The Quick Read: AirDNA projections are reasonably reliable at the market level and much less certain at the single-address level, so a lender treats one as a modeled estimate, not a measured result. Lenders usually discount the number, or fall back on other income evidence, before it counts toward coverage.
- Market-level occupancy and daily-rate trends are the strongest part of the data.
- A property-level projection is a model’s guess, and commentators say it often runs high.
- Real hosting history converts the estimate into proof and is the strongest evidence in a file.
- Short-term-rental purchase leverage tops out at 75% LTV, with a 1.00 coverage floor on purchases, subject to lender guidelines.
- A projection that falls short can shrink the loan, so underwrite your own deal below the raw number.
What Does the AirDNA Number Actually Measure?
It measures modeled activity on public Airbnb and Vrbo listings. It does not measure what a host was paid. AirDNA collects calendars, prices, reviews and availability, then estimates occupancy, daily rate and revenue per listing, neighborhood and market, per a methodology explainer from magicbnb.
DSCR Calculator
Run the numbers in your market
Rate source: Freddie Mac 30-yr average via FRED® — Federal Reserve Bank of St. Louis · effective Oct 15, 2026
Prefilled with starting assumptions — enter your own rent or nightly figures, taxes, insurance, and HOA for a more accurate picture.
Program parameters update from Lendmire’s centralized guideline source. Taxes and insurance are editable estimates.
As of Oct 15, 2026 · General Freddie Mac market benchmark, not a Lendmire loan offer. Rent, nightly rate, occupancy, taxes, and insurance are editable estimates. Short-term rental figures are estimates only and vary significantly by season, property type, management approach, and local short-term-rental rules — confirm local regulations before relying on them. Qualifying income for short-term rentals varies by program — some use appraisal market rent, others use documented STR history or projections — and is confirmed in underwriting. Not a Loan Estimate, approval, or commitment to lend. Program availability and eligibility are subject to lender guidelines, credit approval, property review, and underwriting.
Here is the catch. The tool has no direct booking access. A model has to guess whether a blocked night is a real guest stay or an owner block. That inference is where most of the error comes from.
AirDNA says its data matches Airbnb’s SEC-reported figures with 94.9% accuracy and Vrbo’s with 98.7%, per AirDNA’s own accuracy page. Treat that as a company claim. Awning’s review points out the more important limit: those figures describe how well the data matches platform-level reporting. They do not describe how closely a forecast for one property matches what one owner earns.
So the 95% headline answers one question. Your file asks a different one.
Key Terms Defined
Projection: A modeled estimate of a property’s short-term-rental revenue, built from comparable listings rather than the property’s own records.
Hosting history: The property’s actual platform earnings over a trailing period, shown through payout statements or platform reports.
Rent schedule (Form 1007): An appraiser’s estimate of long-term market rent for the property, used as a conservative fallback.
Haircut: A lender’s discount applied to a projection to allow for model error before the income counts toward coverage.
Coverage ratio: The property’s rent used for lender review divided by its full monthly obligation: principal, interest, taxes, insurance and any HOA dues.
Is the Error Random or Does It Lean One Way?
It leans high. Commentators put property-level misses at roughly 15 to 30 percent in thin markets, and several describe the skew as optimistic. That range is a commentator estimate, not a published study.
Awning’s review cites property-level estimates running well off in thin markets. VaultSTR says projections skew optimistic because they describe established comparable listings, not launch-year performance. Two of these sources may be repeating the same underlying claim, so do not count them as independent confirmation.
Why would the number run high? Three mechanisms show up repeatedly.
| Source of bias | Direction | Effect on a DSCR file |
|---|---|---|
| Comps are established listings | Upward | Projection assumes mature performance |
| Owner-blocked nights counted as stays | Upward | Occupancy looks stronger than reality |
| Matching by distance over amenities | Either way | Unusual properties misfire |
An academic paper on differing views of lodging reality found upward bias in AirDNA’s published metrics. It also found AirDNA defines occupancy, daily rate and RevPAR differently than the traditional lodging benchmark. The paper is several years old, so use it for the structural point only: “occupancy” in AirDNA is not the same yardstick a hotel analyst uses.
