Worked Example: Conventional Community Bank Construction-to-Permanent Short-Term Rental Feasibility Study in New York

This is a worked example, not a client engagement. It takes a hypothetical cluster of six new two-bedroom cabins on one wooded parcel in Ulster County, in the Catskills tourism region of New York, owned by an investor entity that hires a full-service manager, and shows how a feasibility consultant would test it against a community bank construction-to-permanent loan: why the deal is conventional rather than SBA, which sizing test binds, how thin the coverage is at the assumed nightly rate and occupancy, and what would decide the answer. Every project number is a stated assumption, every market, tax, regulatory and rate figure is cited to a public source, and no town, business, lender or address is named.

At a glance. Six new cabins in Ulster County on a community bank loan that funds construction and converts to a five-year fixed rate. Coverage, not the assumed 70 percent loan-to-cost limit, sizes the debt: at the assumed $265 nightly rate and assumed 52 percent occupancy the cabins support an assumed loan of $960,000 at 1.26x coverage, 39.3% of cost, with the owner bringing 60.7% as equity (Tables 4 and 8). At 70 percent of cost coverage would be 0.71x (Table 8), and every sensitivity in Table 9 falls below the assumed 1.25x minimum, in a range from 0.72x to 1.06x.

• Total project cost: $2,441,000
• Loan (debt service coverage test binds): $960,000
• Equity: $1,481,000
• Stabilized net operating income after reserve: $106,134
• Debt service coverage at stabilization: 1.26x
• Break-even occupancy at 1.00x: 47.0%
• Coverage if the bank lent 70 percent of cost: 0.71x

What a Feasibility Consultant Tests First on a New York Short-Term Rental Construction Loan

A community bank lends on a ground-up cabin cluster under a written real estate lending policy, and the interagency real estate lending guidelines, Appendix A to Part 365 of the FDIC's regulations, say what it should contain. For development and construction projects it should require feasibility studies and sensitivity analyses of income to interest rates, vacancy rates and operating expenses ^[1], and set minimum requirements for the borrower's hard equity and minimum standards for cash flow and debt service coverage ^[2]. The coverage minimum itself is the bank's own number, not a regulatory one, and that paragraph is the study's table of contents: income, coverage, equity, and the three sensitivities the rule names.

The consultant's first question is whether stabilized net operating income covers the permanent debt service at the bank's minimum. The result is close to the line. Stabilized net operating income after reserve is $106,134 and annual debt service is $84,309, so coverage at stabilization is 1.26x against an assumed 1.25x lender minimum (Table 8). With no program floor in a conventional loan, the lender's assumed 1.25x minimum binds (Table 8). A loan sized to the assumed 70 percent loan-to-cost limit would be $1,708,700 and would cover only 0.71x (Table 8). The project cannot borrow what the cost test would allow.

The second question is the rate. A construction-to-permanent loan at a community bank commonly fixes the permanent rate for five years and then resets it, so the model takes the 5-year Treasury constant maturity yield, 4.64 percent on November 6, 2026, from the Federal Reserve's H.15 release ^[3] and adds an assumed 2.75 point spread, giving 7.39% on an assumed 25-year amortization, or $8.78 of annual debt service per $100 of loan (Table 8). The 10-year Treasury in the same release, 4.86 percent ^[4], is not used because the bank is not fixing for ten years; the sensitivity section tests a two-point rise at the reset instead.

Credit conditions are the third thing to weigh. The October 2026 edition of the Fed's loan officer survey recorded more banks stiffening than easing their terms in each commercial property category, construction and land development among them ^[5]. A study showing coverage one hundredth above the minimum should expect to be asked what happens when the first winter is slow.

Decision. The study would size the loan on debt service coverage, not on loan-to-cost, because coverage binds first (Table 8), and would present the achieved nightly rate and occupancy as the inputs that decide whether the deal clears the bank's minimum at all.

