Owner Nights Against Rental Nights: The Real Cost

The Subtraction Nobody Underwrites#

Every rental-pool projection in Bali and the Gilis starts from the same number: nights available times occupancy times average daily rate. That arithmetic describes a hotel room, not a residence with an owner attached. A hotel room has no one who can walk in, take the key, and remove a week from the denominator without asking the general manager. A branded residence does. The owner's contract grants exactly that right, and the projection that ignores it is not conservative, it is incomplete.

The gap is not academic. Published guidance puts personal-use allowances anywhere from two to four weeks a year, the example a law-firm review of branded residences gives, to 30 to 60 days or more, the range a branded-residence marketplace describes, with some agreements capping use and others paying the owner only for the days the unit sits in the pool. Either mechanism produces the same fact on the ground: nights that would have carried a paying guest instead carry the person who signed the purchase and sale agreement. The revenue line for those nights is not smaller. It is zero, then it is backfilled from somewhere else in the model, usually the assumption column nobody stress-tests.

None of this is unique to Bali or the Gilis. It is unique to any real estate product that bundles a personal-use right into an income-producing asset, from condo-hotels in Miami to fractional chalets in the Alps. What differs by market is the price of the mistake. In a market where a villa operator advises booking the Christmas-to-New-Year fortnight six to nine months ahead, on minimum stays of up to seven nights, the cost of an owner treating the calendar as a personal vacation home rather than a shared income asset is higher than in a market with a flatter demand curve. Bali's monthly curve is flatter than that framing suggests: AirDNA scores the island 87 out of 100 on seasonality, where a higher score means a smaller gap between the highest- and lowest-revenue months, benchmarked against other Indonesian markets. Outside the festive fortnight, which monthly averages cannot isolate, the timing of an owner's nights therefore matters less in Bali than in a sharply seasonal market.

What a Rental Pool Agreement Actually Promises#

A rental pool agreement is usually a revenue- or profit-sharing contract, and unless it carries a fixed rent or a guarantee it promises no income. A law-firm review of branded residences describes it as attached to the unit's sale and purchase agreement and notes that the programme can pay a fixed rent, a variable rent based on gross revenue or net operating profit, or a combination of both. Where developers guarantee returns, the same review adds, the guarantee is usually limited to a ramp-up stabilisation period, and where foreign investors cannot acquire ownership title and hold a long lease instead, the agreement becomes a sub-lease by nature. Where the rent is variable, the central clause is the split. HVS describes the tariff of a rented unit in a resort ownership programme as typically split in the range of 50/50 between the resort owner and the unit owner, and a branded-residence marketplace puts the owner's share at 40% to 60%, with the remainder covering operating costs, marketing, reservations, housekeeping, maintenance and operator fees. The owner receives an income stream that looks passive. It is passive only in the sense that the owner is not making beds. Every other variable, including how many nights the owner personally withdraws, remains the owner's decision.

Some agreements also define which weeks are off-limits to the owner, and that clause can matter more than the headline allowance. A branded-residence marketplace notes that mandatory rental programmes require owners to make their residences available for rental for specified periods; where those periods cover the highest-demand holiday windows, they protect the pool in the weeks when guest demand and owner demand for the same unit peak together. Where no such clause exists, nothing forces the owner's personal nights into the low season, and the owner is free to book the week the market would have paid the most for, subject only to availability and any cap on total annual nights. HVS, in an illustrative case study of a resort whose owners take 25 nights a year, assumed most owners would stay in the shoulder and off seasons to protect their own returns. That is a behavioural assumption, not a contractual one.

Where the Calendar Actually Breaks#

Peak season in Bali is not evenly distributed, and neither is the cost of an owner-use night. A Bali property-buying guide reports two high seasons, July-August and December-January, and BPS counts of foreign arrivals to Indonesia, as charted by Kontan, peaked in July-August and in December in 2025, with August the busiest month. Measured short-term rental performance draws a different map: AirROI places the peak Airbnb season for Uluwatu, Seminyak and both Gili islands in August, September and October, and lists December among Uluwatu's three softest months. Within the Christmas-to-New-Year window specifically, a villa operator describes roughly 20 December to 5 January as the most compressed demand spike of the Bali year and advises booking a quality private villa six to nine months ahead, and villa and travel sources report minimum stays of three to seven nights over the same dates. Rates rise across the market for those dates, but no villa source reviewed for this article publishes a multiple over baseline. A travel-cost guide puts the Christmas-to-New-Year uplift in hotel rates at a third or more above baseline, and cites a published comparison in which private two-bedroom villas cost about 60% more in high season than in low season, a figure it offers as direction rather than a quote. That is not a soft peak. It is a demand spike compressed into roughly two weeks in which the well-run properties fill first.

