Model Study: Acquiring and Repositioning a 72-Room Boutique Hotel in Portland, Oregon, with SBA 7(a)

Model study: an illustrative feasibility analysis built on public data for Multnomah County, Oregon; not a client engagement.

Project summary

All figures in this table are model results or assumptions, not market facts.

The credit question

A lender reading a file like this is not asking whether Portland has visitors. Multnomah County drew $4,251.1 million of direct travel spending in 2025, according to the state tourism agency's annual travel impact report [3]. The question is whether a buyer can pay $95,000 a room for an independent hotel in a county where visitor spending on accommodations is still below its 2019 level in nominal dollars [3], spend another $26,400 a room renovating while guests are in the building, and cover a capital stack shaped by SBA's rules. Four things have to be true.

First, the market has to hold at roughly its 2025 level and recover slowly from there. Hotel, motel and short-term rental person-nights in the county fell from 7,374,390 in 2023 to 6,973,330 in 2025 [3]. Downtown hotel occupancy is published only in paid data, which this model study does not use, so the model assumes market occupancy of 62.0% in 2025 and 2026, rising slowly after that. The modelled hotel needs 62.7% occupancy at its Year 3 rate to reach 1.25x coverage. The renovated hotel must win at least its fair share, and the market cannot slip.

Second, the renovation has to buy a rate premium. The seller's trailing year is modelled at 61.0% occupancy and $158 ADR. The plan raises ADR to $181 and occupancy to 66.0% by Year 3. If none of that premium arrives, coverage falls below 1.00x from Year 3 (scenario 1 in the stress tests).

Third, the structure has to fit the rules. The $5,000,000 maximum for a 7(a) loan [1], not the 10% minimum injection [2], sets the buyer's cash. Historical coverage of 1.09x is below the 1.25x SBA requires for an Initial Acquisition [2], so the file depends on the Special Purpose Property route, which lets a lender rely on projections when the appraised value fully collateralizes the loan [2].

Fourth, the building has to be what the budget assumes. Portland requires a whole-building seismic upgrade of an unreinforced masonry bearing wall building of two or more stories once permitted alteration or repair costs within two years exceed $30 per square foot, a base figure the code adjusts each year for construction costs [4]. The modelled scope is $28.29 per square foot.

Location and county economy

Multnomah County contains Portland, the Oregon Convention Center and Portland International Airport, and anchors the Portland-Vancouver-Hillsboro metro area.

County indicators

Population and income. The county lost about 21,000 residents between 2020 and 2025 and has been nearly flat since 2023, adding 694 people in 2024 and 919 in 2025 [5]. That is a mild negative for the bar and cafe and a reason food and beverage is held at $14 per occupied room in the model. Median household income rose to $90,423 in 2024 [6] from $72,469 in 2019 [7]. Income is not the constraint here; downtown activity is.

Employment. Travel is a modest share of the county's job base. The state's travel impact report counts 21,840 direct travel jobs in 2025, 3.3% of the county's 654,140 jobs, and 8.7% fewer travel jobs than the 23,920 of 2019 [3]. Travel-supported jobs in accommodation and food service were 12,480 in 2025, against 14,000 in 2019 [3]. That gap is consistent with lower lodging demand than in 2019 and with properties running leaner staffing. The model grows rooms labor cost 3.5% a year, an assumption.

Access. Portland International Airport handled 18,563,132 passengers in 2025, up 6.0% from 2024, while international passengers fell 7.6% to 769,491 [8]. Domestic air access is improving; international inbound is not yet a driver for an independent hotel.

Market analysis

Demand drivers

Visitor spending. The state tourism agency's 2025 travel impact report puts Multnomah County direct travel spending at $4,251.1 million and travel employment at 21,840 [3]. Visitor spending by hotel and motel guests in the county grew from $1,106.1 million in 2021 to $1,789.1 million in 2025, a 3.7% gain in the last year [3]. Statewide, travel spending rose 1.3% to $14.6 billion [9], and the Portland Region accounted for $5,549 million, up 0.2% [3].

Multnomah County travel indicators, 2019 to 2025

Source: [3]. Nominal dollars; STVR is short-term vacation rental. The report notes that county estimates are subject to revision [3].