The Site and Its Market: Ulster County and the Catskills Region of New York

The Catskills is a visitor economy, and the public record says so in dollars. Visitor spending in the Catskills tourism region rose from $1,181 million in 2020 to $2,628 million in 2024, up 7.1 percent in 2024 alone, in the Economic Impact of Visitors to New York 2024 study that Tourism Economics prepared for Empire State Development ^[6]. All eleven New York tourism regions grew in 2024, and the increases were largest in New York City at 7.9 percent, followed by the Catskills at 7.1 percent ^[7]. Lodging was 51 percent of regional visitor spending, and seasonal second homes generated a further $520 million of economic activity ^[8]. The same study says tourism encompasses 16 percent of all jobs in the Catskills region ^[9]; the inference, which is the analyst's and not the source's, is that a cabin's demand rests on visitors because the regional economy does.

Within the region, Ulster County is the largest market and the fastest growing. Ulster County represented 43 percent of the region's tourism sales in 2024, with $1.1 billion in direct tourism spending, followed by Sullivan County at 39 percent ^[10], and Ulster led county spending growth at 8.8 percent ^[11]. Visitors spent $625.7 million on lodging in Ulster County in 2024 ^[12]. The Ulster County Comptroller reports growing tourism visitation and higher than ever occupancy tax collections ^[13], with receipts up from $2 million in 2019 to $8.3 million in 2024 after the rate doubled ^[14].

The resident base is small and nearly flat, which is why this example builds no local demand case: Census Bureau Vintage 2025 estimates put Ulster County at 183,330 residents on July 1, 2025, about 0.8 percent above its April 1, 2020 base ^[15]. The market that matters is two hours south, the New York-Newark-Jersey City metropolitan statistical area of 20,112,448 residents ^[16].

The ownership of the existing stock points the same way. Granicus (Host Compliance) listing data reviewed by the Ulster County Comptroller show 1,008 Ulster County short-term rentals with owners who receive mail outside the county ^[21]; in the same Granicus data the Ulster County Comptroller puts that share at 36 percent, most of them with New York City addresses ^[22]. The site itself is only what the model needs: about 15 wooded acres, assumed to cost $210,000 (Table 3), with its own wells and septic, in a municipality whose zoning allows a non-owner-occupied cluster; no Ulster County land price was sourced.

Decision. The study would present Ulster County as the leading and fastest-growing lodging market in a region whose economy is one sixth tourism, would source demand to the New York metro rather than to local residents, and would flag the land price, the parcel's tax rate and the municipality's rules as site facts this example could not supply.

Demand: Who Books a Catskills Cabin and at What Rate

A hotel feasibility study starts from a market's published occupancy and average daily rate. A short-term rental study in the Catskills cannot, because this example found no public county-level rate or occupancy series for Ulster County or the region; the licensed data is published at town level, and the case names no town. The assumed rate and occupancy are therefore anchored to national figures and the distance is stated.

AirDNA reports national short-term rental occupancy of 53.6 percent in September 2026, up 3.7 percentage points year over year, and an average daily rate of $254, up 7.2 percent ^[23]. In its July 2026 midyear outlook AirDNA forecasts national occupancy of 57.4 percent in 2026 and 57.5 percent in 2027, just above the pre-COVID average ^[24], and expects smaller cities, rural markets and regional drive-to destinations to outperform as travelers prioritize value ^[25], which describes a Catskills cabin two hours from the metro. Against that, AirDNA reports that mountain and lake resort demand grew only 0.6 percent year over year in August and September 2026 ^[26], while mountain and lake resort average daily rates grew 5.1 percent, behind only coastal resorts at 5.4 percent ^[27]: rate is doing the work in resort markets, demand is not. Two further AirDNA observations bear on a two-bedroom new build: AirDNA reports that through the first five months of 2026 studio to two-bedroom listings were roughly flat while four- and five-plus-bedroom homes led on occupancy and rate ^[28], and that new listings, lacking reviews and repeat guests, have continued to compete on price ^[29]. The inference, which is the analyst's, is that a new cabin should expect to open below established hosts' rates, and the ramp table is built on it.