An owner-use night taken inside that window does not compete with an average night. It competes with some of the highest-value nights the calendar produces all year. Thirty nights withdrawn in February, when AirROI's Uluwatu series for the twelve months to July 2026 shows the lowest average nightly rate and the lowest monthly revenue of the year, cost the rental pool less than the same thirty nights taken across the festive fortnight and August. The number of nights withdrawn is the metric every buyer asks about. The timing of those nights is the metric that actually moves the yield, and it is rarely disclosed in the same paragraph as the usage allowance.

The Bali-Gili Demand Curve, in Real Numbers#

The market underneath this calculation is measurable, and the measurements are not uniform across the archipelago. A Bali villa vendor's analysis of AirDNA 2025 data for professionally managed listings, published by PARADYSE Homes, benchmarks Uluwatu villas at an average daily rate near $288 and occupancy near 66%, and Seminyak-Kerobokan at a higher rate on fewer nights, near $307 against occupancy closer to 58%. Those figures describe the managed top of the market. AirROI's series for all Airbnb listings over the twelve months to July 2026 puts Uluwatu at $244 and 46.2% occupancy, and Seminyak at $278 and 38.0%. The distance between the two sets of figures is itself the first underwriting question, since a projection built on the managed benchmark assumes the scheme performs at the top of its market.

The Gili Islands complicate the picture further because the measured market there is small and mixed. AirROI tracks 110 active listings on Gili Trawangan and 69 on Gili Air for the twelve months to July 2026, at average nightly rates of $149 and $163 and occupancy of 34.6% and 37.7%, and describes both as boutique markets. Those are samples of the whole short-term rental stock, not a benchmark for a branded scheme, and Victaura Research found no third-party series that isolates professionally managed or branded inventory on the islands. An allocator underwriting a Gili project is, by necessity, underwriting on a Bali proxy and a thinner set of local signals, which is itself a factor that should widen any return range quoted, not narrow it.

Demand at the country level is recovering rather than surging. BPS-Statistics Indonesia recorded 15.39 million foreign visitor arrivals in 2025, up 10.8% on 2024 and still below the 16.1 million of 2019, at about 95% of the pre-pandemic volume. That is a market climbing back to a prior baseline, not one compounding past it. A rental projection built on continued double-digit demand growth is extrapolating a recovery, not a trend, and the distinction matters most in exactly the peak weeks where owner-use decisions do the most damage.

Thirty Nights, Not Thirty Days#

The question here is narrow: what happens to the declared yield when thirty nights come out of high season. Treat "declared yield" as gross rental revenue before the operator's split, calculated against a fixed unit and a fixed year, so the effect of timing is isolated from every other variable. Using the managed Uluwatu benchmark above as a stand-in for a professionally managed villa product, an unencumbered calendar produces roughly 241 rental nights across the year at a blended $288 average daily rate, about $69,400 of gross rental revenue. That is an illustrative construction built on a vendor's reading of AirDNA data, not a specific project's audited numbers, and it should be read that way throughout this section.

Move the thirty withdrawn nights through three points on the calendar and the revenue impact is not linear. Placed inside the Christmas-to-New-Year fortnight and July-August, and assuming the villa would otherwise have been full at a premium of 30% to 35% over the annual blend, the nights would all have rented, and at an elevated rate. Both inputs are Victaura Research assumptions for an upper case, not measurements. Placed in the shoulder months, where occupancy is assumed to track the 66% average, only a portion of those nights would have found a paying guest even absent the owner. Placed in February and March, where the model assumes weaker demand, the marginal cost of the same thirty nights falls further; AirROI's Uluwatu series shows February as the lowest-rate month of the year, but not a low-occupancy one: at 51.0%, its occupancy sits mid-table among the twelve months, above December and July. The label on the brochure, "thirty owner nights," is identical in all three cases. The number that reaches the owner's account at year end is not.

Modelling the Subtraction#

The table below turns that logic into a single illustrative comparison. Its baseline comes from a published third-party benchmark, not from any specific development's rent roll, and its seasonal inputs are Victaura Research assumptions, constructed to show direction and magnitude, not to forecast an actual return. The method: hold the unit, the year, and the thirty withdrawn nights constant, and vary only where on the calendar those nights fall. The revenue-impact column is the estimated gross rental revenue given up as a share of an unencumbered-year baseline built on the same benchmark, and the source line under the table separates the measured input from the assumed ones.