One more limit. The model was reportedly trained on data gathered before the platforms hid booked-versus-blocked detail, according to RentalRecon. Nobody has published an independent, audited comparison of projections against realized revenue. Any precise accuracy percentage you see quoted is indicative at best.
How Do Lenders Actually Use the Projection?
Lenders use it as one of several income paths, then discount it. Across the wholesale network Lendmire works with, the pattern is consistent: DSCR lender review runs primarily on property-level rental income covering the payment, subject to lender guidelines. For short-term rentals, the evidence hierarchy usually looks like this.
1. Documented platform history, typically about 12 months of hosting. 2. A market-based projection report where history does not exist. 3. The appraiser’s long-term rent schedule as a conservative fallback.
Programs differ. Some accept a projection report, some rely on an appraisal-based short-term-rental analysis, and some want trailing history only. The factor applied to the projection is lender-dependent, which is why two lenders can produce different qualifying income from the same address. One educational page notes that the number underwriting adopts is generally more conservative than the raw figure, and that a deal that only works at the full estimate may not survive its own loan file.
Think of the haircut as the lender’s built-in allowance for model error. That is where “accuracy” turns into a financing question.
On the network’s short-term-rental programs, purchase leverage reaches 75% LTV. Refinance runs around 70%, and cash-out on short-term-rental collateral is 70%. Borrowers should expect a 640+ credit score and about 12 months of hosting history, with a 1.00 coverage floor on purchases and 1.00 on refinances. These are typical ranges from select lenders, not guarantees, and every file is underwritten individually. For the full mechanics, see Lendmire’s complete DSCR loans guide. Every figure here varies by lender and program — guidelines, property type, leverage, and credit profile all apply.
Is the Lender’s Number the Same as the Free Tool’s Number?
Usually not. Expect the lender’s figure, not the one a free calculator showed you at contract. Different report providers and different comp sets can return different numbers for the same address. A lender may also pull its own report, so your pro forma and the file can diverge before any haircut is applied.
BNBCalc’s review says the headline comes from an algorithm, not from the comps displayed on screen. It also says the matching leans on distance more than amenities. That review comes from a competing tool vendor, so weigh it accordingly. Still, the practical lesson holds: the comps you see may not be the comps that drive the number.
Many investors confuse two different questions. Accurate for your pro forma is one question. Accurate enough for qualifying income is another. A projection can be fine for choosing which market to buy in and still be too soft to carry a loan.
What Does the Underwriter Check in the Report?
A reviewer checks whether the report fits the subject property. Experienced processors look at the same handful of items on every short-term-rental file.
- Comp fit: unit type, bedroom and bath count, and distance from the subject.
- Report age: projections change between pull dates, so a stale report is a gap.
- Market depth: how many comparable active listings exist. Fewer than a few dozen comps widens the error considerably, per RentalRecon.
- Property fit: amenities, condition and management quality that make the property diverge from its comps.
- Legality: short-term rental rules can vary by city, county, HOA, and property type, so investors should confirm local rules before relying on projected rental income.
Thin, rural, highly seasonal and unusual-property files fail most often at the comp-fit step. A rural cabin with a handful of comparable listings is a harder file than a standard condo in a deep market.
What Happens When the Number Is Wrong or Moves?
Projections change between pull dates, and a file can pend when the figure drops after contract. This is the quiet timing risk on short-term-rental purchases. An investor who priced the offer off the first report can find the lender’s later pull lower.
When you believe the projection is low, the legitimate response is documentation, not argument. Useful evidence includes:
- Trailing platform statements from the property or a near-identical unit you operate.
- A property manager’s written income history for comparable local properties.
- An appraiser’s short-term-rental analysis, where the program accepts one.
Evidence like this converts an estimate into proof. Actual platform income is the strongest path in the file. If none of it exists, the rent schedule fallback is the conservative backstop, and some deals work on long-term rent alone.
Lendmire’s piece Does a DSCR Lender Use the Old Lease or Short-Term Projections? covers how that income-basis choice plays out on existing rentals.
How Much Can a Projection Miss Move the Coverage Number?
Quite a bit, because coverage is a ratio and the denominator does not move. Run the numbers on a hypothetical short-term-rental purchase where the raw projection clears roughly 1.25x against the full monthly obligation. These are modeled assumptions, not market data.
