Most bad short-term rental purchases are not bad properties. They are good properties bought on a revenue number that was never real.
The underwriting is not complicated, but it is unforgiving about inputs. Here is the sequence, in order, with the places people go wrong.
Step one: revenue from true comparables
The single most common error is using a market-level average. A tool that tells you the average listing in Sevierville earns $52,000 a year is telling you almost nothing, because that average blends a one-bedroom cabin with a nine-bedroom lodge.
Build the number from properties that actually compete with the one you are buying:
- Same bedroom count, and same sleeps capacity
- Same sub-market, meaning the same drive time to the demand driver
- Comparable amenity set, especially hot tub, pool, game room, and view
- Actively booked with a trailing twelve month history, not a listing that went live in March
Find eight to twelve of them. Take the median average daily rate and the median occupancy, not the mean, because a single superhost outlier distorts a small sample badly. Multiply ADR by occupancy by 365.
Then discount it. A new listing with no reviews will not hit comparable performance in year one. We model the first twelve months at 65 to 75 percent of the comparable-derived figure and only use the full number from year two.
If a seller hands you their own trailing revenue, verify it against platform data rather than accepting the statement. Owner-reported figures frequently include cleaning fees as revenue while excluding the cleaning cost, which inflates the top line by 10 to 15 percent.
Step two: build the expenses from the bottom
Do not use a percentage. Build the line items and then check what percentage they came to.
| Line | Basis | Annual |
|---|---|---|
| Cleaning (net of fee) | Turnover cost above recovered fee | $2,400 |
| Management or co-host | 15% of gross | $11,700 |
| Platform commission | 3% host fee | $2,340 |
| Utilities and internet | Full year, guest-inclusive | $4,800 |
| Supplies and consumables | Per-stay basis | $2,900 |
| Repairs and maintenance | 5% of gross | $3,900 |
| Furniture replacement reserve | 4% of gross | $3,100 |
| Software and monitoring | PMS, pricing, locks, noise | $1,400 |
| Permits and licensing | Local requirement | $600 |
| Total operating | 43% of gross | $33,140 |
Two lines get skipped constantly. Furniture replacement is real, because a property hosting three hundred guests a year consumes sofas and mattresses on a three to five year cycle. And the cleaning line is rarely fully recovered, because you will absorb cost on partial refunds, deep cleans, and gaps between the fee you can charge without hurting conversion and what the cleaner actually costs.
The management line deserves its own scrutiny because it is the largest single expense and the most variable. Full-service management at 20 to 25 percent of gross buys you genuine passivity. A co-host at 10 to 15 percent leaves you handling pricing strategy and guest messaging. Self-management costs nothing and takes ten to twenty hours a month. Model whichever structure you will actually sustain in month eighteen, not the one that makes the spreadsheet look best in month one, because switching mid-year damages your review velocity and can disturb your tax position.
Step three: debt, taxes, and insurance
Get an actual insurance quote during diligence rather than estimating. In coastal Florida and Alabama, short-term rental policies with wind and flood coverage regularly come in at two to three times what an inland buyer assumes, and it is the single most common reason a deal that penciled in the spreadsheet stops penciling at closing.
Reassess property taxes at the new assessed value, not the seller's current bill. A long-held property can carry a tax figure that will double the year after you buy it.
Step four: the three tests
We accept a deal only if it clears all three.
- Cash-on-cash of 12 percent or better on stabilized year-two operations. The mechanics of that calculation are in our cash-on-cash breakdown.
- Positive net cash flow in the first full year after ramp, before any tax benefit.
- Debt service coverage of 1.25 or better, which is also what a DSCR lender will require.
The second test is the one that matters most. A deal that only works because of the tax deduction is a deal that stops working the moment your circumstances change. The tax position should improve a good investment, not rescue a bad one.
Have a specific property in mind?
Send it over. We will run it through the same model we use internally and tell you plainly whether it clears.
Apply NowThe sensitivity check
Before committing, rerun the model with occupancy down 15 percent and ADR down 10 percent simultaneously. That is roughly what a soft year plus new supply looks like.
If the property still covers debt service, it is durable. If it goes deeply negative, you are buying a bet on the market holding, and markets do not always hold. That single stress test is why we decline more deals than we bring forward, and it is a large part of what the underwriting stage is actually for.
Keep reading
Frequently asked questions
How do you estimate revenue for an Airbnb before you buy it?
Build it from true comparables, not city averages. Identify eight to twelve active listings with the same bedroom count, same sub-market, and similar amenities, then pull their trailing twelve month occupancy and average daily rate. Multiply the median ADR by the median occupancy by 365 and discount the result for a first-year ramp.
What expense ratio should you use for a short-term rental?
Model operating expenses at 35 to 50 percent of gross revenue before debt service, depending on management structure. Self-managed properties land near the bottom of that range, co-hosted properties in the middle, and full-service managed properties at the top. Anything modeled below 30 percent is missing line items.
What return should an Airbnb deal produce to be worth buying?
Our internal filter is a minimum 12 percent cash-on-cash return on stabilized year-two operations, positive net cash flow in the first full year after ramp, and a debt service coverage ratio of at least 1.25. A deal that only works on the tax benefit and not on the operating numbers is a deal we decline.