Figures here are illustrative. They describe how something works, not what any place has returned.

Somebody walks into a house on a hillside and knows within a minute that they want it. That reaction is real information — it predicts, quite reliably, that other people will react the same way. It predicts almost nothing about whether owning it will make money.

The two questions run on different variables, and the trouble is that the first is answered instantly and the second takes a fortnight of work. Whichever question is answered first tends to become the conclusion.

What actually determines the return

• What was paid for the land, against what the land is worth without the building on it.
• The cost to build, and how far that ran past the estimate.
• How many nights a year it earns, which is a climate and a distance question before it is a marketing one.
• What the operator takes before anything reaches an owner.
• Debt, and what it costs in the months nothing is earned.
• The reserve, and whether anybody funded one.

None of those is visible from the verandah. Several are actively obscured by it: a building that cost far too much is frequently the most impressive one on the site, because that is where the money went.

Seasonality is the term people underestimate

A place that is extraordinary for five months and unreachable for two has an occupancy ceiling set by weather. That ceiling is not a marketing problem and no amount of photography moves it.

The mistake is to model an annual average. The costs are monthly and the revenue is seasonal, which means the question is not whether the year works — it is whether the reserve carries the months that do not.

Where the money goes before it reaches an owner

In any operated property, revenue passes through a sequence before it becomes a distribution: operating costs, the operator's share, debt, tax, reserve, and only then whatever remains. Each stage is contractual and each is senior to the owner.

An owner evaluating a leisure asset should be able to state that sequence for the specific property, in order, with the percentage at each stage. If nobody can produce it, the position being offered is not understood by the person offering it.

A hundred rupees, followed to the end

The figures below are invented and the shape is not. Every stage is a real contractual claim that exists in most operated leisure assets, ordered as it is actually paid.

• 100 arrives as revenue. This is the number quoted in the brochure.
• 38 leaves as operating cost — people, power, water, supplies, upkeep. It is largely fixed, which is why occupancy matters so much.
• 10 leaves as the operator's fee, usually charged on revenue rather than on profit. Note the order: it is taken whether or not anything is earned below it.
• 6 leaves as sales and distribution — the platforms that fill the calendar take a percentage of what they fill it with.
• 12 leaves as interest, if the asset carries debt at a typical loan-to-value.
• 8 leaves as the reserve for major repair, assuming somebody funded one. Where nobody did, this line reads zero and reappears as a capital call in year seven.
• That leaves 26 before tax, against an asset that may have cost twenty-five times the annual revenue to acquire.

Twenty-six per cent of revenue sounds healthy. Against the acquisition cost it is roughly a one per cent yield before tax, and that is the number that matters, because it is the one being compared against every other place the money could sit.

Move one variable and watch it break. Drop occupancy by a fifth — one bad monsoon, one road closed, one year the flights got expensive. Revenue falls to 80. Operating cost barely moves, because a building costs almost the same to keep whether or not anybody is in it. The operator's fee falls with revenue; interest does not fall at all. What was 26 becomes single digits, and the reserve is the first line anybody proposes to skip.

What the reserve actually is

The eight in that list is the least discussed line and the one that decides how the asset looks in year fifteen. Roofs, plant, pumps, waterproofing and vehicles all have replacement cycles measured in years, and none of them announce themselves.

An asset without a funded reserve is not cheaper to hold. It has deferred a known cost into an unknown year, and the year it lands is disproportionately likely to be a year revenue was already poor — because the same weather that emptied the calendar is what found the roof.

This is the single most useful question to ask of any operated property: what is in the reserve, what is it forecast against, and who decides when it is spent. The answer is usually either precise or absent, and both are informative.

The exit nobody modelled

Almost every projection for a leisure asset ends with a sale, and almost none of them explain who the buyer is.

That matters more here than in most classes. The pool of people who want a specific extraordinary building in a specific location, at the price required to make the model work, is small and does not grow steadily. It is also correlated with exactly the conditions under which a holder might want to sell.

An exit assumption is a claim about a buyer. Written properly it names who they are likely to be, what they would be buying it for, and what they would pay it on — a multiple of earnings, a price per square foot, or a comparable transaction. Written the usual way it is a terminal value in a spreadsheet, arrived at by applying a growth rate to the acquisition price and calling the result a market.

Illiquidity is not automatically a defect

Real assets are slow to sell, and this is usually presented as a cost. Sometimes it is. It is also the reason they are not repriced by sentiment every afternoon, and the reason a holder is not tested on their conviction weekly.

The defect is not illiquidity. It is illiquidity that was not priced — a position entered at a valuation that assumed an exit nobody had modelled.

The test worth applying

Would this be a sound position if it were ugly? If the numbers only work when the building is beautiful, what is being bought is the building, and that is a legitimate purchase — it is simply not an investment, and it should not be underwritten as one.