How Rental Returns Work in Bali Real Estate

5 June 2026 · 8 min · Rental Returns

A yield percentage is the output of a model

Whoever builds the model decides the answer.

That is why the same villa can honestly be advertised at 15% and honestly modeled at 4%. Nobody has to lie. They just have to choose favorable occupancy, quote gross instead of net, count management but not platform commission, ignore tax, and leave out the fact that a leasehold consumes itself.

This article builds the model properly, with every line visible. Swap in your own numbers; the method is the focus.

Start with nights, not percentages

An occupancy percentage is an abstraction. Nights are real, and they are the only thing that generates revenue.

There are 365 of them. Subtract the ones you keep for yourself. What remains is what can be sold, and it can only be sold at rates the market supports in that particular month.

This matters because Bali’s seasonality is severe. A property running near capacity in July may sit under half-full in February. Modeling a single annual occupancy figure against a single average rate hides that entirely, and it almost always hides it in a flattering direction.

Build it by season instead.

Now subtract your own use. Three weeks in shoulder season costs money. At 65% occupancy those nights would have sold about 14 of them, costing roughly $3,220 in foregone revenue.

Adjusted gross revenue: about $56,100 from 216 nights sold.

Note what the blended rate did. Averaging $258 across the year looks unremarkable, but it is only achievable because high season carries it. Anyone quoting you peak rates as though they apply year-round is describing a property that does not exist.

Sanity-check your assumptions before you trust them

Published Bali market data spans a wide range, and the range is the point.

Commentary on well-managed properties in prime areas describes occupancy of 65–75%. Platform data for Canggu in 2026 has been reported at an average nightly rate near USD 214 with average occupancy around 41%, and median annual revenue per listing of roughly USD 25,800 — in a market with over 4,000 listings in that area. The top decile reportedly clears 84% occupancy.

Both pictures are accurate. One describes the best-run properties, the other the middle of a crowded market. Assume you are the median until you have evidence otherwise, and stress-test occupancy down to 60% to see what a bad year does. If the investment only works above 80%, it does not work.

Two more corrections most models miss. First, year one often underperforms, because properties typically take 12–18 months to stabilize pricing, reviews and seasonal patterns. Second, a property without a private pool underperforms on both rate and occupancy, so do not borrow comparables from a different product.

Separate gross from net — completely

This is the point where projections fail. Here is the full cost stack against $56,100 in gross revenue.

Net before tax: $19,765.

Watch the top two lines. Management commonly runs 15–25% of gross and platform commissions add a further 15–20%. That is up to 45% before a single light bulb is replaced.

This is also why the widely quoted benchmark of “operating costs at 35–50% of gross” needs interrogation: many versions of that figure exclude platform commission, or assume management includes housekeeping when your contract may not. Ask precisely what a quoted percentage covers. Two models with identical headline cost ratios can differ by fifteen points of net yield.

Then tax

Rental income is taxable, and the treatment depends on your residency and structure.

A tax resident holding an Indonesian tax number is generally described as paying a 10% final tax on gross rental income. A non-resident earning Indonesian-source rental income is generally subject to withholding at 20%, unless a tax treaty with their home country reduces it.

That distinction is not a footnote.

Note the tax is calculated on gross, not on profit. That makes it heavier than it first appears, and it means a poorly performing year is taxed at the same rate as a good one.

Confirm current treatment with an Indonesian tax adviser rather than relying on any article, including this one.

The two lines nobody puts in the brochure

Setup capital. Furnishing a two-bedroom to rental standard is not included in the purchase price. Budget realistically; call it $32,000 here. Your invested capital is $382,000 instead of $350,000, and every yield figure should be calculated against the larger number. That alone drops our example from 4.04% to 3.70%.

Lease amortization. On a 30-year leasehold, roughly $11,667 of your capital is consumed each year as the remaining term shortens. This is a real economic cost rather than an accounting concept, and it is why a villa with 8 years left on its lease sells for far less than the same villa with 25 years left.

Set that against $14,155 of net income: in cash terms the property earns, but a meaningful share of that is the return of capital you are simultaneously losing to the clock. The investment case then rests on two things the brochure treats as certainties — capital appreciation, and the ability to extend the lease at an affordable price.

Why payback period is the most misleading number in Bali marketing

Payback is usually quoted as purchase price divided by annual income, producing a comfortable-sounding figure.

Run it properly on our example. $382,000 of invested capital against $14,155 net after tax is a 27-year payback on a 30-year lease.

This is not a worst-case scenario. These figures come from average assumptions, competent management and basic arithmetic. It also explains why a handful of specific factors matter so much.

• Direct bookings cut platform commission. Shifting a third of bookings off platforms is worth roughly $2,800 a year here.
• Management terms. The gap between 15% and 25% is over $5,600 annually.
• Format. Apartment residences report operating cost ratios nearer 30%, because shared infrastructure spreads fixed costs — structurally better than a standalone villa on this line.
• Rate, not occupancy. Filling low season by discounting adds nights and costs; raising achievable rate in high season drops almost entirely to the bottom line.
• Lease length. A longer initial term or a fixed-price extension changes the amortization math more than any operational improvement.

That last line is the one to sit with. Operational skill moves the model at the margins; the lease terms move it structurally.

Model three scenarios, always

One number invites false confidence. Build three.

• Conservative — occupancy at 55–60%, rates 10% below assumption, costs 10% above, no appreciation
• Realistic — your best honest estimate, the model above
• Optimistic — strong management, higher direct-booking share, favorable market

Then ask one question: can you live with the conservative case? If the answer is no, the optimistic case is irrelevant. That is the entire discipline.

What to ask before accepting anyone’s projection

Put these to whoever hands you a number.

• Is this gross or net, and net of exactly which costs?
• What occupancy and nightly rate does it assume, month by month?
• Does it include platform commission separately from management?
• Does it include tax, and at which residency rate?
• Does it include furnishing, and is the yield on purchase price or total invested capital?
• Does it account for personal-use days?
• Does it account for lease amortization?
• Is it based on actual performance of comparable properties the operator runs, or is it a projection?

The final question is the most important. Ask for the actual distributions from buildings the operator is running today, rather than a spreadsheet about yours.

The honest summary

Bali rental property can produce solid cash returns. Well-located, well-managed properties genuinely outperform, and the demand base is real.

But the gap between the advertised number and the deposited number is created by the modeling, not by the market. Start with nights. Separate gross from net completely. Put tax, furnishing and lease decay on the page. Then judge what remains.

A property that looks good under those conditions is probably good. A property that needs optimistic assumptions to look good has already told you what it is.

The figures in this article are illustrative, built from published market ranges to demonstrate a method. They are not a projection for any specific property and should not be relied on as one. Actual results vary with occupancy, seasonality, nightly rates, operating costs, management quality, taxes and market conditions, and are not guaranteed. Tax treatment depends on individual residency and structure and changes over time. This article is general information and does not constitute legal, tax, financial or investment advice. Consult a qualified Indonesian tax adviser and an independent notary before committing funds.