- Average Airbnb occupancy over the 12 months to August 2026: Ubud 36.6% at $113 a night, Canggu 36.2% at $214 (AirROI, data as of 08.08.2026). That is the whole market, unmanaged listings included.
- Investor-facing market reviews show different numbers: 10–18% gross, 4–6% net self-managed and 10–15% net professionally managed (Magnum Estate, 03.06.2026). The gap is operations, not magic.
- Annual revenue per listing: Ubud $11,613, Canggu $21,244. Year on year the market fell 19–23% while Ubud supply grew 43% (AirROI, 08.08.2026).
- Foreign arrivals in H1 2026 were about 3.2 million, down 2.42% year on year (BPS Bali, 04.08.2026). This is a mature market, so any model built on ever-rising demand is broken.
- Four numbers decide everything: nightly rate, occupancy, management fee and the full cost list. If any of them comes from a slide rather than a report, the yield has not been verified.
"15–17% a year" is the most common number in Bali villa decks. It is not always a lie: as a gross yield that range exists. The problem is that the buyer hears "this is what I will receive" while the seller means "this is what the property produces before every cost". Here is what counts as realistic in 2026 and how to audit someone else's model in fifteen minutes.
Where 15–17% comes from
A villa model has three levels, and the number falls at each one.
- Gross revenue — nightly rate × occupancy × 365. This is where the attractive percentages live.
- Operating profit — minus utilities, staff, pool and garden, consumables, repairs.
- Owner's net — minus the management fee, platform commissions, taxes and the repair reserve.
Market reviews say it plainly: gross yields of 10–18% in strong areas, net 4–6% self-managed and 10–15% professionally managed (Magnum Estate, 03.06.2026). So 15% net is the ceiling of a professionally run property, not an average expectation.
Real 2026 benchmarks
The most useful check is to compare the promise against whole-market short-let data. Below is AirROI's 12-month view (August 2025 – July 2026), data as of 08.08.2026.
| Metric | Ubud | Canggu |
|---|---|---|
| Average annual occupancy | 36.6% | 36.2% |
| Average nightly rate | $113 | $214 |
| RevPAR | $44 | $79 |
| Annual revenue per listing | $11,613 | $21,244 |
| Active listings | 2,818 | 3,920 |
| Revenue year on year | −19.3% | −23.1% |
One caveat, without which these figures mislead: this is the entire market, including listings rented occasionally, without professional photography, pricing or management. A properly managed villa in a good location performs well above average — but nowhere near 85% year-round.
The average Ubud listing earns $11,613 a year. A $139,000 villa promising 15% would need $20,850 — nearly double the market average. Possible, but it requires evidence rather than a slide.
The second layer is seasonality. Ubud peaks in August at 52.6% occupancy and $123 a night, and bottoms out in February at 31.3% and $99. A model with one blended number for all twelve months hides both the low season and the cash gap inside it.
The third layer is demand itself. Bali received about 3.2 million foreign visitors in H1 2026, down 2.42% year on year (BPS Bali, 04.08.2026), while Ubud supply grew 43%. This is a mature market: it does not grow on its own, and "demand will catch up" is no longer a plan.
Five checks in fifteen minutes
1. Nightly rate. Open three or four comparable villas in the same area with the same bedroom count and look at their calendars and prices for March and August. If the model's rate exceeds neighbours' peak pricing, the model is already wrong.
2. Occupancy. A realistic working band for a managed villa is 50–65% a year. Anything above 70% needs an explanation: long stays, corporate contracts, retreat programmes. "Good marketing" is not an explanation.
3. Management fee. The Bali market runs 15–25% of gross revenue, up to 30% with some full-service operators. Check that it is deducted and what it covers.
4. The cost side. The model needs staff with the thirteenth salary, electricity, water, internet, pool, garden, pest control, insurance, PBB, licences and a repair reserve of 1–3% of value per year. The full reference is in our guide to owner running costs. Fewer than ten lines is not a model.
5. Taxes. Rental tax applies to gross revenue, not profit: 10% for Indonesian tax residents, 20% withholding for non-residents. Plus tax at home.
Red flags in a sales deck
- "Guaranteed yield". A guarantee is a company obligation, not a property feature. Ask what backs it and what happens if it is withdrawn.
- One occupancy figure with no monthly breakdown. It hides the low season.
- Yield with no cost side. The classic move: show gross and call it "yield".
- ROI that ignores the lease term. A villa with 20 years left and one with 50 years left are different assets at the same price — see prices and cost per year of ownership.
- Portfolio-level reports. Ask for per-property statements from the booking platforms, not a summary table in a PDF.
Calculation template. A two-bedroom villa in Ubud at $139,000. Rate $150 a night (above the market median thanks to newness and fit-out), occupancy 55% — that is 200 nights and $30,000 gross. Less 20% management: $6,000. Less $9,000 of annual owner costs. Less 10% rental tax on gross: $3,000. That leaves roughly $12,000, or 8.6% net. At a 5% management fee it works out near 11.8%. Neither number is 17% — and both are honest.
What we check as an agency
DOMA is a real estate agency: we do not build and we do not issue yield guarantees of our own. Our job here is to make the seller's model verifiable — benchmark the quoted rate against open area data, request per-property reports from the developer, lay out the cost side line by line, and restate the result as a net figure after tax and lease term.
The legal half of the same audit is covered in how to vet a developer and villa due diligence.
Bottom line
Bottom line: 15–17% is a gross yield in a good location, not money in the owner's account. The 2026 reference points are 36% market-wide occupancy and 50–65% under professional management, 4–6% and 10–15% net respectively, in a market on a plateau. Check four numbers — rate, occupancy, fee and costs — and insist each one comes from a report rather than a slide.
Model your own scenario with costs and taxes in our calculator — two minutes.
This material is informational and is neither a public offer nor investment advice. Market figures come from open sources as of the publication date and change over time.
FAQ
Are 15–17% annual returns realistic in Bali?
As a gross yield, yes — strong locations do run 10–18%. As a net yield, almost never: after management, utilities, staff, taxes and a repair reserve you are left with 4–6% self-managed and 10–15% professionally managed (Magnum Estate, 03.06.2026). Confusing gross with net is the main source of disappointment.
What is the real occupancy of Bali villas?
Across the whole Airbnb market for the 12 months to August 2026, Ubud averaged 36.6% and Canggu 36.2% (AirROI). Investor decks quote 70–85%, which refers to professionally managed prime properties, not the market. Model 50–65% and check what the seller assumed.
What is usually inflated in a financial model?
Three things: the nightly rate (August peak instead of the annual median), occupancy (85% instead of 50–65%) and costs (no thirteenth salary, no repair reserve, no tax on gross rent).
What does an honest model contain?
Month-by-month seasonality rather than one average; rate and occupancy with a source; the management fee; a full operating cost list; taxes; and the leasehold horizon. Plus a clear split between what is fact from delivered properties and what is forecast.
How do I sanity-check a promised yield fast?
Compare the quoted rate with neighbours' open Airbnb calendars, multiply by 55% occupancy, deduct 20–25% management plus the annual owner costs from our cost guide, then tax. That result is your realistic base. If it is half the promise, the problem is not your arithmetic.