The model's assumptions sit against those anchors as follows. The assumed average daily rate is $265 a night before cleaning fee and taxes, about 4 percent above the September 2026 national figure of $254 that AirDNA reports ^[23], on the reasoning that a new cabin with a hot tub in a resort region should price above the average listing. The assumed stabilized occupancy is 52 percent, below both the 53.6 percent September 2026 reading and the 57.4 percent forecast for 2026 that AirDNA publishes ^{[23] [24]}, for a mountain market with a winter trough in the segment AirDNA reports as flat ^[28]. The assumed average stay is 2.5 nights, for weekend-led drive-to demand (Table 3). None of these is a Catskills number; they are national figures adjusted by judgment, and the study would replace them with a monthly survey of comparable cabins in the county before a lender relied on them. No New York lodging rate series that is public and free to republish could be sourced for this example; the closest public measure is the Tax Department's open data, in which Ulster County traveler accommodation vendors reported taxable sales and purchases of $98,997,672 for June to August 2026 against $93,418,822 a year earlier, both preliminary ^{[30] [31] [32]}. Those are receipts, which the Department cautions may be only a portion of an industry's gross sales ^[33], not a cabin rate.

Decision. The study would carry the $265 rate and 52 percent occupancy as assumptions anchored to AirDNA's national figures and would state that no regional or county series was found, so the lender reads the revenue build as a hypothesis rather than a market fact.

Competition: Ulster County Supply and the Local Rules That Cap It

The competitive set is the county's existing stock, measured by the Ulster County Comptroller with two data vendors, and the rules that decide where new stock can be added.

AirDNA data reviewed by the Ulster County Comptroller show the monthly average number of short-term rentals in the county rising from 942 in 2015 to 3,013 in 2024 ^[34]. The Granicus (Host Compliance) count of 2,827 in May 2025 that the Ulster County Comptroller also reports ^[36] comes from a different vendor and method, and the study would not read it as a decline from the AirDNA peak; each measures a county stock of roughly three thousand units. Within that stock, the AirDNA data reviewed by the Ulster County Comptroller show 98 percent of Ulster County short-term rentals are entire homes ^[37], so a cabin competes with houses, including the larger houses that AirDNA reports as leading on occupancy and rate ^[28].

The rules are the sharper constraint. The Ulster County Comptroller reports that at least 16 municipalities in the county have adopted some form of short-term rental regulation, and that at least 6 of those have adopted either a cap or a requirement that the unit be owner occupied ^[38]. The county's Housing Smart Communities Initiative recommends owner occupancy or a cap at 2 percent of the housing stock ^[39], and 15 of the county's 24 municipalities were above that mark ^[40]. An owner-occupancy requirement bars this project outright; a cap bars it wherever the cap is full; and a county, city, town or village may enact a local law prohibiting or further limiting short-term rental units ^[41]. The inference, which is the analyst's, is that site selection is the project's first feasibility test, and the study would document the municipality's law rather than assume it.

National supply is slowing, which helps a project that gets built. AirDNA counts 1.74 million available national short-term rental listings in September 2026, up 1.6 percent year over year ^[42], the slowest listing growth since the pandemic as higher mortgage rates curb new supply ^[43], and AirDNA expects available listings to grow just 2.7 percent in 2026 ^[44]. The exception is the segment this project sits in: AirDNA reports that small town and rural markets still have the highest listing growth rate, 5.3 percent year over year for August and September 2026 ^[45]. Rural cabins are where the new competition is being built.

New York City, where so many Ulster County owners live, is the contrast case: Local Law 18, adopted January 9, 2022, requires hosts to register and bars booking platforms from processing transactions for unregistered rentals ^[46], with enforcement from September 5, 2023 ^[47]. AirDNA counts 215,543 active Airbnb, Vrbo and Booking.com listings across New York State markets as of October 2026 ^[48] and shows the New York City market with 14,070 active listings, 77 percent occupancy and a $182 average daily rate ^[49]; nothing sourced here says what Local Law 18 did to that market.