Who Bears the Cost of a Pooled Calendar#

Many branded residence programmes pool revenue across the property rather than unit by unit, and the accounting period of the pool decides who pays for an owner's peak-season absence. Where a rental pool can count as a collective investment scheme, a law-firm review notes, developers may instead need to account for each unit separately. In the rental pool distribution clause that a local real estate agent reproduces for Whistler's Phase 2 hotel condominiums, each owner is entitled, for each day the unit sits in the rental pool, to a share of that day's room revenue whether or not that specific unit was rented. Shared day by day, a unit withdrawn on a sold-out night simply drops out of that night's denominator: its owner forgoes the night's share, the operator loses its split on the displaced booking, and the other owners' per-unit share broadly holds. Pooled over a month or a year and shared by days in the pool, the same withdrawal removes a high-value night but only an average day from the denominator, so every other owner absorbs a fractional share of the difference, spread thin enough that no single co-owner notices, but real enough to show up in the aggregate distribution at year end.

This is the argument for asking an operator, before signing, over what period the pool is shared and how owner-use nights are distributed across it, not just within one unit. In a pool shared over the year, a handful of owners who treat their units as personal holiday homes during the highest-demand weeks create a structurally different risk for everyone else than owners whose personal use clusters in the shoulder and low season. Even detailed owner statements may not show it: a Whistler agent notes that the most detailed hotel-condominium statements there give no year-end summary of owner use, so the monthly statements have to be requested, and Victaura Research found no publicly disclosed, project-level breakdown of aggregate owner-use nights by month for any Bali or Gili Islands branded scheme during this research. That absence is disclosed here rather than papered over, and it belongs on the list of questions the allocator raises directly with the operator rather than assumes.

The Industry's Usage Cap Describes a Right, Not a Habit#

Branded residence programmes cap or price personal use in different ways, but a cap is a ceiling, not a forecast. Published guidance runs from two to four weeks a year in a law-firm example to 30 to 60 days or more in a branded-residence marketplace, with the specific number set project by project, and where a pool is shared by days in the pool, as in the Whistler clause, an owner earns no share for the days the unit is out of it. The same law-firm review notes that the formula for charging operating costs back to owners needs to account for how long, and in which season, the owner used the unit, the one mechanism in the guidance reviewed here that prices timing. None of that documentation predicts when inside the year the owner will actually show up. A cap that reads "up to twenty-eight nights" is silent on whether those nights land in the second week of January or the last week of December, and in a pool shared over the year the two produce very different outcomes for every other owner.

A comparable industry offers a partial check on how owners actually behave once they hold a usage right. In the timeshare sector, a model built entirely around scheduled personal use rather than passive income, the American Resort Development Association's 2024 United States Owners Report finds that owners spent an average of 6.5 nights at the timeshare resort on their most recent timeshare vacation, and that those vacations are spread fairly evenly across the year, with June and July the most popular months. That is survey evidence from an adjacent structure, not proof of behaviour inside a branded-residence rental pool, and Victaura Research is treating it as directional context rather than a substitute for project-level data. It does suggest owners with a usage right tend to take it in blocks long enough to matter, and that some of those blocks land in the summer peak.

What the Brand Actually Buys, and What It Does Not Fix#

A recognised operating brand can lift occupancy and rate in tourism-dependent markets, but verifiable public measures of that lift in resort Asia are scarce. A branded-residence marketplace attributes the lift to global reservation systems, loyalty programmes and revenue management that an individual owner cannot replicate. An occupancy range for branded Phuket schemes in hotel rental programmes circulates in trade coverage attributed to Knight Frank; Victaura Research could not trace it to a Knight Frank publication, including the Residence Report 2025/26, and does not use it. The brand's distribution system, loyalty base and quality standard plausibly do real commercial work; what the public record lacks is a clean measure of how much of it reaches the owner's occupancy.

None of that changes the arithmetic of an owner-use night. A brand that fills 90% of the calendar with paying guests still loses the same marginal night when the owner exercises a usage right during the highest-demand week. The brand raises the ceiling on what the rental pool can earn. It does not, and structurally cannot, prevent the pool's participants from removing inventory from that ceiling whenever their contract lets them. Knight Frank's survey of September 2026 counts 903 branded residence schemes at the end of 2025, up from 354 in 2015, and expects the sector to pass 1,000 during 2026 and to reach about 1,800 schemes by 2031. That growth is a story about brand economics and developer fees. It is a separate story from whether any individual buyer's usage habit erodes the yield the brochure quoted.

