Private Money LendersRead time: 30 seconds
- STR DSCR uses AirDNA projections, not historical Airbnb host statements alone. The lender pulls a third-party market projection and uses the lower of host history and AirDNA projection.
- The 75% rule cuts gross revenue to net. Most STR DSCR programs use 25% as an assumed operating-expense ratio (cleaning, supplies, utilities, platform fees, taxes) and underwrite to 75% of gross.
- Seasonality is stress-tested. Lenders look at the lowest three consecutive months of projected revenue, annualized — not the trailing twelve-month average.
- The result is conservative. A property generating $90K trailing twelve-month gross may underwrite to $52K effective income — and DSCR is computed off that number, not the gross.
- Properties with under 12 months of host history are underwritten almost entirely off AirDNA. New listings get the most conservative treatment because there is no host data to anchor against.
Why STR DSCR is different
A long-term rental's income is a signed twelve-month lease. The lender looks at the lease, pulls market rents from a comp set, and uses the lower of contract rent and market rent as the underwritten income. The variability is small, the data is stable, and the underwriting is well-understood across the industry.
A short-term rental's income is a stream of nightly bookings: variable, seasonal, market-dependent, regulation-sensitive, and entirely a function of the listing's nightly rate, occupancy rate, and platform conversion. The same property can produce $50K, $90K, or $130K of gross revenue across three different operators in the same year. None of those numbers map cleanly to "what should this property borrow."
Lenders responded by building a separate STR DSCR product class with its own underwriting machinery. The three core differences from long-term DSCR:
- Income source is a third-party market projection rather than a lease.
- Operating expenses are assumed at 25% of gross rather than dollar-itemized from the borrower's P&L.
- Seasonality drives a stress test that produces a downside underwritten income, not just an average.
The result is a structure that prices STR DSCR loans slightly higher than long-term DSCR (typically 25–75 bps of rate premium) and caps LTV slightly lower (typically 70–75% vs 75–80% on long-term DSCR), reflecting the higher revenue volatility.
How AirDNA feeds underwriting
AirDNA is the dominant third-party data provider for STR market analytics. Their Rentalizer product produces a property-specific projection that lenders use as a primary input into the underwriting model. The projection includes:
- Projected annual revenue at a specified occupancy rate (typically 60–70% submarket median).
- Average daily rate (ADR) derived from comparable active listings within a configurable radius.
- Occupancy rate based on actual booking data scraped from Airbnb and Vrbo.
- Seasonality curve month-by-month, showing peak and trough months.
The lender pulls the AirDNA report directly. The borrower can submit their own Rentalizer report, but the underwriter independently verifies the inputs (bedroom count, bathroom count, amenities, exact location) before relying on it. Most lenders use AirDNA's percentile selector to pick the comp set quality — 50th percentile is the median STR in the submarket, 75th percentile is the upper quartile. Conservative underwriting uses 50th percentile; aggressive uses 75th. Ask your lender which percentile they price off.
The blend between AirDNA projection and host history:
| Property type | Income input | Typical conservatism |
|---|---|---|
| New listing, no host history | AirDNA 50th percentile projection only | Highest — no anchor |
| 1–11 months host history | Lower of (annualized host data, AirDNA projection) | High |
| 12+ months host history, trending up | Lower of (TTM host data, AirDNA projection) | Medium |
| 12+ months host history, trending down | Lower of (last 6 months annualized, AirDNA projection) | High — trend is unfavorable |
The 75% rule
STR gross revenue is not net cash flow. After cleaning fees, platform commissions, occupancy and lodging taxes, supplies, utilities, internet, professional photography refreshes, maintenance, and dynamic-pricing software, a typical STR operates at 65–75% of gross.
Rather than ask each borrower to itemize their operating expenses, most lenders apply a flat 25% assumed operating-expense ratio. The 75% of gross becomes the effective rental income (ERI) that feeds DSCR.
From ERI, the calculation continues exactly like a long-term DSCR underwrite:
Most STR DSCR programs require a minimum DSCR of 1.10 or 1.15 (slightly higher than the 1.00–1.10 minimum on long-term DSCR, reflecting the higher cash-flow volatility). A DSCR of 1.17 in this example would qualify; a 1.05 would not.