Decision. The study would treat the municipality's short-term rental law as a go or no-go condition ahead of any market analysis, and would describe the competitive stock as roughly three thousand entire-home listings on either vendor's count.

Cost Basis: What Six Cabins Cost to Build in Ulster County

Rounded to the dollar; lines may not foot by $1.

Every input that is not a tax rate or a Treasury yield is an assumption; the cost lines are the ones a contractor's bid would replace first.

The total project cost is $2,441,000, or $406,833 per cabin (Table 4). The largest line is cabin construction at an assumed $215,000 per cabin, about $307 per square foot on a 700 square foot plan, for $1,290,000 across the six (Table 4). That assumption is set against the only public construction cost series at county level: the reported construction valuation of single-family permits in the Census Bureau's Building Permits Survey, about $405,600 per unit in Ulster County in 2025 ^[52] and about $406,300 in 2024 ^[53], against about $305,100 in Greene County ^[54] and $335,900 in Sullivan County in 2025 ^[55]. A 700 square foot cabin is a fraction of a typical permitted house, so the assumed per-cabin figure sits well below the per-unit valuation, but per square foot it is not cheap, and the study would get bids. Site work and utilities are assumed at $330,000, and furnishings, a hot tub and equipment at an assumed $35,000 per cabin, including the fire extinguisher and posted safety plan the registry conditions would require (Table 3).

The soft and carrying lines are where a lender reads the sponsor's realism. Design, engineering, permits and approvals are assumed at $120,000, including the local site plan or special use approval (Table 3). Contingency is assumed at 5 percent of site work and cabins, $81,000 (Table 4), thin for a rural new build and a line a bank would likely ask to see raised. The interest reserve during construction is assumed at $45,000; an operating reserve for the ramp years is assumed at $50,000, sized with the working capital line to cover the year 1 and year 2 shortfalls in the ramp table; pre-opening, marketing and working capital are assumed at $35,000; and third-party reports, closing and financing costs are assumed at $70,000 as one combined line (Table 3).

The sources of funds are the striking part of the table. The bank loan is an assumed $960,000, 39.3% of cost, and owner equity is $1,481,000, 60.7% of cost (Table 4). That split is the arithmetic of coverage, explained in the debt service section, not a lender's loan-to-cost policy, and because the loan is fixed by income, every dollar of construction cost above the assumptions is a dollar of additional equity, not of additional debt.

Decision. The study would carry the per-cabin construction cost as the assumption most in need of a bid, and would show the sponsor that cost overruns fall entirely on equity at this loan size.

Revenue Build: Rate, Occupancy, Cleaning Fees and the Taxes the Guest Pays

Six cabins offer 2,190 available cabin-nights a year, and at the assumed 52 percent occupancy 1,139 are booked; at the assumed $265 rate they produce $301,782 of nightly revenue, and the assumed 2.5-night average stay makes 456 stays (Table 5). Each stay carries an assumed cleaning fee of $135, for $61,495, and other guest fees are assumed at 1.5 percent of nightly revenue, $4,527 (Table 5). Total revenue at stabilization is $367,804, $61,301 per cabin and $137.80 of nightly revenue per available cabin-night (Table 5). The operating statement charges an assumed $120 of housekeeping cost per turnover against the cleaning fee.

What the guest pays is more than what the owner books, because the tax rules changed in 2025. Effective March 1, 2025, New York State and local sales tax is imposed on sales of short-term rental unit occupancy when the rental rate is more than $2.00 per unit per day ^[56]; the Tax Department attributes that change to Chapter 99 of the Laws of 2025 ^[57]. Its definition of a short-term rental unit includes a cabin and a cottage ^[58], and cleaning fees, extra person charges and pet fees may be taxable as part of the charge for occupancy ^[59]. The state rate alone is 4 percent ^[60], and the combined state and local rate in Ulster County is 8 percent under Publication 718 effective March 1, 2025 ^[50]. Localities may charge an additional tax on short-term rental occupancy, administered by the locality ^[61], and Ulster County does: its occupancy tax rate rose to 4 percent effective February 1, 2024 ^[51]; neighboring counties mostly charge 5 percent ^[62]. The Comptroller states that this revenue is collected by businesses on behalf of the county and is not intended to be a cost borne by the operator but by the guest ^[63], which is why the model keeps it, and the sales tax, outside revenue and expenses. The memo line of Table 5 shows the result: at the assumed $265 rate, 8 percent sales tax and 4 percent county occupancy tax lift the guest's nightly cost to $296.80, before the cleaning fee and any tax on it (Table 5), the price the guest compares against hotels.

