Is Airbnb still profitable in 2026? Conditionally yes. Short-term rentals still clear real net margins for operators who bought at a sane price, in a market with more than one reason to visit, under rules that are unlikely to change, and who run the property like a business. The same property bought on peak-year projections, financed at today's rates, with no reserve, loses money. Profitability is no longer a property-type question. It is an underwriting and operations question, and it gets answered address by address.
Last updated: 2026-09-18
I am writing this for the operator who already owns one or two units and is deciding whether to hold, refinance, add another, or get out. Nothing below is a forecast. Where I use a number, it comes from a named source you can open yourself. Where I use an assumption, I label it an assumption.
The Answer, Stated Plainly
Demand did not disappear; the operator's cost side is what moved. Airbnb reported in its Q2 2026 financial results that revenue grew 17 percent year over year to $3.6 billion, gross booking value grew 16 percent to $27.2 billion, and Nights and Seats Booked grew 10 percent year over year, with what the company described as a moderate increase in average daily rate. That is platform-level demand, not your listing. But it settles one argument: travelers are still booking short-term stays in volume, and at rates that are not collapsing.
What tightened is the operator side of the equation. Financing costs more. Insurance costs more. Cleaning, maintenance, and management cost more. In markets where a lot of units came online at once, each listing takes a smaller share of the same nights. None of that is fatal. All of it eats the cushion that used to hide sloppy underwriting.
So the conditions are: cost basis, demand durability, regulatory durability, and operator quality. Miss one and you can survive. Miss two and the property becomes a slow bleed.
What Actually Changed Since the Peak Years
Four things moved, and they moved together.
Financing. Freddie Mac's Primary Mortgage Market Survey put the average 30-year fixed rate at 6.95 percent for the week of September 17, 2026. A deal underwritten at a 4 percent note does not survive being re-cut at 7. On a $400,000 loan amortized over 30 years, the jump from 4 percent to 6.95 percent adds about $740 a month in principal and interest — my own arithmetic on the Freddie Mac rate, not a published figure. That $740 is often the entire difference between a listing that funds its reserves and one that does not.
Insurance. Insurance is the line operators understate most, and the published data only shows the start of the climb. The Insurance Information Institute, reporting National Association of Insurance Commissioners data, shows the average U.S. homeowners premium rising from $1,311 in 2020 to $1,569 in 2022 — an 11.2 percent single-year jump, and the latest year in that series. Short-term rental coverage sits above standard homeowners pricing, and in coastal and wildfire-exposed counties the spread is wider still. If your pro forma carries a flat insurance number from three years ago, your pro forma is wrong.
Regulation. Registration and permitting regimes are now normal rather than exotic. New York City runs short-term rental registration through the Mayor's Office of Special Enforcement. Austin requires a city license by property type, and as of July 1, 2026 the city began asking platforms to remove unlicensed properties from their listings. The relevant point for underwriting is not that rules exist. It is that rules are the one variable that can take revenue to zero instead of down 15 percent.
Supply and occupancy. I am not quoting a national supply-growth or occupancy-change number here. I could not tie one to a live primary source, and a national average would not describe your ZIP code anyway. Pull your own market's active-listing count and comp-set occupancy from whatever data tool you already pay for, compare it to the same month last year, and use that. A number you sourced beats a number I paraphrased.
Where the Math Still Works
The math still works where four conditions hold at once: a sane cost basis, several unrelated demand drivers, durable rules, and an operator who runs the unit like a business. They compound, so a miss in one gets expensive fast.
Condition | What good looks like | What to check before you commit |
|---|
Cost basis | Purchased at or below current comparable sales, not at a peak-year price | Your all-in cost per bedroom versus resale comps from the last 90 days |
Demand drivers | Three or more unrelated reasons to visit, spread across the calendar | Hospital, university, employer, park, venue, or airport within reasonable drive time |
Regulatory durability | Registration exists, is administered, and caps new entrants | Current ordinance text plus the last 12 months of council agendas |
Operator quality | Direct bookings, reserve funding, fast turnovers, rate automation in place | Your own review score trend and reserve balance, not a vendor promise |
Markets with several unrelated demand drivers hold occupancy better than single-driver markets, because the failure of one driver does not empty the calendar. A ski town with no summer economy is a six-month business paying twelve months of carry. A mid-sized city with a hospital system, a university, and a convention calendar has three chances every month to fill a night.