Seasonality stress
The trailing twelve-month average obscures the worst three months. A beach property earning $14K in July might earn $3K in February. The lender wants to know whether the property covers debt service in February, not just on average.
The standard test: take the AirDNA monthly projection, identify the three lowest consecutive months, annualize that figure, and verify DSCR holds at the annualized trough. The annualized trough is computed as:
A trough DSCR of 0.56 means the property covers only 56% of its monthly debt service during its three worst months — funded by the borrower's reserves, not by operations. Most lenders require either a minimum trough DSCR (often 0.85–1.00) or a minimum reserves balance equal to 6–9 months of PITIA. Reserves are the standard mitigant; a property whose trough DSCR drops below 0.80 is rarely fundable without meaningful borrower liquidity.
Worked underwriting example
A four-bedroom cabin in Broken Bow, Oklahoma. Purchase price $620,000. Borrower contributes 25% down ($155,000) and finances $465,000 at 8.50%, 30-year amortization. Taxes, insurance, HOA, and pool service total $1,100/month. Monthly PITIA: $4,675.
Host has 18 months of operating history showing $94,000 trailing-twelve-month gross. AirDNA 50th-percentile projection for the property: $87,200 annual gross. Lower of the two: $87,200.
| Underwriting line | Value |
|---|---|
| Underwritten gross | $87,200 |
| × 75% expense haircut | $65,400 |
| ÷ 12 = monthly ERI | $5,450 |
| PITIA | $4,675 |
| DSCR (annual) | 1.17 |
Now the seasonality test. AirDNA's monthly projection shows June–August averaging $11,400/month, December–February averaging $4,200/month. Three lowest consecutive months (Dec–Feb): $12,600 total → annualized × 4 = $50,400 → × 75% = $37,800 → ÷12 = $3,150 monthly ERI in trough.
| Trough metric | Value |
|---|---|
| Monthly ERI in trough | $3,150 |
| PITIA | $4,675 |
| Trough DSCR | 0.67 |
| Required reserves (9 months PITIA) | $42,075 |
The annual DSCR clears (1.17 > 1.10), but the trough DSCR is 0.67. The property funds, but only if the borrower has $42,075 in documented post-close reserves. That requirement is often the binding constraint on STR DSCR loans, not the DSCR ratio itself.
STR DSCR loan caps and pricing
How most STR DSCR programs price relative to long-term DSCR:
| Term | Long-term DSCR | STR DSCR |
|---|---|---|
| Max LTV (purchase) | 75–80% | 70–75% |
| Max LTV (cash-out refi) | 70–75% | 65–70% |
| Rate premium (vs long-term) | — | +25 to +75 bps |
| Min DSCR | 1.00–1.10 | 1.10–1.15 |
| Reserves requirement | 3–6 months PITIA | 6–9 months PITIA (often binding) |
| Min FICO | 660–680 | 680–700 |
| Prepayment penalty (typical) | 3-2-1 step-down | 3-2-1 step-down |
What can go wrong
Three failure modes worth flagging:
- Regulation risk. Cities and HOAs change STR rules. A property qualified at $87K of projected revenue may earn $0 the day a municipal ban takes effect. Lenders mitigate by checking current regulations and avoiding markets with active ban legislation; borrowers can mitigate by reading the latest HOA bylaws and local STR ordinances before purchasing.
- Saturation risk. A submarket can absorb only so many listings. As supply grows, occupancy falls. AirDNA's projection reflects current submarket dynamics; if 200 new listings come online in the next 12 months, the projection becomes stale. Submarkets we currently flag for saturation: parts of the Smoky Mountains, certain Florida Panhandle cities, and selected mountain ski towns.
- Operator dependence. The same property under a professional operator may earn 30% more than under an inexperienced host. AirDNA's projection assumes median operator performance. A property bought with the assumption that the new owner will perform at the top quartile is being underwritten optimistically.
