STR DSCR income: AirDNA projection vs. trailing 12-month history

Two investors can buy nearly identical cabins, submit nearly identical files, and walk away with different loan amounts for one reason: the lender chose a different way to measure short-term rental income. Some programs underwrite off a forward-looking market projection. Others want the property's own trailing twelve-month booking history. A third group takes whichever number is lower. Which method your lender uses can move your qualifying income by 20 percent or more, and on a DSCR loan, income is the entire ballgame.

This guide breaks down the three common STR income methods, runs the same property through each, and explains how to match a deal to the right program before you spend money on an appraisal.

The three ways lenders count STR income

Projection-based. The lender orders an AirDNA-style market report, or uses an appraisal addendum built on comparable STR listings, and takes a percentage of the projected gross. Haircuts typically run 20 to 30 percent of the raw figure. This is the only workable path on a purchase where the seller never operated the home as a rental, and it is the method most investors have in mind when they search for short-term rental DSCR loans.

History-based. The lender uses the property's actual trailing twelve months of platform payouts, usually backed by Airbnb or VRBO earnings statements and bank deposits. Some programs accept as little as 12 months; others want 24 for a refinance. The number is real, so the haircut is lighter, often 10 to 20 percent, but you need a track record.

Lesser-of. The lender runs both and underwrites the lower figure. This is common among conservative shops and in markets where projections have historically run ahead of actual results. It protects the lender and can quietly cost the borrower, because a strong operator in a soft-looking market gets pulled down to the projection, and a new owner in a hot market gets pulled down to history.

Before you apply, ask which of the three a program uses. It is a one-line question, and the answer determines whether your file is even in the right place.

The same cabin, three answers

Take a four-bedroom cabin priced at $520,000. The AirDNA-style report projects $84,000 of annual gross revenue. The prior owner's trailing twelve months show $71,000 of actual payouts. The appraiser's long-term rent schedule comes in at $2,900 a month, or $34,800 a year.

Run it through each method with a 25 percent haircut on the projection and a 15 percent haircut on history:

• Projection-based: $84,000 less 25 percent is $63,000, or $5,250 a month.
• History-based: $71,000 less 15 percent is $60,350, or about $5,029 a month.
• Lesser-of: the history figure, about $5,029 a month.
• Long-term rent fallback: $2,900 a month.

Now the debt side. At 70 percent LTV the loan is $364,000. On a 30-year fixed at an illustrative 7.75 percent, principal and interest is about $2,608. Add $215 a month of property tax, $430 of STR-rated insurance, and a $95 HOA, and PITIA is about $3,348.

The ratios: 1.57 on the projection, 1.50 on history, and 0.87 on the long-term rent. The first two clear any program's 1.0 to 1.25 minimum comfortably. The third fails outright, which is why a lender that defaults to the long-term rent schedule on an STR property will decline a deal that is fine on the numbers that matter.

When the projection wins, and when it misleads

Projection-based underwriting is the right tool when the property has no STR history: a purchase from an owner-occupant, a converted long-term rental, or a second home that was only used personally. It is also the better method when the operating history is weak for reasons that will not repeat, such as a prior owner who ran the listing badly, blocked the calendar for family use, or had a renovation mid-year.

The weakness is that a projection is only as good as its comp set. Reports that pull from the top quartile of listings in a ZIP code will overstate what an average-managed property earns. Lenders know this, which is why haircuts exist. As a buyer, run your own sanity check: compare the projection's implied nightly rate and occupancy to what three to five genuinely similar listings are posting for the next 90 days of calendar. If the report assumes 62 percent occupancy and comparable listings look half booked in peak season, expect the lender's reviewer to notice too.

When history wins

History is the stronger lever on a refinance or a cash-out. If you bought a home in a market like Asheville and have run it for 14 months at $93,000 of gross payouts while the market projection says $78,000, you want a lender that underwrites your actual number. That is a $15,000 swing in gross income, and at a 20 percent haircut it is the difference between roughly $5,200 and $6,200 of monthly qualifying income.

The documentation burden is real, though. Expect to provide platform earnings exports, 12 months of bank statements showing the matching deposits, and sometimes a property-management agreement and occupancy-tax filings. Deposits that do not reconcile to the platform statements are the most common reason a history-based file stalls in underwriting. If you run the property through an LLC, keep the deposits in the entity's own account from day one, a point covered in detail in the guide to LLC vesting and entity structure.

Seasonality is the quiet variable

A flat annual number hides the shape of the year. A Smoky Mountain cabin might earn 55 percent of its annual revenue between June and October and run thin from January through March. A lake or beach property can be even more lopsided. Lenders that apply one occupancy assumption across twelve months tend to underqualify strongly seasonal markets and overqualify steady ones.

Two practical consequences follow. First, when you shop programs, prefer lenders that accept month-by-month data and underwrite the annual total rather than a stress month. Second, build your own reserves to cover the slow season. Most STR-friendly programs want six to twelve months of PITIA in reserves, and a property with a 1.50 annual ratio can run well below 1.0 in its worst quarter. The Pigeon Forge deep dive shows how that plays out in a heavily seasonal market, and the Tennessee state page covers lender coverage and state-level context for the Smokies and beyond. Investors comparing a seasonal leisure market against a steadier urban STR market like Nashville should weigh occupancy consistency as heavily as peak-season rate.

Pricing: what the STR label costs

STR-purpose DSCR loans generally price 25 to 100 basis points above a comparable long-term-rental loan, and leverage caps tend to sit 5 points lower: 70 to 75 percent LTV rather than 75 to 80. Prepayment penalties are common, usually a 3-2-1 or 5-4-3-2-1 step-down. On the $364,000 loan above, a 50 basis point rate premium costs about $152 a month, or roughly $1,820 a year. In the example, the rent premium over long-term rent is roughly 80 percent, which covers that spread many times over, but it belongs in your model before you submit an offer, not after.

Run the sensitivity yourself. Plug the projection figure, the history figure, and the long-term fallback into the DSCR ratio calculators with your actual taxes and insurance, and look at how many basis points of rate or points of LTV you can lose before the ratio drops under the program minimum. If a 1.15 minimum is the line and your cushion is 0.05, the file is one appraisal surprise from a restructure.

Matching the deal to the lender

A simple decision rule works for most files:

• No operating history, purchasing: look for projection-based programs and ask for the haircut percentage in writing.
• 12 to 24 months of clean history, refinancing or cashing out: look for history-based programs, and confirm they will not also cap you at the projection.
• Strong history in a market where projections run low: avoid lesser-of programs.
• Weak history in a hot market with a credible projection: avoid history-only programs.
• Any file where the long-term rent would fail the ratio: confirm the program has a true STR income path before ordering anything.

Because these rules vary by lender rather than by market, the fastest filter is the lender directory, where STR-capable programs can be compared on the fields that matter here: income method, haircut, LTV cap, and reserve requirement. Applying to three lenders that all use the wrong method for your property wastes the same application fees as applying to three that fit.

Pre-application checklist

Pull the trailing twelve months of platform payouts and reconcile them to bank deposits. Order or run your own market projection and note the implied occupancy and nightly rate. Get a written insurance quote that names STR use, since a standard landlord policy can be declined at claim time. Confirm local licensing, occupancy-tax registration, and HOA rules, because a lender will ask and a prohibition kills the deal regardless of ratio. Finally, calculate your ratio under all three methods so you know in advance which number a lender can pick that sinks the file. If the worst case still clears the minimum, you can apply almost anywhere. If only the best case clears it, apply only where the best-case method is the one in the program guidelines.