Regulation is uncomfortable but useful once you are inside it. A market that caps permits and enforces them protects the operators who hold the permits. The risk is not being regulated. It is buying before the rules land and discovering your unit is on the wrong side of them.
The Margin Structure of an Operating Listing
Below is a planning model, not a measured result and not a promise. Every line is an assumption you should replace with your own actuals or your own verified comp data. I use a single-unit, financed, three-bedroom listing at a $60,000 annual gross booking figure because it makes the arithmetic legible.
Line item | Planning assumption (% of gross) | Notes |
|---|
Gross booking revenue | 100% | Assumption: comp-set derived, not vendor projection |
Platform and payment fees | 3–5% | Varies by fee structure and channel mix |
Cleaning (net of guest-paid fees) | 3–8% | Assumption: turnover cost not fully recovered |
Supplies, consumables, restocking | 2–4% | Rises with group size |
Utilities and internet | 5–8% | Climate dependent |
Repairs and maintenance | 4–6% | Assumption: excludes capital replacements |
Capital reserve | 4–6% | Roof, HVAC, appliances, mattresses |
Insurance | 3–6% | Assumption: STR-rated policy, not homeowners |
Property taxes | 6–12% | Wide jurisdictional spread |
Management or your own labor priced at market | 10–20% | Price your hours even if you self-manage |
Debt service | 25–40% | Assumption: financed at prevailing rates per Freddie Mac PMMS |
Add the middle of those ranges and you can see the structural problem: there is very little room left. That is the plain state of a financed listing in 2026. Cash flow on a leveraged single unit is thin, and the return shows up in amortization, tax treatment, and equity — not in a large monthly deposit. The listings printing meaningful monthly cash are usually ones bought years ago, bought with a large down payment, bought cash, or operated at a rate premium the comp set cannot match.
This is educational, not tax or legal advice. How depreciation, material participation, and occupancy taxes apply to your situation is a conversation for your CPA; the IRS guidance on residential rental income is the starting reference, not the answer.
The Profile of Deals That Stopped Working
Almost every failed short-term rental I have looked at failed on underwriting, not on the platform. I have watched the same four failures repeat.
The first is buying on peak-year projections. A property is underwritten on the best twelve months the market ever had, plus a revenue estimate from a tool with no comp adjustment. Then a normal year arrives and revenue lands 20 percent under plan against fixed costs.
The second is a reserve of zero. The deal only closes if you assume nothing breaks. Something breaks. A $9,000 HVAC replacement becomes a credit card balance, and the credit card balance becomes a distressed sale.
The third is single-point demand. One employer, one event, one season. When that driver moves, the calendar does not soften — it empties.
The fourth is regulatory blindness. The permit question was answered by the listing agent instead of the ordinance. In a registration market, "the neighbors do it" is not a compliance position.
Notice what is missing from that list: the platform. None of these failures were caused by Airbnb. They were caused by underwriting that needed a perfect year.
Your Decision Checklist for Holding or Entering Now
Work these five steps in order and the hold-or-sell answer usually writes itself. Run them in sequence. Stop at the first one that fails, because the later steps cannot rescue an earlier miss.
Step 1 — Rebuild the revenue line from your own comp set
Pull 12 months of actual booked nights and average daily rate for the closest 10 comparable units, not city averages. If you own the unit already, use your own trailing twelve months. Write the number down before you look at any pro forma, so the pro forma cannot anchor you.
Step 2 — Reprice every expense at today's cost
Get a current short-term rental insurance quote, a current cleaning quote, and your actual utility bills. Add a capital reserve line even if you have never funded one. Read the bills, not the pro forma. Price your own labor at what you would pay a manager.
Step 3 — Answer the regulation question from the ordinance itself
Read the current ordinance and the last year of council or commission agendas for the jurisdiction. If registration exists, confirm your unit qualifies today and that the permit transfers. If a moratorium is being discussed, treat the deal as a coin flip regardless of the numbers.