- Projection accepted at face value: coverage clears roughly 1.25x.
- Projection reduced by a lender haircut: coverage falls toward 1.00x, near the purchase floor.
- Projection misses and the haircut stacks on top: coverage slips below 1.00x on short-term income alone.
Below 1.00, the file does not necessarily end. A separate select-lender path takes coverage below 1.00, with leverage and terms adjusted. Interest-only structures and extended terms through select lenders can also change the ratio. All of that remains subject to credit approval, property review and lender guidelines.
Here is a point investors miss. Clearing 1.00 is not positive cash flow. The ratio compares rent to principal, interest, taxes, insurance and HOA only. Repairs, vacancy, management, utilities and capex sit outside it. BNBCalc notes that Rentalizer uses flat expense defaults and leaves line items at zero. Your real operating costs are yours to model.
A bigger down payment can lower the monthly obligation and lift the ratio. It never erases leverage caps, credit floors, reserve rules or property eligibility. The strongest files clear both tests: enough equity and enough rental coverage. Reserves commonly run about 6 months of PITIA, stepping up to about 9 months above $1,500,000, and vary by lender, leverage and loan size.
DSCR vs. conventional financing
There are two common ways to finance an investment property, and they qualify you differently — here’s how investors weigh them.
Why investors choose it
- Qualifies on the property’s rental income — no personal tax returns, W-2s, or pay stubs needed to document income.
- No personal debt-to-income ceiling to clear, so existing mortgages and obligations don’t cap your borrowing the same way.
- Can be closed in an LLC, keeping the property inside a business entity.
- Built for scaling — not held to the limit on number of financed properties that conventional financing applies.
- Underwriting centers on the deal: generally qualifies when the rent covers the payment, a 1.00x coverage ratio being a common baseline (confirmed in underwriting).
- Designed specifically for investment property, including long-term and, where the program allows, short-term rentals.
Where it’s strong
- Often the lowest ongoing financing cost for a buyer who fully qualifies on personal income — a fit for a first property or a cost-first purchase.
Trade-offs for investors
- Requires full personal income documentation and must fit within a debt-to-income limit — salary, existing debts, and other mortgages all count.
- Typically held in your personal name rather than a business entity.
- Caps how many financed properties you can carry, which can become a ceiling as a portfolio grows.
- Evaluates you as a borrower as much as the property, which usually means more paperwork.
How investors usually choose: a first or single property often optimizes for the lowest financing cost; portfolio builders often optimize for leverage, vesting in an LLC, and scaling past conventional caps. The right answer depends on your goals, the property, and current guidelines — both paths run through select lenders in Lendmire’s wholesale network, with eligibility and terms confirmed in underwriting.
How Should You Pressure-Test Before You Submit?
Discount the raw projection yourself, then compare it to local actuals. A practical sequence works.
1. Pull the projection and note the comp set behind it. 2. Compare against operators in the same submarket: actual occupancy, daily rate and seasonality. 3. Discount the raw figure before you set the offer price. A downward adjustment of roughly 10 to 15 percent is a common rule of thumb, but treat it as a soft guide rather than a standard. 4. Run coverage on three inputs: projection, discounted projection and rent schedule. 5. Document any variance you can explain with real records.
Awning’s review notes the tool is hard to justify for a single property in a market you already know well, but useful for the market-level data lenders recognize. That fits how files work. Use the data to choose the market. Use your own records to carry the loan.
Experience from the broker side: short-term-rental files in tourist-heavy markets often pencil on the raw projection and tighten once the lender’s report, haircut and fresh insurance quote all land. Running the downside case before contract is the habit that keeps these files clean. Any rule on income basis, minimum market grade or documentation differs by lender, so Lendmire’s team compares options across the network rather than assuming one program’s approach applies everywhere.
When Is Another Income Path the Better Fit?
When the market is thin, the property is unusual or you have no hosting history, lead with something other than the projection. A standard long-term lease can carry a file on its own. A conservative rent schedule can do the same when the short-term numbers feel soft. Investors who already operate a comparable unit locally should lead with that history.