Who collects matters for the expense lines. A booking service that facilitates the sale of a short-term rental unit occupancy in New York State must register as a sales tax vendor and collect the state and local sales tax ^[64], and an operator must itself register and collect unless an exception applies ^[65]. The practical reading, which is the analyst's, is that direct bookings off platform make the owner a registered collector, and the assumed $9,000 a year for registration, accounting, software and licenses in Table 3 includes the filings. New York City's unit fee of $1.50 per unit per day applies only within the city ^[66].

The ramp table is the honest part of the revenue build. Year 1 occupancy is assumed at 40 percent and year 2 at 47 percent, against 52.0% at stabilization (Table 6). The rate is assumed at 92 percent of the stabilized figure in year 1 and 97 percent in year 2, giving $243.80 and $257.05 against $265.00 (Table 6), on AirDNA's report that new listings without reviews compete on price ^[29]. Net operating income after reserve is $40,189 in year 1 and $78,327 in year 2 against $106,134 at stabilization, so against annual debt service of $84,309 coverage is 0.48x and then 0.93x, with cash shortfalls of $44,121 and $5,982 (Table 6). Those shortfalls are what the assumed $50,000 operating reserve and the working capital inside the assumed $35,000 pre-opening line are for, and a bank would want them funded at closing.

Decision. The study would keep both taxes outside revenue as pass-throughs borne by the guest, show the guest's all-in nightly cost of $296.80 beside the owner's assumed $265 rate, and present the ramp as two years in which the cabins do not cover their debt.

Debt Service Coverage, Loan Sizing and Sensitivity

The operating statement is a manager's statement, not an operator's. Total revenue of $367,804 bears booking channel and payment fees at an assumed 3.0 percent, $11,034, and a third-party management fee at an assumed 20 percent of nightly and other revenue, $61,262, the fee that makes the owner passive (Table 7). Housekeeping and laundry are $54,662 at the assumed $120 per turnover; utilities, repairs, supplies and insurance are assumed at $4,200, $4,500, $1,000 and $2,800 per cabin; and real property tax is an assumed $6,000 per cabin, $36,000 in total, on an assumed assessed value of $300,000 per cabin at about 2 percent, because the parcel's own tax rate was not sourced (Table 7). Total operating expenses are $246,958, a replacement reserve of 4 percent of revenue, $14,712, is deducted before coverage, and net operating income after reserve is $106,134, 28.9% of revenue and 4.35% of total project cost (Table 7). A yield on cost of 4.35% is the first number a lender would circle: it is below the 7.39% rate the bank would charge, which means the project cannot carry much debt by construction.

The sizing table runs two tests and takes the lesser. Test 1, the assumed 70 percent loan-to-cost limit, allows $1,708,700, for coverage of 0.71x; test 2, the loan that the assumed 1.25x minimum supports, is $966,804, and the assumed loan of $960,000 is that figure rounded down (Table 8). At the assumed $960,000 loan, annual debt service is $84,309, coverage is 1.26x against the assumed 1.25x minimum, and cash flow after debt service is $21,824 (Table 8). That cash flow is 1.5% of the owner's $1,481,000 of equity under the assumed structure (Table 8), before appreciation and income tax. On these assumptions the project is an equity-heavy bet on nightly rates, not a leveraged income investment.

