How every figure is computed

Definitions

Recorded price

The recorded price is the median price per square foot of registered ready sales in the trailing 12 months, requiring at least five observations. It uses the same cutoff as rent and yield. If that year has too few sales, no recorded-price benchmark is published.

Ready units only. A buyer comparing what to pay for a finished apartment is not helped by a launch price agreed years ago on a payment plan, so off-plan sales are excluded here — unlike the gross yield denominator, which pools both because a yield needs the broadest price base it can defend.

Five ready sales is the floor for this recorded-price benchmark; other metrics state their own evidence floors. A price computed from two sales moves by a third when one penthouse trades, and publishing that as guidance would be the most damaging thing this index could do. Floor, view, condition and layout all move a real price and none of them is in this figure.

Gross yield

Gross yield divides the median annual rent from new registered Ejari contracts starting in the last 12 months by the median price of the building's registered sales in the last 12 months (at least 3 sales) for the same bedroom type, with ready and off-plan sales counted alike.

The rent leg is the median of NEW registered Ejari contracts starting in the last 12 months for the same building and bedroom type. Renewals are excluded to keep new and renewing tenancies as separate cohorts. At least 5 contracts must sit under it (20 for a villa community) or no yield is published. A contract counts for the building when its registration names it, or, for the many contracts filed with no project name, when its floor area matches exactly one project's registered units and that project holds only this building. Checked against contracts that do carry a name, that match is right 99.7% of the time.

The price leg is the median of the building's registered sales in the same trailing 12 months, requiring at least 3. There is no fallback to older sales. Both legs end on the same data cutoff; future-starting contracts are excluded. Ready and off-plan sales count alike. The period, sample count and latest sale date are printed beside the yield.

The same price benchmark is the denominator of the modelled short-term yield shown beside it, so the two returns are comparable.

“Gross” means before every cost: service charge, agency fees, maintenance and vacant months all come out of it in practice.

Gross yield, all unit types

The all-unit-types yield pools every bedroom type in the building: the median new Ejari rent across all contracts over the median registered sale price across all sales, in the same 12 months.

The rent median and the price median can rest on different unit mixes — a year where mostly studios let and mostly 2-beds sold pulls the ratio one way — so the pooled yield is not an average of the per-type yields and can sit above or below all of them. Where a per-type yield exists it is the figure to compare; the pooled one is kept for rankings and cross-area tables, where one row per building is needed.

Yield spread

Yield spread is the building's gross yield minus the median gross yield of its area for the same bedroom type, in percentage points.

Points, never percent: “1.4 points above the area” is a different statement from “20% above the area”, and only the first is what this measures. The spread is not shown where the building's own yield is suppressed — a spread quoted against a broken denominator would inherit the problem.

New-tenant rent premium

The new-tenant rent premium compares the median rent on new registered contracts against the median on renewals in the same building, over the same 12-month window.

The groups may contain different homes. Their median difference is not an individual tenancy's increase, a guaranteed gain on turnover or the permitted increase under the official rental index. It is only shown where the gap is material (at least 2%), because Dubai rents cluster on round numbers and tiny gaps are noise.

Short-term premium

The premium divides modelled net short-term revenue (nightly rate × 365 × assumed occupancy, minus platform fee, cleaning and utilities) by the median registered long-let rent for the same building and bedroom type.

Above 1×, the nightly let earns more than the lease; below 1× it earns less. Rankings and hub medians exclude rows below 1×, so those figures describe buildings where short-term renting beats the lease — not all buildings.

Nightly rate per sqft

Nightly rate per square foot divides the median nightly rate by the unit's measured floor area (derived from the building's own sales where possible, else the Dubai median area for that bedroom count).

It exists for comparison: absolute nightly rates mostly measure unit size, and per-sqft strips that out.

Net short-term revenue

From the nightly rate a guest sees, net revenue deducts the platform fee, then cleaning (2 turnovers a month) and utilities, both scaled to the unit's measured floor area.