Step 4 — Stress the model two ways
Cut revenue 20 percent and hold costs flat. Then hold revenue flat and raise insurance plus taxes 15 percent. If either scenario turns negative and you have fewer than six months of reserves, the deal fails the test even if the base case looks fine.
Step 5 — Decide, and write the decision down
Hold, improve, refinance, or sell. Give the decision a review date and a trigger — for example, two consecutive quarters below the stressed revenue case means you list the property. Deals rarely die from a bad month. They die from a year of not deciding.
If the checklist tells you the property does not work, that is a useful answer. Some operators reading this should sell into a stable market rather than defend a thin margin for three more years. I would rather you hear that from me now than learn it from a forced sale.
I have spent more than 18 years in real estate and have focused on short-term rentals since 2016, and I have trained well over 10,000 operators through CashFlow Diary. The pattern has not changed with the market cycle: the operators who survive are the ones whose numbers were conservative before they needed to be. The market decides the ceiling. Your underwriting decides the floor.
If you want a second set of eyes on the deal in front of you, book a complimentary one-on-one diagnostic call at https://cashflowdiary.com/diagnostic/ and bring your actual numbers — revenue, expenses, debt service, and the ordinance.
Frequently Asked Questions
Is Airbnb still profitable in 2026?
Conditionally yes. Airbnb's Q2 2026 results show nights and seats booked up 10 percent year over year and revenue up 17 percent, so demand is intact at the platform level. Whether your listing is profitable depends on your cost basis, your financing, your insurance and tax lines, the durability of local demand and regulation, and how well the unit is operated. A well-bought, well-run unit still works. A unit bought on peak-year projections with no reserve usually does not.
What margin should I plan for on a financed short-term rental?
Plan conservatively and verify with your own actuals. The planning model in this article shows how little room remains after fees, cleaning, utilities, maintenance, reserves, insurance, taxes, realistic management cost, and debt service at rates near the 6.95 percent 30-year average Freddie Mac reported for the week of September 17, 2026. Treat every line as an assumption to replace, not a benchmark to trust.
Is my market saturated?
Saturation is local and measurable, so measure it instead of debating it. Compare active listings and comp-set occupancy in your submarket against the same month last year, then compare your own occupancy against the comp set. If the comp set is flat and you are down, that is an operations problem. If the comp set is down and you are matching it, that is a supply problem, and the answer is differentiation or exit.
Does regulation make a market uninvestable?
Usually the opposite, once you are compliant. Enforced registration regimes like New York City's or licensing regimes like Austin's cap new supply, which protects permitted operators. The danger is buying ahead of a rule change, or buying a unit that cannot qualify. Read the ordinance and the recent council agendas before you sign, and confirm whether the permit transfers with the sale.
Should I use rate automation or set prices myself?
Use rate automation, then override it with local knowledge. Demand-responsive pricing tools react to booking pace and comp-set movement faster than you will manually, but they do not know about the regional event next April or the road closure in July. Let the tool hold the baseline and set your own floors, minimum stays, and event-week rates.
Sources
Airbnb Q2 2026 financial results — platform revenue growth of 17 percent to $3.6 billion, GBV of $27.2 billion, nights and seats booked up 10 percent, and the moderate ADR increase cited above.
Freddie Mac Primary Mortgage Market Survey — the 6.95 percent average 30-year fixed rate for the week of September 17, 2026 used in the debt-service discussion.
Insurance Information Institute: homeowners and renters insurance facts — National Association of Insurance Commissioners data showing the average homeowners premium rising from $1,311 in 2020 to $1,569 in 2022, an 11.2 percent single-year increase and the latest year published.
NYC Mayor's Office of Special Enforcement — the New York City short-term rental registration regime referenced as an enforced permitting example.
City of Austin short-term rental licensing — license requirement by property type, and the July 1, 2026 start of platform removal for unlicensed properties.
IRS Topic No. 415, renting residential and vacation property — the starting reference for rental income and expense treatment mentioned in the margin section.