The interagency guidelines name three sensitivities, interest rates, vacancy rates and operating expenses ^[1], and the sensitivity table runs them with two revenue cases added. Every one of them fails the assumed 1.25x minimum. Revenue down 10 percent, modelled as occupancy falling by a tenth, takes net operating income to $83,520 and coverage to 0.99x; revenue down by a fifth takes income to $60,907 and coverage to 0.72x (Table 9). A rate two points higher at the five-year reset lifts annual debt service to $99,770 and cuts coverage to 1.06x (Table 9). Operating expenses up 10 percent across every line cut income to $81,438 and coverage to 0.97x, and a nightly rate 10 percent lower at the same occupancy cuts income to $83,773 and coverage to 0.99x (Table 9). The range runs from 0.72x to 1.06x: a ten percent miss on either rate or occupancy, well inside the error of a national anchor applied to a county with no published series, takes the cabins below the point where they pay their debt at all. That is how thin this is.

The break-even table states the margin in a manager's terms. Fixed operating costs, the lines that do not move with bookings, are $120,000 a year, the sum of the assumed per-cabin and administration lines, and each booked cabin-night contributes $198.57 after the channel fee, the management fee, the reserve and the net of cleaning fee over cleaning cost (Table 10). At the assumed $265 rate the cabins cover fixed costs and debt at 1.00x at 47.0% occupancy and reach the assumed 1.25x minimum at 51.8%, against the assumed 52 percent base (Table 10). At the assumed 52 percent occupancy, the rate at which coverage falls to 1.00x is $239.14 and the rate that meets the assumed 1.25x minimum is $264.11, against the assumed $265 (Table 10). The cushion to 1.00x is real; the cushion to the covenant a bank would actually write is nil.

Decision. The study would report that coverage sets the loan at an assumed $960,000 (Table 8), that the deal meets the assumed 1.25x minimum with no margin and fails it under every tested shock (Table 9), and that the sponsor's choice is between bringing 60.7% of cost as equity (Table 4) or proving a county rate and occupancy above the assumptions before asking for more debt.

What the Conventional Program Requires and Why This Deal Is Not an SBA Loan

Under 13 CFR 120.110(c), passive businesses owned by developers and landlords that do not actively use or occupy the assets acquired or improved with the loan proceeds cannot borrow under the SBA's business loan programs unless they qualify as Eligible Passive Companies under section 120.111 ^[67]. As this example is structured, an investor entity owns the cabins, hires a full-service manager at an assumed 20 percent fee, and does not operate them; the owner neither uses nor occupies the assets, and the Eligible Passive Company exception covers a holding company leasing to the borrower's own operating business, not an investor leasing to the public through a manager. On the regulation's own test the venture is passive, and so it is conventional. The qualifier belongs in the same paragraph: the regulation's test is passive or active, not residential or commercial, and whether an owner who runs cabins as an operating lodging business could be treated differently is a different question that turns on SBA's operating procedures, which this example does not quote. It decides only that an investor who hires a manager is the passive case.

The lender type is a design choice: a community bank portfolio construction-to-permanent loan, because the project needs construction funds for six detached units and then a permanent loan sized on the cash flow of a managed lodging operation. Investor mortgage products built for completed one- to four-unit homes are not modelled.

Once the lender is a bank, the interagency guidelines bind. The supervisory loan-to-value limits are 80 percent for a commercial construction loan and 85 percent for a one- to four-family residential construction loan or an improved property loan ^[68]; improved property includes non-owner-occupied one- to four-family property and other completed income-producing property ^[69]. The assumed 70 percent loan-to-cost limit sits inside every one of them, and the coverage test binds long before any of them do (Table 8). The assumed 1.25x minimum is one bank's policy number inside the guidelines' instruction to set minimum coverage standards ^[2], and the 60.7% equity in Table 4 is the hard equity the same instruction would measure.