Service charge and mortgage are deliberately not deducted: neither comes out of the long-let rent it sits beside, and a service charge is owed whichever way the unit is let, so removing it from one side only would break the comparison. Neither figure is net of income tax or furnishing.

Occupancy assumption

No Dubai source publishes per-building occupancy — the Land Department registers contracts, not nights — so annual short-term figures are modelled at a single stated 70% occupancy.

The model uses nightly rate × 365 × occupancy, deducts proportional platform fees, then fixed annual cleaning and utility costs. Changing occupancy requires recalculating net revenue and yield; halving occupancy does not halve either net figure. The assumption is not evidence of occupancy achieved by this building.

Management fee

The cost model charges no management fee, matching a self-managed unit. Professional short-term management in Dubai typically charges 15–25% of revenue after the platform fee; deduct it from revenue after the platform fee, then subtract fixed cleaning and utility costs and recalculate yield.

500m rows

Airbnb publishes approximate coordinates, so even the tightest set — listings within 50 metres — is the building and its immediate neighbours. Where that set is too small to measure a bedroom type, the row widens to the surrounding 500-metre cluster and is marked “500m”.

Ejari

Ejari is the Dubai government registry where every tenancy contract must be filed. Rents on this site come from those registered contracts, which makes them agreed prices rather than listings — and lets new contracts be separated from renewals, which is what the turnover-upside metric measures.

Off-plan price gain

This compares the median price per sq ft of every registered off-plan sale in a building with the median price per sq ft of its ready resales in the last 12 months. Per sq ft, so a launch of studios is not compared with resales of larger flats. The two groups are different units and different buyers, not one buyer's purchase and sale. The 4% DLD transfer fee, agency fees and financing are not deducted, so it is not a realised investor return.

The years shown are the gap between the two groups' median sale dates, not one owner's holding period. Only buildings with both groups appear, so the list leans to buildings that resold.

Median of registered records

Prices and rents on this site are medians of registered records: Dubai Land Department sale transactions and Ejari rental contracts. Asking prices on listing portals run above what transacts and include stock that never sells; registered records are what was actually paid.

Gross payback

Gross payback is 100 divided by the gross yield: a 7% yield pays the price back in roughly 14 years of rent, before any costs, void periods or price movement. It is a way to feel the size of a yield, not a projection.

Break-even occupancy

Break-even occupancy inverts the short-term cost model against the building's registered long-let rent: the share of nights that must be booked for the nightly let to match the lease, after platform fee, cleaning and utilities.

Cleaning is modelled as a fixed monthly cost, so the figure is approximate — read it as a threshold, not a precise operating target. Where the answer exceeds 100%, no figure is shown: the lease wins at any occupancy.

Villa community

A villa has no building in the Land Department's transaction records: measured on the 2026-08 export, `building_name_en` is blank on every one of 186,128 villa sales. What the records do carry is the master project (73.9% of sales), and that is the name people search. So villa figures are published per community, keyed on the same locality table the apartment area hubs use, with the community's name taken from the property portals that rank for it and the Land Department's own registration kept as an alias.

Where the Land Department leaves the community blank, the project name usually carries a launch phase — "DAMAC ISLANDS 2 - BAHAMAS 1" — and a curated fold table maps those phases to the community they belong to. Emirates Living is one registration carrying three places: its Springs phases publish as The Springs, its Emirates Hills project as Emirates Hills, and the residual — larger plots at a higher price — as The Meadows.

A community publishes once it holds 3 registered sales (or, for the rent page, 3 new contracts) in the trailing 12 months, the same floor as a tower; its yield needs 20 contracts, since five spread across a thousand-villa community are not evidence about that community. The headline row counts every registered record, including the 14–18% of recent villa sales filed without a bedroom count; the per-bedroom rows count only those with one, so they never sum to the headline, and every page says so.

Title type (apartment or villa)

The apartment index is every record the Land Department types as `Unit` and Ejari types as `Flat`. The villa index is every record typed `Villa`, plus Ejari's `Complex Villas` label for villas inside a managed compound. Nothing is classified on its name, and no record is in both.