Spending on accommodations in 2025 was still 10.7% below 2019 in nominal dollars, and hotel, motel and rental person-nights fell 5.4% between 2023 and 2025 [3]. Rate, not volume, has carried the recovery. That argues against assuming a large occupancy gain and for rate gains only where a product change justifies them.

Where visitors spend. In 2025 visitor spending in the county on arts, entertainment and recreation rose 13.1% and on food service 6.5%, while visitor air transportation spending fell 1.8% [3]. Leisure activity is growing faster than travel overall, which suits a boutique hotel with a lobby bar more than a hotel built around weekday corporate demand.

County lodging tax budget. Multnomah County's Convention Center Fund receives a portion of the 11.5% transient lodging tax on hotel and motel rooms in the county and a 2.5% increment of the motor vehicle rental tax [10]. Its budgeted tax revenue rose from $41.43 million in fiscal 2024 to $44.28 million in fiscal 2026, and the fiscal 2027 adopted budget sets it at $43.46 million [11], [12], [10].

Convention Center Fund tax revenue, adopted budgets

Budgets are forecasts, not collections, but the county does not expect lodging-related tax revenue to grow in fiscal 2027.

The convention center. The county reports that Oregon Convention Center visitors generated $358 million of economic impact in fiscal 2025 and targets $378 million in fiscal 2027 [10]. A 72-room independent hotel takes overflow and small groups, not convention blocks, and the model does not depend on them.

Existing supply and its condition

Public sources do not publish a current count of Portland's competitive hotel rooms, their occupancy or their average rate. Those series come from paid data that this model study does not use, so the model treats market occupancy as an assumption and does not count the competitive supply.

Three points follow. The county data point to a demand problem rather than a supply problem: person-nights fell 5.4% from 2023 to 2025 while spending by hotel, motel and rental guests rose 5.2% [3], so the model assumes no net new competitive supply through 2030. If competing hotels trade at distressed prices, an appraiser may cite those sales as a lower bound on value. And competitors under pressure may cut rate, so the stress tests include a case with no renovation premium.

A paid competitive-set report with the subject's own penetration history is the key document missing from any public-data study; a lender would require one.

Demand method: fair share and penetration

For an existing hotel the method is fair share: the subject's share of room-nights sold against its share of supply. Without a public supply count, the model works with occupancy: the penetration index is the subject's occupancy divided by the market's. An index above 100% means the hotel takes more than its share.

Market and subject occupancy, 2025 to 2030

Both occupancy columns are assumptions or model results, not measurements of the competitive market.

The penetration path is deliberately modest: 94% in the renovation year, then 101% to 103%, a small premium for a renovated product. The plan assumes the market recovers 3.0 points over five years and the hotel moves with it.

Rate positioning: independent versus branded

The seller's trailing ADR of $158 is an assumption standing in for the seller's actual statements. The plan takes ADR to $181 by Year 3, a 14.6% increase over three years, or about 4.6% a year. That requires the renovated hotel to sell as an upgraded product. With no public benchmark for downtown rates, this is the assumption a lender should test hardest.

Staying independent avoids franchise royalty, program and reservation fees but means higher online travel agency commissions and more of the hotel's own marketing: 6.0% of rooms revenue and 6.5% of total revenue after stabilization. The comparison below holds rate and occupancy constant.

Independent versus soft-brand affiliation, Year 3 (same ADR and occupancy)

Soft-brand fees, savings and capital are illustrative assumptions, not quotes from any franchisor.

Under these assumptions a soft brand would cost $133,692 of NOI in Year 3 and cut coverage to 1.18x unless it lifted ADR by $7.71 per occupied room, about 4.3%, plus $432,000 of brand-standard capital. The table does not credit a brand's loyalty-driven occupancy. The independent plan is better only if the owner runs direct booking well; scenario 5 tests the failure case.

For SBA, a lender must review any third-party management agreement for passive-business risk unless it is part of the franchise disclosure documents of a brand listed on SBA's Franchise Directory [2]. An owner-run independent hotel avoids the question.

Forecast

Year 1 runs at 58.8% occupancy and $164 ADR with rooms out of service, Year 2 at 64.0% and $174, Year 3 at 66.0% and $181, Year 4 at 67.0% and $186, then flat occupancy with ADR up 3.0% a year. The plan does not assume a return to pre-2020 demand.

