least 51 percent ^[76]. A study should measure the leased space against the rentable area and print the result.

The 504 structure is the fourth test. Under the usual pattern the borrower puts in a tenth, a bank lends half and the 504 loan supplies the other 40 percent ^[77]. The borrower's part rises to 15 percent or more in two cases, when the business (or the operating company behind an Eligible Passive Company) has run for two years or less ^[78] and when the building qualifies as limited or single purpose; where both hold, it rises to 20 percent ^[7]; either condition also requires the bank loans to reach 50 percent of project cost or more ^[6]. The site's SBA program page places hotels among the special-purpose buildings that have few other uses ^[79]. Whether a particular hotel building counts as limited or single purpose is not the study's call, and a study should show the structure both ways when the point is open. Size is the fifth: SBA caps any single 7(a) loan at $5,000,000 ^[80], so a hotel project whose loan would be larger needs another structure, 504 among them.

For a Business and Industry guarantee, tourist and recreation facilities, hotels and motels operated as public or private commercial enterprises among them, are eligible projects ^[81]. Owner-occupied hotels and motels "are only allowed when the pro rata value of a direct owner's living quarters, based on square footage, is deducted from the use of loan proceeds" ^[82]. Rural status turns on population, excluding places of more than 50,000 inhabitants and their urbanized fringe ^[83], though fringe areas the Under Secretary finds rural in character are not excluded ^[84] and those findings are made project by project ^[85], and the B&I program article reads how that definition and the rest of Part 5001 bear on a project's location.

Licensing is the last gate, and it is a state's. In Florida each public lodging establishment must hold a license from the state's division, and the license cannot be transferred from one place or one individual to another ^[86]. A new establishment needs its license in hand before it opens its doors ^[87], operating without it is a second degree misdemeanor ^[88], and licenses are renewed every year ^[89]. A study should list each license the hotel needs, who must hold it and when, and place the dates in its opening schedule, because a change of owner can mean a new application rather than a transfer.

The Hotel and Motel Record in the SBA Loan Cells

Every loan figure in this section belongs to one NAICS code, 721110, which holds hotels and motels together; the dataset cannot separate limited-service from full-service properties, or a motel from a resort, so each count below describes the whole code. From FY2017 to FY2026 the national 7(a) cell for the class counts 8,942 ^[90] approvals for $21,538,178,600 ^[90], and the 504 cell counts 2,885 ^[91] approvals with debentures of $6,679,237,290 ^[91]. The median 7(a) approval in the code is $2,200,000 ^[90], with the middle half of loans between $1,144,250 ^[90] and $3,602,575 ^[90]; the median 504 debenture is $2,048,000 ^[91], and a debenture is only the CDC's part of a project that also carries a bank loan and the borrower's contribution.

Brands run through the record. Of the class's 7(a) approvals, 64.69 percent ^[90] carry a franchise code, and of its 504 approvals 62.98 percent ^[91] do. The cell counts approvals that carry a franchise code; it does not break them down by brand, and nothing in it compares how the franchised loans performed with the rest.

The yearly counts move. The 7(a) cell holds 1,128 ^[90] approvals in FY2017, 1,361 ^[90] in FY2021, 613 ^[90] in FY2023 and 814 ^[90] in FY2026. The 504 cell holds 372 ^[91] in FY2017, 199 ^[91] in FY2021, 382 ^[91] in FY2025 and 264 ^[91] in FY2026. Nothing in the cells explains these movements, and this article offers no explanation of its own. How the cells are built and read is set out in How to Read a Loan-Outcome Page: Counts, Floors, Ranks and the Keys Behind Every Figure, and how a NAICS code becomes a class in the article on codes and classes. The national pages for the class are 7(a) hotel loans and 504 hotel loans.

Loan Data Behind This Article

Every figure in this table comes from the Loan Analytics loan dataset, analytics-loan-data freeze-3 (commit 360c856), and carries its key. SBA rates are approval-cohort floors and are withheld when fewer than 30 loans stand behind them.