The split follows the title because that is the only classification the register itself makes, and it is the one that governs what a buyer actually acquires. It also decides the grain each index can publish at: a unit title names a building, so apartments publish per building; a villa title names a plot, so villas publish per community.

Townhouses fall on both sides. A plot-titled townhouse sits inside its community's villa figures. A stacked or strata-titled townhouse is registered as a unit and appears in the apartment index under its own name — which is why the Dubai South apartment tables carry rows called "The Pulse Townhouses Cluster 35" and "Urbana III Stacked House". Measured on the 2026-08 Ejari export, contracts in townhouse- or stacked-house-named projects split 9,891 filed as Villa against 929 filed as Flat: the second group is real apartment-index stock, not a mapping error.

Read a row from the index it appears in, and carry the title type with the figure. Do not reclassify a row by reading its name, and do not treat a townhouse-named row in an apartment table as evidence the index is wrong about it.

Rent grain (villas)

Ejari villa contracts name their community less often than sales do — 49% carry a master project, and the 58,891 leases filed as "Complex Villas" almost never do. So villa rent is attributed in three rungs and each page states which one its yield rests on. `community`: the contract's registration resolves to this community. `project`: the registration has no community but its project name does, through the same fold table the sales use. `vicinity`: contracts across the community's Land Department area, pooled in only where the first two rungs hold fewer than 20 new contracts, and deduplicated against them.

The label is decided by majority: a community whose 12-month new contracts are mostly pooled is labelled `vicinity`, and its yield should be read as the area's rather than the community's. Measured on the 2026-08 export, 90 of the 125 sale-eligible community rows rest on their own contracts, 9 on project-name matches and 23 on the area.

Why similar gross yields can share the same rent evidence

Worked evidence edition captured 5 September 2026, using registered data to 2 September 2026. These two Al Barari one-bedroom records were selected to explain source grain; they are not a representative market sample or recommendations.

B2 one-bedroom record: AED 150,000 median new annual rent ÷ AED 1,796,178 median registered sale price × 100 = 8.35% gross. The rent benchmark uses 20 project-level new Ejari contracts and the price uses 25 building-specific sales, in the same trailing 12 months to 2026-09-02. Ready and off-plan sales count in the price denominator.

C1 one-bedroom record: AED 150,000 median new annual rent ÷ AED 1,800,482 median registered sale price × 100 = 8.33% gross. The rent benchmark uses 20 project-level new Ejari contracts and the price uses 25 building-specific sales, in the same trailing 12 months to 2026-09-02. Ready and off-plan sales count in the price denominator.

Both rows reuse the same project-level rental benchmark. They do not establish two independent tower rents or 40 distinct tenancies. The small yield difference comes from the sale-price denominator. Check a building's source label before interpreting its position against a neighbour.

Cost scenario, not observed charges: if annual operating costs were AED 20,000, B2's rent after those costs would be AED 130,000, or 7.24% of that price. This scenario excludes financing, acquisition and exit costs, tax, and vacancy unless included in the chosen cost. It is not this building's measured net yield.

The dated evidence extract retains the input figures, source URLs, original response hashes and selection rule. Current record pages can change after refresh; this example's edition stays fixed. SHA-256: b796aaaac195d7dfa859b7decf5b0d0a211070ed753e7fa1bc98da655f6d3788. A correction requires a new dated edition.

• Gross yield definition — numerator, denominator, grain and exclusions.
• Building-specific yield ranking — uses each building's own rent evidence; pooled examples above need not qualify.
• Al Barari rental benchmarks — compare contract medians and identify shared rental evidence.
• B2 rental record and C1 rental record — compare the same bedroom and read the project-source caveat.

Commercial property

Service charges and net yield

Every yield on this site is gross: annual rent over price, before the service charge, transfer and agency fees, vacancy and finance.