Reading a Rental Pool Agreement Like an Allocator#

The document that matters is the rental pool agreement, not the sales brochure, and it rewards a specific reading order. Locate the revenue split first, since every subsequent number in the projection is a fraction of that split, not the gross figure most marketing materials lead with. Locate the personal-use clause second, and read it for two things separately: the total annual nights allowed, and whether any of those nights are restricted from the blackout or peak-demand calendar. An allowance with no peak-season restriction is a materially different economic instrument than an identical night count confined to shoulder and low season.

Then find the mechanism, if any, that prices heavy personal use. Some agreements simply cap total nights. Others let the owner use the unit as often as desired but reduce that owner's share of pooled income proportionally, which aligns incentives better but requires the allocator to model their own likely usage honestly rather than assuming a conservative one. Ask, specifically, whether the operator discloses aggregate owner-use nights across the pool historically, by month. Few operators volunteer this. Its absence is not evidence of a problem. It is evidence that the number has not yet been asked for, and an allocator underwriting a specific unit should ask before signing, not after the first December statement arrives short.

One more clause deserves a direct question: force majeure and closure risk. Bali and the Gilis carry event risk beyond ordinary seasonality, from volcanic ash closures at Bali's airport to boat-access disruption on the smaller islands during rough-weather periods, and a rental pool agreement may or may not specify how lost nights from such closures are treated relative to owner-use nights in the same accounting period. An agreement silent on the interaction between force majeure days and personal-use allowances leaves room for a dispute precisely when the numbers are already under pressure. Raising it before signing costs nothing. Raising it after a closed airport and a depleted owner-use allowance collide in the same quarter costs the allocator leverage they no longer have.

Honestly Disclosed: What This Analysis Cannot Verify#

The illustrative model in this article rests on published market benchmarks, and every benchmark has a boundary the reader should know. AirDNA's and AirROI's figures are drawn from short-term-rental platform listings, AirDNA naming Airbnb, Vrbo and Booking.com, a channel that overlaps with, but is not identical to, branded residence rental pools, which often run through direct booking, an operator's own website, and offline travel agent networks alongside or instead of the major platforms. The $288 and 66% baseline is a villa vendor's reading of AirDNA data for professionally managed listings, not AirDNA's own published figure and not a specific branded project, and it sits well above AirROI's Uluwatu average for all Airbnb listings. AirROI's Uluwatu series rests on 89 active listings and its Gili Islands series on about 180 across two islands, and neither isolates managed or branded stock. Those gaps are disclosed rather than filled with an estimate dressed as a fact.

The thirty-night model itself is a construction, not a measurement. It assumes a single unit, a single year, and a clean substitution of owner nights for rental nights, none of which holds precisely in an operating rental pool where cancellations, minimum-stay rules, and pooled-inventory allocation across many units complicate the picture. Run on AirROI's measured monthly series for Uluwatu for the twelve months to July 2026, thirty nights at August's occupancy and rate cost about 11% of a year's gross rental revenue and thirty nights at February's about 7%, a narrower gap than the table's assumption-driven 16% and 5%. The same series does not support the table's calendar beyond August: thirty nights at December's or July's averages cost about 7%, no more than February, while September and October, placed in the table's shoulder, cost about 10%. The series rests on 89 listings and a single year, and monthly averages cannot isolate the festive fortnight, which is where the table's upper case sits. The ARDA timeshare comparison sits in an adjacent industry with a different ownership structure and should be read as context for how owners with a usage right tend to behave, not as a data point transferable to branded residences. Beyond the labelled assumptions in the table, where a number could not be sourced to a named institution, a named report, and a stated year, it was left out of this article rather than estimated.

Closing#

The principal underwriting a branded residence in Bali or the Gilis is underwriting two calendars at once, and only one of them is priced in the brochure. The rental calendar is the one every projection shows: nights, rate, occupancy, split. The usage calendar, the weeks the owner or their family and guests actually occupy the unit, is contractual, personal, and almost never modelled against the specific weeks it is most likely to collide with peak demand. Closing that gap does not require a better brochure. It requires a rental pool agreement read for its blackout calendar and its usage-pricing mechanism before the purchase, and an honest personal forecast of when, not just how often, the unit will actually be used.

Skin in the game disclosure. Victaura, through its parent Greystone B.V. (Netherlands), holds an active operating position in prime resort property. Readers should assume commentary may be influenced by, or benefit, Greystone's position. This document is classified as marketing material under MiFID II Article 24(3). It is not investment advice.