Site and technical

Site requirements. For an existing hotel the site test is legal status, condition and location. A zoning confirmation letter should show the hotel is a legal use and could be rebuilt after a casualty. Walkable access to downtown, the convention center and light rail matters; parking is limited and sits in a small other-revenue line of $6 per occupied room.

Zoning and permits. A rooms renovation with no change of use or added floor area needs building and trade permits rather than a land use review; the model assumes the hotel is a legal conforming use, to be confirmed. Two code questions decide the scope.

Seismic. Portland City Code chapter 24.85 sets seismic triggers for existing buildings [4]. For unreinforced masonry bearing wall buildings of two or more stories, permitted alteration or repair costs above $30 per square foot within a two-year period require the whole building to be improved to the ASCE 41 basic performance objective for existing buildings [4]. Site improvements, mandated accessibility improvements and some other mandated work are excluded from that cost, and the dollar amounts are adjusted each year by a construction cost index for Portland [4]. At the assumed 42,000 square feet, the base threshold is $1,260,000. The modelled building scope, including its contingency, is $1,188,000, leaving $72,000 of headroom, or about 6%, against the base figure; the indexed figure is higher, so the test is conservative. A 20% overrun would cross the base figure. From January 1, 2029, an alteration permit valued above $175,000, excluding mechanical, electrical, plumbing and finish work among other items, will require an ASCE 41 seismic evaluation, with some building types exempt [4]. The model assumes a concrete or steel structure, to be confirmed by a seismic engineer before closing.

Lodging taxes. Guests in the City of Portland pay a 6% city transient lodgings tax and a 5.5% county tax, which together make the 11.5% total the city's Revenue Division collects [13], [14]. The Portland Tourism Improvement District includes all hotels in the city and adds a 3% assessment on rent for stays after June 30, 2021, with stays of more than 30 days excluded [15]. Guests may be charged the 3% only when it is itemized separately on the bill as the Portland Tourism Assessment [13]. The state lodging tax is 1.5% for nights in 2026 and 2.75% for nights beginning on or after January 1, 2027; the added 1.25% is a nature conservation fee that receipts must list separately [16]. A Year 1 guest therefore pays 17.25% on top of the room rate, up from 16.0% in 2026; a $164 room costs $192.29. The taxes pass through and sit outside modelled revenue, but they raise the price guests compare with hotels outside the city.

Construction approach. The works replace beds, furniture, flooring, lighting and televisions and upgrade bathrooms, room heating and cooling units, corridors and life-safety systems: $15,000 a room of building work and $9,000 a room of furniture and equipment, each with a 10% contingency. Construction cost inflation is the main budget risk, which is why the stress tests include a 20% overrun.

When the construction component of a 7(a) loan exceeds $350,000, SBA requires, before construction starts, evidence that the contractor has furnished a 100% performance bond and a labor and materials payment bond and carries builder's risk and workers' compensation insurance. SBA waives that requirement when the lender retains a third-party construction management firm with funds control, or uses its own construction management department [2]. The modelled building work alone exceeds $350,000, so the test applies whatever the lender decides about furniture. For a phased renovation, a third-party monitor is the practical route.

Schedule. Closing is assumed in January 2027, with four 35-night phases of 18 rooms from January through May, the low season.

Renovation phasing and displacement, Year 1

Model assumptions: 52% winter occupancy on the rooms taken out of service, and 35% of displaced guests moved to available rooms in the hotel rather than lost.

The 2,520 out-of-order room-nights are 9.6% of inventory, but only 852 paid room-nights are lost, about $139,689 of rooms revenue. A full closure would lose far more and would interrupt the review history and online ranking an independent hotel depends on.

Development budget and sources and uses

The budget combines an assumed $6,840,000 acquisition with renovation, soft costs, the SBA fee and reserves. The price allocation matters because SBA treats real estate differently for maturity.

Development budget

All amounts are model assumptions with the rationale given in the assumptions table. The SBA upfront fee is computed from SOP 50 10 8.1's maximum fee chart [2].

Real estate is $7,288,000, 75.4% of cost; that split decides the loan's maturity.