The Land Department's owners-association budgets are public and current: the 22 September 2026 export holds 2026 budgets for 1,994 residential property groups. The export states no unit, so we tested it. A group's total does not grow with its project's floor area, so its categories are rates per square foot a year, not budgets in dirhams. Parking and meter lines are flat annual amounts, and the income line is left out because the export does not say which way it runs, so the charge we use errs high.

Where the Land Department's own project keys tie one budget to a building, the building page shows the yield after that charge: the charge per square foot over the sale price per square foot, taken off the gross yield, with the budget year beside it. No budget is matched to a building by its name. The figure is still before agency fees, maintenance, vacancy and finance, so it is not a net return.

Market signals: supply against leasing, and same-building moves

A building's rent uses only the leases that can be tied to that building, by its name or its registered floor areas. Most registered leases cannot: of 538,473 apartment leases in the 12 months to 5 October 2026, 90,308 could be tied to one. Area-level leasing is therefore counted from every registered lease in the area, placed by its Land Department area and master project the same way buildings are. About a third of Dubai's apartment leases are in older districts with no freehold sales, such as Karama and Deira; they count toward Dubai's total and toward no area page.

Planned completions are apartment units in projects the Land Department register lists as neither complete nor cancelled, by the year of the registered completion date. Projects already past that date and projects without one are counted separately, not spread over future years. Leases are every registered new and renewed apartment lease in the area in the last 12 months. A planned date is not a delivery forecast. The ratio is withheld below 100 leases a year.

Each building counts once per bedroom type when it has at least five ready sales (or five new leases) in both the last 12 months and the 12 before. Its change is its own median price per sqft (or annual rent) now against then, so the mix of buildings that happened to trade cannot move it. Off-plan sales are left out. The pooled median uses every ready sale (or new lease) in the place, with at least 20 in each year. Where the pooled change and the typical building's change differ by 3 points or more, the page says so: the difference is which buildings traded, not what happened to prices.

How the building summary reads each indicator

A building page opens on a summary that reads each indicator as Strength, Fair or Watch. Each reading comes from a fixed threshold, never from judgement, and sits beside the figure it reads. A row with too little evidence shows the gap and no reading.

Returns

Gross yield at least 1 point above the Dubai median reads Strength; 1 point or more below reads Watch.

Fees repaid

Purchase fees repaid by a year or less of net rent read Strength; more than two years reads Watch.

Rent

New-lease rent up 3% or more on the previous 12 months reads Strength; down 3% or more reads Watch.

Price

Median price per sqft up 3% or more on the previous 12 months reads Strength; down 3% or more reads Watch.

Letting

Annual leases covering 60% or more of registered units in 12 months read Strength; under 35% reads Watch.

Liquidity

Sales covering 12% or more of registered units in 12 months read Strength; under 5% reads Watch.

Supply

Homes due in the area over three years under 25% of the homes let there read Strength; 50% or more reads Watch.

Earlier buyers

Context only: what earlier buyers paid against today. Never a verdict.

Model constants

The short-term cost model runs on a handful of stated constants. They are listed here with their live values — the page reads them from the model itself, so it cannot drift from what the site computes.

Assumed occupancy

Share of nights assumed booked when a nightly rate is annualised — an assumption, not a measurement; revenue scales linearly with it.

70%

Platform fee

Airbnb's share of the guest-facing booking price, deducted before any other cost.

15.5%

Management fee

The model assumes a self-managed unit; professional management typically takes 15–25% of revenue after the platform fee.

0%

Cleanings per month

Turnover cleans charged per month, fixed regardless of occupancy.

2

Cleaning fee

Reference cost for a one-bedroom, scaled to each unit's measured floor area.

AED 200 per clean

Utilities

Reference cost for a one-bedroom, scaled to each unit's measured floor area.

AED 700 per month

Turnover-upside floor

Minimum new-vs-renewal gap shown; Dubai rents cluster on round numbers, so smaller gaps are noise.

2%

Questions

What does “Recorded price” mean on Dubai Wealth Index?

