Where Villa Pro-Formas Break in 2026: How to Read the Occupancy Assumption

The Signal

In almost any villa financial model, the variable that moves the return the most is not the purchase price or the ADR: it is occupancy. A single percentage point of occupancy flows through to the revenue line almost linearly, and yet it is the figure most often rounded up in sales brochures.

The contrast with 2026 data is uncomfortable. Developer brochures tend to assume occupancy of 65–70%. Airbnb data puts the median for a real listing in Lombok at around 51%. And hotel occupancy in Nusa Tenggara Barat (the province that includes Lombok) was 41.98% in December 2025, according to Statistics Indonesia (BPS). In parallel, in Bali the average ADR has fallen about 14% year-on-year due to supply saturation. None of those three figures is catastrophic, but all three sit well below the number printed on the first page of many models.

This article is not about whether Lombok or Bali perform. It is about how to read the occupancy assumption in a pro-forma before transferring capital, what tests to apply to it, and why the gap between a 70% brochure figure and a 51% market reality can change a purchase thesis entirely.

What Has Changed in 2026

Three blocks of data help calibrate any occupancy assumption this year.

1. Bali Shows the Mechanics of Saturation

Bali's short-term rental market is the laboratory for what happens when supply grows faster than demand. In 2026 there are on the order of 39,000 active Airbnb listings, an increase of around 18% year-on-year. International arrivals also rose, roughly 13.9% year-on-year to about 3.3 million in the first half, but less than supply. The result is measurable: average ADR fell around 14% year-on-year, average monthly revenue per property dropped 6.13%, and average discounting moved from about 15% in 2024 to nearly 19% in 2025.

The relevant point for an investor is not the negative headline, it is the mechanism. When supply runs ahead of demand, the adjustment comes less through occupancy and more through price: owners cut rates and raise discounts to fill nights. A pro-forma that assumes high occupancy and stable ADR at the same time ignores that, in a saturated market, holding one of the two variables tends to cost the other.

2. Enforcement Starts to Separate the Wheat from the Chaff

Since 31 March 2026, Airbnb and Booking.com remove from their platforms any property without valid NIB and Sertifikat Standar documentation in Indonesia. The medium-term effect points to a bifurcation: professionally managed, compliant properties hold occupancy and rate, while generic, unmanaged product suffers rate and night compression. For reading a pro-forma this matters, because the market average occupancy includes both groups. The right question is not "what is the average occupancy", but "which of the two groups will this villa compete in".

3. Lombok Starts from a Lower and More Seasonal Base

Lombok is at an earlier stage of the cycle, with demand more concentrated in specific seasons and events. The median occupancy of an Airbnb listing runs around 51%, with an ADR near IDR 2.8 million, and the real range swings between 40% and 90% depending on zone and seasonality. Gross yields in the Kuta/Mandalika band sit around 12–16% in 2026, down from 14–18% in 2023, and prime-zone net yields at 6–9%: among the highest in Southeast Asia for comparable product, but far from the gross headline. The operational conclusion: in Lombok occupancy is potentially higher at peaks, but less predictable, and that demands a more conservative pro-forma, not a looser one.

How to Read the Occupancy Assumption

A model is read with questions, not with trust. These are the five tests we apply to any occupancy assumption before accepting it.

First: where does the percentage come from? A 70% should trace back to comparable data for the same zone and typology, not to the best month of the villa next door. If the number does not come with a source and a market comparable, it is an aspiration, not an assumption.

Second: is it stabilized occupancy or year-one occupancy? A new villa does not fill from the first month. The ramp to stabilized occupancy usually takes between twelve and twenty-four months, and years 1–2 often run 10 to 15 points below the mature figure. An honest model shows the ramp; a sales model shows only the stabilized year and presents it as the starting point.

Third: is the ADR coherent with that occupancy? Occupancy and rate buy each other. A pro-forma that holds both 68% occupancy and ADR at the top of the range, in a market where average discounting is rising, is counting the same good news twice. Always cross the two variables.

Fourth: is the figure gross or net? The jump from gross to net eats a large part of the headline. From revenue you must subtract the management fee (15% in our model), the PBJT tax (10% in Indonesia), maintenance, platform (OTA) commissions, and real vacancy. A net yield of 7–14% is a reasonable range in Lombok depending on typology and location; any net number above the market gross is a warning sign, not an opportunity.