A short-term rental's gross revenue is a marketing number. The underwritten number — the 75% of the lower of AirDNA and host history, stress-tested against the worst three months — is what determines what the property can borrow.PML Underwriting Team
Glossary
- AirDNA
The dominant third-party data provider for short-term-rental market analytics. Their Rentalizer product produces a property-specific revenue projection used by most STR DSCR lenders.
- ADR (average daily rate)
The average price per night booked across the trailing 12 months or the projection period. The driver of revenue alongside occupancy rate.
- Occupancy rate
Booked nights divided by available nights. STR submarket medians typically range 50–70% on annual average.
- RevPAR
Revenue per available rental — ADR × occupancy rate. The single best summary metric of a property's STR performance.
- Effective rental income (ERI)
Underwritten gross × (1 - expense ratio). The income the lender uses to compute DSCR.
- Trough DSCR
DSCR computed off the three lowest consecutive months of projected gross, annualized. The seasonality stress test for STR DSCR.
- STR regulation risk
The risk that a municipality or HOA changes its short-term-rental rules in a way that reduces or eliminates the property's STR income.
Frequently asked
Can I use my actual Airbnb host statement instead of AirDNA?
You can submit it; the lender will use it as one input. The final underwritten income is the lower of your host statement and the AirDNA projection. Lenders blend rather than substitute, because a host with 12 months of host data can demonstrate a top-quartile operating result that the median-projection AirDNA report would understate, and a host with 6 months of poor results does not get to wait out the bad months.
What if my listing is new and has no host history?
The lender uses the AirDNA 50th-percentile projection as the income input, often with an additional 10% haircut for the operator inexperience risk. New listings are the most conservative STR DSCR underwrites and the hardest to push above 70% LTV.
How is the 25% expense ratio calculated?
It is an industry assumption blending platform fees (15–18% of gross from Airbnb/Vrbo + payment processing), cleaning costs (typically 5–8% on properties with cleaning passed to guests, higher when absorbed), utilities and internet (2–4%), supplies and consumables (1–2%), professional photography refreshes (~1% annualized), STR taxes that the platform does not collect, dynamic-pricing software (~1%), and maintenance reserves (3–5%). Real numbers vary but the median STR operates near 75–78% of gross, so 25% is a reasonable industry assumption.
Can I lower the assumed expense ratio with a P&L?
On some programs yes — bring 12+ months of itemized expense documentation including platform statements, cleaning receipts, and utility bills. The lender can underwrite to an actual expense ratio if it is clean. Most borrowers do not have the records to make this case and end up underwritten to the standard 25%.
What happens if my city bans short-term rentals after I close?
The DSCR underwriting goes to zero on the STR income. The borrower is expected to convert to long-term rental at the prevailing market rent. If the long-term DSCR clears at 1.10+, the loan remains current. If it does not, the borrower is on the hook for the cash-flow gap or must refinance. This is the single largest tail risk in STR DSCR lending and is why lenders avoid markets with active STR legislation.
Does PML offer STR DSCR loans?
Yes — STR DSCR is a subset of our rental program. Pricing differs from long-term DSCR by 25–75 bps depending on market, host history, and FICO. We use AirDNA as the third-party projection source and apply the 75% / 1.10 / trough-test underwriting framework described above.
How is DSCR underwritten on a property with split STR / LTR use?
If the property has been STR for portions of the year and LTR for other portions, the lender weighted-blends. Annual revenue is the host's combined revenue; expense ratio shifts toward the LTR side (more like 30–35%) because LTR expenses include management, vacancy, and tenant-turnover costs that STRs do not have. The mix is documented and modeled — not averaged blindly.
What if AirDNA does not have data for my submarket?
Rare but it happens, especially in very small towns. The lender pulls the closest comp submarket and applies a haircut for the lack of direct data — usually 10–15% off the projection. Properties in unrated submarkets are difficult to underwrite at high LTV and are often shifted to a long-term DSCR underwrite at the prevailing market rent.
Want your STR underwritten on real numbers, not optimistic ones?
Send us the property address, your Airbnb / Vrbo URL if it is live, and your proposed loan amount. We will pull the AirDNA report and quote the deal — annual DSCR, trough DSCR, required reserves — in writing within four business hours.