What this building actually traded at recently: the median price per sqft over the trailing 12 months, requiring at least five registered sales of ready units. The window is always stated, because a price without one is not evidence.

What does “Gross yield” mean on Dubai Wealth Index?

A year of rent as a share of the purchase price, before any costs. AED 70,000 rent on an AED 1M apartment is a 7% gross yield.

What does “Gross yield, all unit types” mean on Dubai Wealth Index?

The building's yield pooled across every bedroom type: median new rent over median sale price, both taken across all unit sizes. Because the two medians can come from different mixes of units, it can sit outside the range of the per-type yields.

What does “Yield spread” mean on Dubai Wealth Index?

How far this building's yield sits above or below its area's median for the same unit type, in percentage points. Above zero earns more than the typical neighbour.

What does “New-tenant rent premium” mean on Dubai Wealth Index?

The difference between new-contract and renewal-contract median rents at the published grain. Different homes can appear in each group; this is not an individual permitted increase.

What does “Short-term premium” mean on Dubai Wealth Index?

Modelled net nightly revenue divided by the registered long-let benchmark at the same bedroom and published source grain. A 1.5× ratio means fifty percent above that benchmark under the stated occupancy and cost assumptions.

What does “Nightly rate per sqft” mean on Dubai Wealth Index?

The nightly rate divided by the unit's floor area — so a studio and a three-bed can be compared on price, not size.

What does “Net short-term revenue” mean on Dubai Wealth Index?

What reaches the owner after the platform's cut, cleaning and utilities — but before service charge, mortgage, tax and furnishing.

What does “Occupancy assumption” mean on Dubai Wealth Index?

Annual short-term figures assume 70% of nights are booked. That is an assumption, not a measurement. Gross booking revenue scales with occupied nights; fixed costs mean net revenue and yield do not scale proportionally.

What does “Management fee” mean on Dubai Wealth Index?

Short-term figures assume a self-managed unit (0% management fee). A managed unit typically gives up 15–25% of revenue, lowering every yield shown.

What does “500m rows” mean on Dubai Wealth Index?

Measured across the surrounding 500-metre cluster, because this building alone had too few listings of that type. Unmarked rows use listings within 50 metres.

What does “Ejari” mean on Dubai Wealth Index?

Dubai's mandatory rental-contract registry. Every rent on this site is a registered Ejari contract — a price actually agreed, not an asking price.

What does “Off-plan price gain” mean on Dubai Wealth Index?

How far today's resale price per sq ft sits above (or below) the off-plan price per sq ft, before the 4% DLD fee, agency fees and financing. Different buyers on each side, so not anyone's realised return.

What does “Median of registered records” mean on Dubai Wealth Index?

Registered sale prices and long-let rents are medians of DLD and Ejari records. Short-term figures separately use sampled asking rates and a stated model.

What does “Gross payback” mean on Dubai Wealth Index?

Years of gross rent needed to recover the purchase price: 100 ÷ yield. Ignores all costs, growth and vacancies — a rough length-of-bet, not a forecast.

What does “Break-even occupancy” mean on Dubai Wealth Index?

The occupancy at which net short-term income equals the registered long-let rent. Below it, the lease pays more; above it, the nightly let does.

What does “Villa community” mean on Dubai Wealth Index?

The grain every villa figure is published at. The Land Department records no building for a villa, so a community — Arabian Ranches, The Springs, The Valley — is the finest place a registered sale or lease can be attributed to.

What does “Title type (apartment or villa)” mean on Dubai Wealth Index?

Which of the two indexes a record belongs to is decided by its title type in the register — Unit and Flat are the apartment index, Villa and Complex Villas the villa index — never by what the development is called. A strata-titled townhouse is an apartment-index row.

What does “Rent grain (villas)” mean on Dubai Wealth Index?

Which contracts a villa community's rent came from: its own (community), contracts matched to it by project name (project), or contracts pooled from the surrounding Land Department area because it has too few of its own (vicinity).