Sources of funds

The SBA minimum equity injection for an Initial Acquisition is 10%, measured on total project cost, and cannot be reduced or eliminated [2]. Here that is $966,893. Seller debt that is subordinated and on full standby may count as equity, but limited sources such as standby debt can provide no more than half of the required injection [2]. Seller note A is set at exactly that half, $483,446. The buyer's cash of $3,585,479 is 37.1% of cost, far above the minimum. The binding constraint is the $5,000,000 7(a) maximum [1]. Counting seller note A, equity is $4,068,925, 42.1% of cost.

Revenue build and operating expenses

Revenue

Food and beverage (lobby bar and cafe) and other revenue are modelled per occupied room, growing 3% a year.

Revenue build, Year 0 to Year 5

All rows are model results.

Year 1 RevPAR is flat at $96.36 because rooms are out of service. The step to Year 2, a 15.6% RevPAR gain on rooms selling at $174 instead of $158, is the assumption a lender should press hardest, in a market where large competitors are under pressure and public data show nights falling.

Operating expenses

Costs are split into those that vary with occupied rooms (housekeeping, linen, supplies), those that vary with revenue (commissions, marketing, management, reserve) and largely fixed costs (administration, maintenance, utilities, systems, tax, insurance). The fixed block makes margins move sharply with RevPAR.

Operating expense structure

All rows are model results built on the assumptions table.

Rooms cost runs $40.00 per occupied room in Year 0 and $44.35 by Year 3. EBITDA rises from 24.5% of revenue to 29.3%, moderate for a boutique hotel with no restaurant and no brand fees but high labor and insurance costs. The FF&E reserve is 2% in Year 1, 3% in Year 2 and 4% thereafter.

Management is imputed at 3% for an owner-operator so unpaid owner labor is not counted as profit. Property tax ($96,000) and insurance ($1,450 a room) are assumptions; Oregon's assessed-value system can make the actual tax bill differ sharply from what the price suggests, so both should be replaced with actual bills for the tax lot and a current insurance quote that prices earthquake cover.

10-year pro forma

10-year pro forma ($ thousands, model)

Debt service is $640,363 a year: $560,428 on the 7(a) loan and $79,935 on seller note B. Year 1 is $13,792 short, funded from the reserve. Cash-on-cash return on buyer cash is 5.2% in Year 2, 7.0% in Year 3 and 11.7% in Year 10.

Debt structure under the 7(a) program

Debt terms

The $5 million cap. The maximum amount of any one 7(a) loan is $5,000,000, and the SBA portions of all loans to a borrower and its affiliates may not exceed a guaranty of $3,750,000 [1]. Loans over $150,000 may receive a guaranty of up to 75% [1], so a $5,000,000 loan uses the whole guaranty. The 7(a) loan funds only 51.7% of cost, so the buyer and the seller fund the rest. A larger deal would need a second loan, which SBA's rule against piggyback structures restricts: SBA does not allow its loan to sit in a junior or no-lien position behind another loan for the same purpose approved within 90 days [2].

Rate. Prime was 7.00% on November 9, 2026, and the 10-year Treasury yield was 4.88% the same day [17]. The model uses Prime + 2.75%, below the Prime + 3.0% maximum variable rate for Standard 7(a) loans above $350,000 [2]. Each 100 basis points of Prime costs about 0.08x of Year 3 coverage. Fixed rates are also possible; SBA periodically publishes the maximum allowable fixed rate [1].

Fees. The upfront fee on a $5,000,000 loan with a 75% guaranty is $138,125 under the SOP's maximum chart [2]. SBA may announce temporary fee changes for a fiscal year through an Information Notice, and the lender's annual service fee cannot be charged to the borrower [2]. The model keeps the maximum-chart fee as a conservative assumption.

Maturity blending. A 7(a) loan that finances a change of ownership may not have an amortization over 10 years, except as SBA allows for real estate and Special Purpose Properties [2]. When real estate is part of the purchase, the loan can be split into separate loans or blended on a weighted-average basis, rounded to the nearest full year, with only the real estate share allowed beyond 10 years (up to 25), calculated before any equity is applied [2]. Here, 75.4% of uses are real estate: 0.754 x 25 + 0.246 x 10 = 21.31 years, rounded to 21.

SBA also has a Special Purpose Property exemption that allows up to 25 years when 85% or more of total project costs are for real estate [2]. This hotel misses it at 75.4%. A 25-year term would cut debt service by $25,746 and lift Year 2 coverage from 1.29x to 1.35x.