What the Hotel Floors Show Against the Program, and What They Cannot

The charge-off figures are approval-cohort floors: each counts the disbursed loans from the FY2017 through FY2021 approval years with a charge-off recorded by the file's cut date, and any later charge-off can only push it up. On that basis the hotel and motel class's 7(a) charge-off floor, one code covering both, comes to 0.99 percent ^[90]; for every 7(a) loan of those approval years the matching figure is 6.33 percent ^[92]. Over a fixed window of five years from approval, hotel 7(a) loans show 0.27 percent ^[90] against 3.72 percent ^[93] across all 7(a) lending; by dollars, the class floor is 0.49 percent ^[90] and the program's 1.95 percent ^[94]. The hotel 504 floor is 0.55 percent ^[91], and across all 504 loans of those years it is 0.35 percent ^[95]; the hotel figure rests on 6 ^[91] charge-offs among 1,085 ^[91] disbursed loans, so a handful of later losses would move it.

Each of these is a floor for a stated reason. In SBA's public release a disbursed hotel loan that is still running, neither cancelled nor repaid nor charged off, carries the status EXEMPT under FOIA Exemption 4 ^[96], and 1,804 ^[90] of the 4,753 ^[90] hotel 7(a) loans in the charge-off cohort are still in that state. The article on approval-cohort floors shows why a rate built that way errs only on the low side. Cancellation is a different ending, an approval that never disbursed: the class's 7(a) cancellation floor on FY2017 through FY2024 approvals is 8.76 percent ^[90], with a ceiling of 8.81 percent ^[90], the rate that would hold if the approvals not yet disbursed all fell away, against 11.24 percent ^[97] for the program; the 504 class floor is 13.13 percent ^[91], ceiling 19.16 percent ^[91], against 13.60 percent ^[98]. Cancellation is not charge-off, and the two measure different endings, so this article never adds them.

What the floors can tell a study is limited and worth stating exactly. When the class floor sits under the program figure, that is no evidence a particular hotel is a safer credit, and a higher floor would not prove the opposite: the file records no occupancy, rate, brand or market for any loan, and it offers no cause for the gap between one class and another. The floors frame the question, and the failure modes have to come from the market evidence. The agencies' 2023 policy statement on commercial real estate workouts supplies one in its hotel example: a permanent loan of $7.9 million on a stabilized 3-star hotel ^[99], whose occupancy fell after a well-known 4-star hotel opened within a mile, while the borrower cut room rates and still lost bookings ^[100]. By the end of the five-year term coverage had fallen to 0.95x ^[101], and the appraisal on file had not caught the lost occupancy ^[102]. That is the second-entrant column of Table 2 played out, followed by the rate column of Table 7. The Kauai figures add the event, and DBEDT's note that Maui County hotels were still affected by the August 2023 wildfires in September 2026 ^[103] adds the length of a recovery. A study's stress columns should be built from these three: new supply, rate cutting and a disruptive event, each sized from evidence.

A Short Model: An 84-Room Hotel Financed Under 504

Table 7 carries the demand model's output into a test-year income statement and a 504 coverage test. The subject is an illustrative 84-room limited-service hotel built by a new company, at an assumed $135,000 a room. Because the borrower is new and the building is taken here as limited or single purpose, the borrower contributes 20 percent ^[7], the bank first lien takes the 50 percent the rule then requires ^[6] and the debenture covers the rest. The first lien carries an assumed spread over 4.64 percent, the 5-year Treasury constant maturity on November 6, 2026 ^[8], and amortizes over 20 years, the hotel end of the handbook's range ^[9]. The debenture carries an assumed increment over 4.86 percent, the 10-year constant maturity on that date ^[10], standing in for a rate SBA fixes with the Treasury Secretary's approval ^[11], and runs 25 years ^[12].

The base column sells 67.6% of its rooms at $159.18, a RevPAR of $107.57 and a RevPAR index of 97.2 against the set. Rooms revenue of $3,298,047 and total revenue of $3,429,968 leave net operating income of $1,076,173 after an FF&E reserve at the low end of the handbook's band ^[5], a margin of 31.4% that sits inside the handbook's range for limited-service profit margins ^[48], a rough comparison because the handbook does not say which line its margin is measured on. Franchise fees at the brand's assumed 10.0% of rooms revenue come to 9.6% of total revenue, above the band's upper end, so the brand's own terms govern ^[4]. Debt service on both liens is $822,017 and coverage is 1.31x. Set the site's 504 minimum of 1.15x ^[13] beside the 1.25x lender assumption and the assumption, being stricter, is the test that counts; the projection clears it by $48,652 of net operating income.

