Underwriting a long-term rental is an exercise in reading a lease. Underwriting a vacation rental is an exercise in reading a calendar. The property has no tenant, no fixed monthly rent, and no contractual promise that next February will resemble last February. What it has is a pattern: nights booked, prices charged, and revenue earned across seasons that swing wildly. A lender's job is to turn that pattern into a single defensible number — the income the loan can rely on — and then to test whether that number covers the debt. Everything else in the file is commentary on that test.
The test starts with three metrics, and confusing them is the most common underwriting error a borrower can make. Occupancy is the share of available nights that get booked. Average daily rate, or ADR, is the average nightly price guests actually paid across booked nights — nights that did not book are excluded, which is why ADR alone flatters a property. Revenue per available rental, or RevPAR, is occupancy multiplied by ADR: the earning power of every night the property was available, booked or not. RevPAR is the number that matters, because it is the only one of the three that cannot be gamed by cutting price to fill nights or by holding price and sitting empty. A property at 80% occupancy and $150 a night earns the same RevPAR as a property at 60% occupancy and $200 a night — $120 — and a lender comparing the two on ADR alone would rank them backwards.
The national picture for 2025 shows why the test has teeth. U.S. short-term rental occupancy landed at 56.9% for the full year, a roughly quarter-percentage-point increase from 56.7% in 2024 — essentially flat, and flat in a way that hides deep divisions. December 2025, the latest single month in the published review, tells the seasonal story in miniature: 1.66 million available listings, up 3.4% year over year; occupancy of 51.0%, down 1.7% from a year earlier; ADR of $248.57, up 3.3%; and RevPAR of $126.97, up 1.8%. Read those four numbers together and the lesson is plain: more supply, softer occupancy, higher prices, and revenue per available night still inching up. A borrower who underwrote on December's ADR without December's occupancy would have overstated the month's revenue by nearly double.
Full-year pricing tells a sterner story. ADR grew just 1.8% across 2025, well below the 2.7% average inflation rate for the year — meaning real nightly rates fell even as nominal rates rose. Only resort markets managed to grow rates ahead of inflation; large-city markets posted the weakest ADR growth, and urban cores alone saw declines in both ADR and the repeat-rent measure that tracks existing operators. Meanwhile supply growth, after years of deceleration, bottomed at 1.8% in July 2025 — the weakest rate on record — before rebounding to 3.4% by December as mortgage rates eased from 6.85% in early June to roughly 6.10% by year end. For an underwriter, that sequence is a warning and a relief in the same paragraph: supply pressure eased at the trough, but it is already rebuilding, and the next property in the market will compete against it.
Averages also hide the distribution, and the distribution is where loans fail. Across the fifty largest U.S. markets in 2025, exactly half gained occupancy and half lost it, ranging from a 4.9% increase in one major coastal market to a 4.4% decline in the national capital. Among full-time hosts active in both 2024 and 2025, the average change in occupancy sat near zero — but nearly one in five hosts saw occupancy fall by 20% or more. A lender that underwrites to the national average is therefore underwriting to a property that does not exist. Competent underwriting is local: comparable properties in the same market, the same location type, and the same price tier, because upscale and luxury properties posted stronger occupancy growth than budget listings in nearly every month of 2025, and resort markets behaved nothing like urban cores.
Seasonality is the second trap, and it operates on the debt-service side as much as the income side. A mortgage payment does not take February off because the property does. The correct underwriting question is never "what does this property earn in its best month" but "what does it earn across twelve months, after cleaning, platform fees, management, maintenance, insurance, taxes, and utilities." Lenders enforce that discipline bluntly: projected income is discounted — commonly by 20% to 25% — before it ever meets the ratio, and one specialist program discounts projections by 20% for new investors while letting experienced operators with nearby properties use the full figure. Reserves add a second cushion, with roughly six months of full PITIA payments in liquid assets a common expectation after closing. A file that survives both the haircut and the reserve test has margin for a bad season. A file that only survives the undiscounted projection does not have a margin at all.
Documentation decides which income figure a lender will even accept. Twelve or more months of actual short-stay income is the strongest evidence, and some programs will use the trailing twelve months of gross revenue in full. Without that history, the file rests on a projection from a data provider, built into an appraiser's market-rent analysis — and the appraisal itself becomes a second underwriting, testing whether the projection resembles what comparable properties actually earn. A third path, long-term market rent, is available at some lenders but answers a different question: it tells the lender what the property would earn as a conventional rental, which for a vacation property in a resort market can be far below its short-stay potential, or, in a soft market, misleadingly above it. Borrowers should find out which path their lender is on before they pay for an appraisal, not after.
Finally, there is the question no spreadsheet asks: may this property legally be rented short-term, and will it still be allowed next year? Specialist lenders verify local permission before funding, and for good reason — a permit that does not transfer at sale, a city cap, or a homeowners' association ban can erase the income the entire loan rests on. Regulation is also the market's defining theme: the leading industry research house describes short-term rentals as entering a new phase of scrutiny and structural growth, with tighter regulation shaping the sector alongside changing traveler expectations. An underwriter who ignores the permit file is not being optimistic. The underwriter is lending against income that may not legally exist.
The discipline, stated plainly, is this: underwrite on RevPAR, not ADR; underwrite on the full year, not the peak season; underwrite on discounted projections, not the data provider's headline; underwrite to the local market, not the national average; and verify the legal right to rent before anything else. A property that passes all five tests is financeable on its own merits, at a ratio a lender can defend and a borrower can live with. A property that passes only some of them is not a bargain waiting for a creative lender. It is a loan waiting for a bad season.
Sources: AirDNA, U.S. Review December 2025 (published January 21, 2026): December 2025 snapshot — 1.66 million available listings (+3.4% year over year), occupancy 51.0% (−1.7%), ADR $248.57 (+3.3%), RevPAR $126.97 (+1.8%); full-year occupancy 56.9% in 2025 vs 56.7% in 2024; full-year ADR growth 1.8% vs 2.7% average inflation; supply growth low of 1.8% in July 2025, rebounding to 3.4% in December; mortgage rates from 6.85% in early June to roughly 6.10% by year end; top-50 market split of 25 gains and 25 declines; nearly one in five full-time hosts with occupancy declines of 20% or more; resort markets only to grow rates ahead of inflation; urban cores declining in both ADR and repeat-rent measures, https://www.airdna.co/blog/us-review-december-2025. AirDNA metric definitions (occupancy as share of available nights booked; ADR as average nightly price paid across booked nights; RevPAR as occupancy multiplied by ADR), https://www.airdna.co/blog/us-review-december-2025. DSCR mechanics and income treatments (PITIA denominator; 1.00 to 1.25 qualifying range; 20% to 25% projected-income haircut; six months of PITIA reserves; appraisal market-rent analysis), https://www.bnbcalc.com/blog/airbnb-business/buying-rental-property/what-is-a-dscr-loan-for-airbnbs, https://www.mtgstrategists.com/blog/dscr-loan-qualification-requirements-2026. Host Financial program terms (12+ months actual STR income, projected income, or market rents; projected income discounted 20% for new investors), https://www.hostfinancial.com/short-term-rental-loans-mortgages/?airdeed. Visio Lending guide (local short-term rental permission verified before funding), https://visiolending.com/resources/short-term-rental-guide/. Phocuswright, Global Short-Term Rentals 2026 summary (market entering a phase of scrutiny and structural growth shaped by tighter regulation), https://www.phocuswright.com/Travel-Research/Research-Updates/2026/shortterm-rentals-enter-a-new-phase-of-scrutiny-and-structural-growth.