Seller notes. Seller note A is on full standby: no principal or interest during the 7(a) term, subordinated, and documented on SBA Form 155 or the lender's equivalent standby agreement [2]. Standby debt may accrue interest and be amortized after the 7(a) loan is paid in full [2], which in the model means at the Year 10 sale. Seller note B pays monthly. Because it is not on standby, it does not count as equity and its payments are included in coverage; SBA also requires any non-standby acquisition debt structured as interest-only to be tested on an amortization of no more than 10 years [2]. Total debt, including seller debt not on full standby, is limited to the business valuation [2]; here it is $5,600,000 against a $6,840,000 price.

Seller role. In an Initial Acquisition the seller may not remain an officer, director, stockholder or employee, but the business may contract with the seller as a consultant for up to 24 months [2]. For an independent hotel, a transition agreement covering revenue management and website handover is worth having.

Why structure decides the file

All rows are model results.

The base structure uses seller note B to cut the buyer's cash by $600,000 but lowers historical coverage to 1.09x. Without seller note B, historical coverage is 1.246x on the full $5,000,000 loan, just short of 1.25x. A 7(a) loan sized to exactly 1.25x on the seller's EBITDA would be $4,982,992. A larger $1.2 million seller note fails the two-year projection test at 1.15x. A 10-year term, which is what a change of ownership without real estate would get, fails at 0.96x. The blended 21-year term is what makes this file work.

DSCR and break-even

Debt service coverage by year

All rows are model results.

SBA defines historical coverage as EBITDA divided by the combined debt service after the transaction [2]; the seller's year covers the new debt 1.09x, below the 1.25x required for an Initial Acquisition [2]. For an owner-occupied Special Purpose Property that is in operation, a lender may rely on projections when the appraised value fully collateralizes the loan, provided projected coverage reaches the required level within two years of funding, and the historical coverage and the variance between historical and projected results go in the credit memorandum [2]. SBA names hotels where the real estate fully secures the loan as an example [2]. On EBITDA, the projections reach 1.45x in Year 2. On the stricter lender view, after a 4% FF&E reserve, they reach 1.29x in Year 2 and 1.46x by Year 4.

Two readings remain open. SBA allows projections where the appraised value fully collateralizes the loan, and also where the purchase price is fully supported by the appraised value of the property [2]. Here the $6,840,000 price includes $740,000 of furniture, equipment and goodwill, and SBA real estate appraisals may not include the value of intangible assets [2]. Whether the test is read on the appraised value or on the 85% value used for collateral, and how the lender treats the non-real estate part of the price, are questions for the lender and SBA. The fallback structure in the model determination covers the case where the projection route is not available.

A careful lender works from the after-reserve column, where Year 2 clears 1.25x by only 0.04x.

Break-even, Year 3 cost structure

All rows are model results.

At the Year 3 ADR, the hotel breaks even on debt service at 56.9% occupancy and reaches 1.25x at 62.7%. The model's assumed 2025 market occupancy is 62.0%. The renovated hotel must at least match the market to meet SBA's standard, with about 9 points of occupancy above the 1.00x break-even. At 66.0% occupancy, ADR can fall to $163.06 before coverage reaches 1.00x and to $174.53 before it drops below 1.25x.

Returns

Returns

All rows are model results.

The exit assumes a sale at the end of Year 10 at an 8.25% capitalization rate on Year 11 NOI, with 3% selling costs. That produces $182,044 a room, almost twice the $95,000 purchase price. That gain is the buyer's reward for downtown recovery risk, not collateral a lender can rely on. A 13.7% levered IRR is modest for this risk, and operating cash does not repay the buyer's cash within ten years; the sale does.

Returns are not very sensitive to the exit cap: at 9.50% the exit value is $11,382,551 and the levered IRR 11.9%. They are far more sensitive to operating performance.

Sensitivity and stress tests

Year 3 DSCR after reserve: occupancy vs Prime rate (7(a) at Prime + 2.75%)

Model results. Occupancy shifts are applied from Year 2 onward. Seller note B is fixed rate; only the 7(a) loan moves with Prime.

The grid shows where the file breaks. A three-point occupancy shortfall still clears 1.25x at today's Prime. A six-point shortfall, which would leave the hotel at 60.0%, below the assumed 2025 market, drops coverage to 1.13x. At planned occupancy, a 5% ADR shortfall gives 1.19x and a 10% shortfall 1.00x. The reserve can absorb one miss for a year, not a sustained one.