The base case passes and every stress column fails. Lowering every segment index by a tenth cuts occupancy to 60.8% and coverage to 1.06x. Lowering ADR by a tenth leaves occupancy alone but takes coverage to 0.98x, because a rate cut saves almost no cost: the model costs the rooms department per occupied room, not per dollar of rate, so nearly all of the lost revenue reaches net operating income. A second hotel of 100 rooms takes occupancy to 58.0% and coverage to 0.96x. Under this series' conventions a failing column is not argued away: the debt is cut to the amount whose payments the net operating income covers at 1.25x, the lender assumption that binds here, fees are recomputed on the smaller loan and equity is measured against it, as the articles on debt service coverage and on sources and uses show. The Connecticut worked example carries the full version for an existing hotel bought and renovated under a Business and Industry guarantee, including its opening ramp and its sensitivity tests.

Where a Hotel Study Should Be Pressed

The set was chosen, not found. A competitive set drawn to include weak properties flatters the subject's indexes. A study should say why each hotel is in the set and why a nearby hotel is not, by location, class, age and the segments it serves, and should show the set's own occupancy and rate before the subject is added.

The pipeline is stale. In the illustration the count of rooms moved the answer more than growth did, as the second-entrant column shows. The pipeline search should be dated, cover projects approved as well as those under construction, and include the brand's own right to open nearby under Item 12 ^[55].

Both indexes are above 100. A new hotel may earn a rate premium or an occupancy premium over its set. A projection that claims both needs a stated reason for each, and the study should also show the result at an occupancy index of 100.

A state series stands in for the market. A statewide or class figure gives direction and seasonality; DBEDT's own report excludes small properties ^[26] and has reclassified others ^[27]. It is not the subject's set and should not be quoted as if it were.

Tax receipts are read as revenue. A room tax base includes short-term rentals ^[32] and leaves out stays of 30 days or more in whole or in part, from the day of written notice where the guest gave it beforehand and after the first 30 days where the guest did not ^[34]; the tax itself is the guest's money. A receipts-based figure needs its exemptions stated and its rooms counted.

The franchise line uses the band when the brand costs more. The handbook's test is the higher of the actual terms and the band ^[4]; a study should take the actual terms from Item 6 ^[3] and show any scheduled increases.

The reserve and the brand's capital demands are missing. Keeping a flag can require upkeep the handbook says belongs in operating expenses ^[43], and the reserve should sit at a stated share of revenue before coverage is tested ^[46].

The amortization is the borrower's wish. Coverage computed on a longer schedule than the lender will grant is coverage that will not exist; the handbook's general hotel parameter is 20 years for income property ^[9].

The license holder and the operator are unclear. The study should name the company that will hold the state license, the franchise and the payroll, and show the license dates in the opening schedule ^[87].

Related Pages on Loan Analytics

• The Loan Analytics methodology series (all articles)
• How a feasibility study is done: the Loan Analytics methodology
• SBA feasibility studies (7(a) and 504)
• USDA feasibility studies
• Bankable feasibility studies for conventional loans
• Hotel and motel: loan data and feasibility studies
• Loan data by state
• Loan data by asset class
• Worked examples (case studies)
• Worked Example: USDA Business and Industry Hotel Feasibility Study in Connecticut
• How to determine demand for a new hotel development
• Hospitality Feasibility: Where ADR Growth Can’t Outrun New Supply

Commission a Feasibility Study

Loan Analytics prepares lender-grade feasibility studies for SBA 7(a), SBA 504, USDA Business and Industry, conventional and EB-5 financing, built on the methods this series sets out. A study is not an appraisal: it tests whether a project's market, costs and cash flow can carry the debt a lender is asked to approve. To discuss a study for your project, contact Loan Analytics.

Daniel Smith, MAI, Loan Analytics