Mixed strategies work too. An investor can underwrite on long-term rent and treat short-term income as upside. That keeps the file independent of a modeled number. It is the lowest-friction route when the projection is the weakest link.
Properties outside these programs are simply not offered. DSCR on manufactured homes, log homes and barndominiums falls outside the network, however strong the projection looks.
Tax treatment can depend on how the funds are used and how the property is held; investors should keep clear records and speak with a qualified tax professional before relying on any deduction.
Frequently Asked Questions
Is AirDNA’s 95% accuracy claim relevant to my property?
Not directly. The company says its data matches Airbnb and Vrbo filings at the platform level, per AirDNA’s accuracy page. That measures totals, not a forecast for one address. A single-property projection carries a much wider error band, especially in thin markets.
Will a lender underwrite the full AirDNA projection?
Typically not. Lenders usually apply a haircut or run their own report, and the factor varies by lender. A deal that only works at the full projection may not survive the loan file, so model your own downside first.
Does a DSCR lender ever count a projection when there is no hosting history?
Some programs do, and others want history or an appraisal-based analysis. On the network’s short-term-rental programs, expect about 12 months of hosting history as the common expectation, subject to lender guidelines. Where history is missing, the program choice matters more than the projection.
What if the projection comes in lower than my pro forma?
Submit documentation. Trailing platform statements, a manager’s income history for comparable properties or an appraiser’s short-term-rental analysis can support a higher figure where the program accepts them. If the coverage still falls short, a select-lender path below 1.00 exists, with leverage and terms adjusted.
Does clearing 1.00 mean the rental will cash flow?
No. The ratio covers principal, interest, taxes, insurance and HOA only. Vacancy, repairs, management, utilities and capex are outside it, so cash flow depends on your real operating costs.
If you are buying or refinancing a rental property and want to see how the numbers work, Lendmire can help you compare DSCR loan options based on the property income, credit profile, leverage, and investor goals. The illustrative calculator on this site gives an editable estimate only. This is not a commitment to lend.
About Lendmire
A non-QM mortgage broker (NMLS# 2371349), Lendmire arranges DSCR financing for real estate investors in 41 markets — 40 states plus Washington, D.C. Because deals are underwritten primarily on property cash flow rather than personal income documentation, the structure suits self-employed buyers and entity-owned portfolios. Lendmire places loans through wholesale investor lenders; it is not a direct lender. Scotsman Guide named Lendmire a Top Mortgage Workplace in both 2025 and 2026.
Get Started
Ready to find the right loan for you?
In about 30 seconds you can review financing options available for your investment property. No commitment required.
Informational only. Not a Loan Estimate, approval, or commitment to lend. Program availability and eligibility are subject to lender guidelines, credit approval, property review, and underwriting.
References
1. magicbnb: Is AirDNA Accurate
5. ResearchGate: Differing Views of Lodging Reality
7. One More Door AirDNA Review
9. RentalRecon: Is AirDNA Worth It
This article is part of Lendmire’s DSCR loan program — full qualification details, guidelines, and scenarios live on the program page.
Related reading: Blanket Loan vs Per-Property DSCR Cash-Outs on Lender Availability · How to Get a DSCR Loan From Application to Closing · How to Come Up With a DSCR Down Payment Without Selling Anything
Brandon Miller
Founder & CEO, Mortgage Loan Originator, Lendmire LLC
- Mortgage Loan Originator · NMLS# 1129696 · Verify on NMLS Consumer Access
- North Carolina Real Estate Broker · License# 343312 · Verify on NCREC
- North Carolina Insurance Producer · License# 19053198 · Property, Casualty, Life, Health · Verify on NAIC SBS
- Lendmire LLC · Firm NMLS# 2371349 · Verify firm licensure
Disclosures. The information presented in this article is general market commentary, not financial, legal, or tax advice. Lendmire is a mortgage brokerage (NMLS# 2371349) — not a direct lender or depository institution — and loan placement is subject to lender underwriting. Nothing in this content represents a commitment to lend. Loan terms, pricing, and program availability vary based on borrower qualifications, property characteristics, and state of subject property, and are subject to change at any time. Lendmire complies with Equal Housing Opportunity requirements. Consumer access: nmlsconsumeraccess.org.