Downside scenarios (DSCR after reserve)

All rows are model results.

Scenario 1, recovery stalls. If occupancy stays at the seller's 61.0% and ADR grows only 2% a year, the renovation earns no return. Coverage falls to 0.97x by Year 3 and stays there, and the levered IRR falls to zero. It matches the 2023 to 2025 decline in county hotel, motel and rental person-nights [3] and the county's flat lodging tax budget [10], so a lender should weigh it most.

Scenario 2, renovation overrun. A 20% overrun on renovation work plus phases running 50% longer adds $345,600 of cost, of which $328,320 comes from the buyer's cash and the rest from a larger standby note, and cuts Year 1 coverage to 0.91x. Coverage recovers from Year 2. The real risk is seismic: a 20% overrun on building work takes the scope to $1,404,000, above the $1,260,000 threshold computed at the code's base figure for a masonry building [4].

Scenario 3, rates. Prime 200 basis points higher from Year 2 lowers Year 2 coverage to 1.15x and Year 3 to 1.24x. The two-year test would fail.

Scenario 4, a demand shock in Year 2. A convention or air-travel slump that cuts Year 2 occupancy to 59.0% and ADR by 4% drops Year 2 coverage to 0.94x. The hotel recovers in Year 3, but the two-year test falls in Year 2.

Scenario 5, distribution. If the hotel cannot shift guests to direct booking and commissions stay at 9% of rooms revenue while marketing rises to 7.5% of revenue, Year 2 coverage is 1.10x, Year 4 is 1.245x, and the hotel reaches 1.25x only in Year 5.

Scenario 6, combined. A stalled market plus higher rates takes coverage to 0.86x.

The base case passes with little margin; five of the six scenarios miss 1.25x in Year 2.

Program compliance

SBA 7(a) eligibility and requirements checklist

On collateral, a $6,100,000 real estate appraisal (an assumption) gives a discounted value of $5,185,000; with existing furniture and fixtures at 10%, the total is $5,223,000, fully securing the loan by $223,000. An appraisal 5% lower leaves a $36,250 shortfall. In that case SBA requires the lender to take available equity in the personal real estate of the owners of 20% or more [2], and the projection route comes into question.

Risks and mitigants

Risk register

Assumptions

Key assumptions

Model determination

As structured, the modelled acquisition and renovation is financeable under SBA 7(a), but only on the Special Purpose Property projection route and with little room for error. It reaches 1.29x after reserves in Year 2 and 1.46x by Year 4, while historical coverage of 1.09x misses the Initial Acquisition standard, and a stalled recovery, a 200 basis point rate rise or a failed direct-booking plan would each break the two-year test.

The $5,000,000 cap forces the buyer to bring 37.1% of cost in cash even with two seller notes, which also keeps the lender's exposure moderate against the real estate. If the appraisal lands where the model assumes, the loan is fully secured, the strongest part of the credit.

A lender would likely set these conditions:

1. A written determination in the credit memorandum that the hotel is Special Purpose Property, that the real estate is integral to the business, and why projections are appropriate, with historical and projected coverage both entered in SBA's loan system [2].
2. An as-is real estate appraisal and a business valuation, both prepared for the lender, supporting a fully secured loan and the purchase price [2]. If either falls short, the fallback removes seller note B and trims the 7(a) loan to $4,982,992, which meets 1.25x on history; the buyer then contributes $4,202,033, 43.5% of cost.
3. A seismic engineer's report confirming the building is not unreinforced masonry, before the renovation contract is signed [4].
4. The $300,000 reserve held until coverage after reserves exceeds 1.25x for four quarters.
5. Third-party construction monitoring with funds control [2] and fixed-price phases.
6. A paid competitive-set report, seller statements verified against IRS tax transcripts [2], and proof that over half of revenue is from stays of 30 days or less [2].
7. A 4% FF&E reserve covenant from Year 3 and a priced rate cap or fixed-rate option.

If the buyer cannot meet those conditions, the simpler answer is a smaller purchase price. Each $100,000 off the price reduces the buyer's cash by about $95,000, because the 7(a) loan is already at its cap and only the standby note scales with cost.