The second body of rules is New York's short-term rental law, which a bank would treat as a condition of the collateral's legality. The Assembly's bill record shows S.820 of 2025 signed as Chapter 99 on February 28, 2025 ^[70], after it passed the Senate on February 24, 2025 and, substituted for A.5686, the Assembly on February 25, 2025 ^[71]. The Ulster County Comptroller describes that chapter amendment as the final framework for implementing the 2024 registry law ^[72] and reports that it requires counties to establish a county-wide registry unless they opt out ^[73]. The statute as in force requires every county that is a covered jurisdiction to establish a registration system, shared registries allowed ^[74]; a county is relieved only by a local law opting out on or before the later of December 31, 2025 or nine months after the section's effective date ^[75], and an opted-out county is not a covered jurisdiction ^[76]. Whether Ulster County established a registry or opted out was not sourced, and that status decides which rules below apply.

Once a county has established its registry, the duties start. Listing or offering a unit without current, valid registration is then unlawful and makes the lister ineligible for registration for twelve months from the date a violation is determined ^[77]. Registration is valid for two years ^[78], and the fees are set by the county, with no amount in the statute ^[79]. Before the county has established a registry, hosts need not comply with those registration requirements ^[80] and no one may be fined for offering an unregistered unit ^[81]. After warning notices for the first and second violations, the county may fine up to $200 for a third violation and up to $500 per day for each later one, with a seven-day cure period after each violation is issued ^[82]. A unit must also include a posted evacuation diagram, a posted list of emergency phone numbers and a working fire extinguisher ^[83], which is why they sit in the assumed $35,000 furnishings line in Table 3, and must carry at least $300,000 of coverage for third-party property damage or bodily injury claims ^[84]. A county, city, town or village may still enact a local law prohibiting or further limiting short-term rentals, whether or not it is a covered jurisdiction ^[41].

Decision. The study would state that the deal is conventional because the investor who hires a manager is the passive landlord 13 CFR 120.110(c) describes, would leave the operating lodging question open, and would carry the state registry and the municipality's short-term rental law as conditions of the collateral.

Risks a New York Short-Term Rental Feasibility Study Would Flag

Local prohibition or a cap. At least 16 Ulster County municipalities regulate short-term rentals and at least 6 cap them or require owner occupancy ^[38], and the statute preserves every local government's power to prohibit or further limit them ^[41]. A rule adopted after the cabins are built is the risk no sensitivity table measures.

Registration lapses. Once the county has a registry, listing without a current, valid registration is unlawful and bars the lister from registration for twelve months ^[77]. For six cabins that is a year of lost revenue (Table 6).

Tax compliance on direct bookings. An operator must register as a sales tax vendor and collect the tax unless an exception applies ^[65], and cleaning and pet fees may be taxable ^[59].

Soft resort demand and price competition. AirDNA reports mountain and lake resort demand up only 0.6 percent in August and September 2026 ^[26], small listings roughly flat through the first five months of 2026 ^[28], and new listings competing on price ^[29]. A ten percent revenue miss gives 0.99x (Table 9).

Rural supply. AirDNA reports small town and rural markets growing listings 5.3 percent year over year in August and September 2026, the fastest of any location type ^[45], into a county stock of about three thousand units on either count the Ulster County Comptroller reports from AirDNA and Granicus (Host Compliance) data ^{[34] [36]}.

Credit climate and rate reset. Banks reported tightened standards for construction and land development loans in the October 2026 Senior Loan Officer survey ^[5], and the 5-year Treasury moved between 4.59 and 4.74 percent in the week to November 6, 2026 ^[3]; a two-point rise at the reset cuts coverage to 1.06x (Table 9).

Property tax. The assumed $6,000 per cabin rests on an assumed $300,000 assessed value, and the parcel's actual rate was not sourced (Table 3). Property tax is $36,000 of the assumed fixed cost lines totalling $120,000 (Tables 7 and 10), so a reassessment moves break-even occupancy directly.

Equity. The coverage-sized loan leaves 60.7% of cost, $1,481,000, to be funded with equity (Table 4) for a stabilized cash return of 1.5% on the assumed figures (Table 8).

Unsourced items. Land price, the parcel's tax rate, a cap rate or appraised value (so no value-based loan test is modelled), the bank's actual spread, coverage minimum and loan-to-cost policy, any Catskills county rate or occupancy series, and Ulster County's registry status were not sourced.