Fifth: does it survive a sensitivity test? No model should be viewed in a single scenario. The useful question is what happens to the return if occupancy comes in at 45%, 55% and 65%. If the villa only works in the high scenario, it is not an investment: it is a bet that the most favorable assumption holds.

What This Means for the Investor

The practical takeaway is simple: ask for the model, not the headline. A developer who works transparently should be able to hand over the full pro-forma with occupancy broken out by year, the source of the comparable, the gross-to-net breakdown, and at least two sensitivity scenarios. If that documentation does not exist, the problem is not the number: it is that there is no model behind it.

This does not mean distrusting every projection. It means treating occupancy for what it is, an assumption with a range, not a promise with a single point. The investor who internalizes that difference stops comparing villas by the brochure ADR and starts comparing them by the quality of the assumption underneath, which is where the return is actually decided.

The Lombok Angle

In Lombok the temptation to inflate occupancy is greater, precisely because the market is younger and there is less history to check against. That is why our Lendang Luar and Kuta models work with occupancy assumptions in the 45–60% band, not 70%, and present the ramp of the first years rather than assuming the mature scenario from month one. Lombok's ADR is lower than Bali's in absolute terms, but the entry cost is lower too, and the variable that sustains the return in an early phase is not the rate: it is the professional management that keeps occupancy above the unmanaged market median.

Put differently: the bifurcation that regulation is accelerating in Bali is exactly the line on which a return is decided in Lombok. Buying a villa is not the same as buying a managed villa, and that difference shows up in occupancy long before it shows up in any other number.

FAQ

Why does occupancy weigh more than ADR in the return? Because it acts on total nights sold, while ADR acts on the price of each night. In practice, a model is more sensitive to a ten-point change in occupancy than to an equivalent change in rate, especially when a high rate pushes discounting up and reduces nights.

Does a 51% median mean the investment does not work? No. It means the realistic starting point for average market product sits in that zone, and that beating that median depends on management, location and typology. A well-managed villa in a prime zone can operate above it; the error is assuming the high number as a base without justifying how it is reached.

What occupancy is prudent to assume for Lombok? As a working reference, a band of 45–60% for managed product in a good location, with the caveat that the figure is seasonal and less predictable than in Bali. Any assumption above that should come with a concrete comparable and a downside sensitivity scenario.

How do I tell a serious pro-forma from a sales one? A serious model shows the ramp by year, separates gross from net with all costs, cites the source of the comparable, and offers sensitivity. A sales model shows a single stabilized number, net without breakdown, and no adverse scenario.

Risks and Caveats

• Supply saturation risk. The Bali case shows that supply can grow faster than demand and compress both rate and occupancy at once. Lombok is at an earlier stage, but the same mechanism applies if the pace of new supply runs ahead of arrivals.
• Seasonality risk. Lombok occupancy is more volatile than Bali's and depends on specific events and connectivity. A model that assumes flat occupancy all year ignores that risk.
• Execution risk in the ramp. Years 1–2 run below stabilized occupancy. A pro-forma that does not model that start overstates initial cash flow and, with it, the capacity to service any financing.
• Regulatory and management risk. Enforcement of NIB and Sertifikat Standar penalizes unmanaged or non-compliant product. A villa that depends on informal channels to fill nights faces an additional occupancy risk that rarely appears in the brochure.

Figures and references cited in this article are based on public sources as of 6 July 2026. Yield and occupancy ranges are estimates based on market data and our own models; actual results may vary. International real estate investment carries market, regulatory, currency and operational risk. This content is informational and does not constitute legal, tax or investment advice.

Further Reading

• Lombok Villa Rental Yield Analysis 2026: the breakdown of gross yield, net yield and ROI that this framework builds on.
• Bali's Supply Correction: What It Means for Lombok Yields: how Balinese saturation reshapes the return equation in Lombok.
• Indonesia Hits a Tourism Record in 2026: What the Demand Data Tells a Villa Investor: the demand side behind every occupancy assumption.

Sources Consulted

• Statistics Indonesia (BPS): Nusa Tenggara Barat hotel occupancy, December 2025.
• Airbnb market data: median occupancy and ADR of listings in Lombok, 2026.
• Bali short-term rental market analysis: listing growth, ADR and year-on-year discounting, 2025–2026.
• Indonesian regulation: NIB and Sertifikat Standar, OTA platform delisting enforcement from 31 March 2026.
• Internal Lombok International Development models: Lendang Luar and Kuta, occupancy assumptions and gross-to-net breakdown.