What Would Decide the Outcome

First, the rate and occupancy the cabins actually achieve. The model's assumed $265 and 52 percent are national anchors from AirDNA adjusted by judgment ^{[23] [24]}, not a Catskills measurement, and the break-even table shows the covenant is met at $264.11 and 51.8% and debt is covered at $239.14 and 47.0% (Table 10). A monthly survey of comparable cabins in the county either supports the assumptions or it does not, and the loan amount follows.

Second, the cost per cabin. The assumed $215,000 sits well below the roughly $405,600 per-unit reported construction valuation of Ulster County single-family permits in 2025 ^[52], as a 700 square foot cabin should, but because coverage fixes the loan at an assumed $960,000 (Table 8), any overrun is equity.

Third, local permission and the registry. The municipality must allow a non-owner-occupied cluster, which at least 6 Ulster County municipalities would not ^[38], and the owner must hold a registration under whatever system the county has adopted, whose status this example did not source ^{[73] [75]}.

Fourth, property tax and the five-year reset: the parcel's rate is not sourced, and a two-point reset gives 1.06x (Table 9).

Decision. The study would conclude that the deal is marginal at the stated assumptions: it meets the assumed 1.25x minimum by one hundredth (Table 8), fails it under every tested shock (Table 9), and should proceed only if a county rate and occupancy survey supports the assumptions, a bid confirms the cabin cost, and the municipality's law and the county registry are documented in the sponsor's favour.

FAQ: Choosing a Feasibility Study Company and Other Questions on New York Short-Term Rental Loans

Can a short-term rental project get an SBA loan? Not as structured here. The SBA regulation at 13 CFR 120.110(c) bars landlord-owned passive companies, those that neither use nor occupy what the loan pays for ^[67], and an investor who hires a manager is that passive landlord. Whether an operating lodging business is treated differently is a question this example does not decide.

Why does coverage, and not loan-to-cost, set the loan? Because at the assumed rate and occupancy the income is too small to service the debt the cost test would allow. The assumed 70 percent loan-to-cost limit permits $1,708,700, which would cover only 0.71x; the assumed 1.25x coverage test permits $966,804, rounded down to an assumed $960,000 (Table 8). Whenever the yield on cost, 4.35% here (Table 7), is far below the loan rate, 7.39% here (Table 8), coverage binds.

How thin is the coverage? Base coverage is 1.26x against an assumed 1.25x minimum (Table 8), and every sensitivity fails the minimum, from 0.72x to 1.06x (Table 9). Break-even occupancy at 1.00x is 47.0% and the assumed minimum needs 51.8%, against an assumed 52 percent base (Table 10).

What taxes does the guest pay, and are they the owner's revenue? Since March 1, 2025, state and local sales tax applies to short-term rental occupancy over $2.00 a night ^[56], at a combined 8 percent in Ulster County ^[50], and Ulster County's own occupancy tax is 4 percent ^[51]. The Comptroller describes the county tax as borne by the guest ^[63], so the model keeps both outside revenue: the guest's nightly cost at the assumed $265 rate is $296.80 (Table 5).

What does the New York registry law require of the owner? The law was enacted as Chapter 99 of the Laws of 2025 ^[70]. Counties must run a registration system unless they opt out by local law ^[75]; once the county has one, listing without a valid registration is unlawful and bars the lister for twelve months ^[77], and a unit must post an evacuation diagram and emergency numbers and keep a working fire extinguisher ^[83]. Whether Ulster County has a registry or opted out was not sourced here.

How should an investor choose a feasibility study company or consultant for a short-term rental project? Ask three things. Does the consultant say where the rate and occupancy come from, and admit when, as here, no county series exists? Does the consultant run the sensitivities the bank's own guidelines name ^[1] and show the break-even, so the lender sees a range from 0.72x to 1.06x (Table 9) rather than a single 1.26x? And does the consultant treat the municipality's short-term rental law and the county registry as conditions to be documented before the revenue is modelled? A feasibility consultant who presents a Catskills cabin at a single point estimate, with no local law in the file, is not writing for the lender.