How every number is computed, what each emirate's data supports, and the honesty rules.
Propefy Invest turns official government property records into a clear, searchable read on the UAE market — prices, activity, and trends, taken directly from the data, not estimated. Every figure traces back to a named government source, listed below. When the data isn't enough to answer a question honestly, we say so instead of guessing.
A quick inventory before the detail below. Every line traces to a named government source — see "Sources" further down.
Coverage is not uniform across the UAE, so we are explicit about what each emirate's open data can support. We never blend the two tiers into a single misleading figure.
Individual registered transactions — price, size, beds, building, status and date. Supports per-community, per-building and per-unit views, registered yield and comparables.
Aggregated indicators published at the area level only. Supports registered sales volume, a per-sqft price median and a month-level price trend where a unit-level transaction register exists. Abu Dhabi's ADREC register also carries individual registered transactions — see the transactions table and each area's own page. Automated valuation and hedonic index are not derivable and are left blank, not guessed. Abu Dhabi carries a separate area-level yield reading, computed from ADREC's own rent/sale price averages (see below).
Abu Dhabi's registered-sale figures come from the ADREC sales register. Two further ADREC aggregate series are layered in alongside it — an area-level yield reading, computed from ADREC's own rent/sale price averages, and transaction-type totals (behind the sales/mortgage split) — each labelled with its own source below. Across every one of these series, Abu Dhabi is served at area- or district-level aggregate grain only: no automated valuation or hedonic index is published for Abu Dhabi at any tier. A separate per-unit registered-sales table is published for Abu Dhabi's activated areas, independent of these aggregate series.
Every figure traces back to a government open-data source. We do not scrape listing portals, agents, or brokers, and we do not buy private datasets.
The 12-month change is the change in the median registered price per square foot over the trailing 12 months against the 12 months immediately before it, for the same community and property type — a plain comparison of two registered-sale medians, with the number of sales behind each shown. A second line reports Propefy's quality-adjusted price index, which holds unit size and layout constant; the two will differ whenever the mix of units sold has changed, and neither is a forecast. In Dubai, a registered apartment, villa, townhouse or hotel-apartment sale whose price per square foot is under a quarter of the median already published for its community and property type is flagged as non-market — the test runs once, when the sale is first loaded, and only where at least 30 sales stand behind that published median. A flagged sale stays in the register and in every sales count, and is left out of both medians and of the quality-adjusted index; sales of every other property type, offices and retail among them, are never flagged.
The trailing-12-month change is computed whenever the 12 months immediately before it hold at least one registered sale — it is left blank only when that prior window holds none at all, since there is then nothing to compare against. Where a table carries it, the change is shown together with the number of sales behind the trailing 12-month window itself, a small or shifting count carrying no separate caveat text of its own, and a move beyond 50% carries an additional notice, rather than being hidden. The quality-adjusted index line is shown only where a community has at least 13 monthly index points on record, and its level is anchored to that community's own first stable quarter — index levels are not comparable between communities, only each community's own change over time is.
This rating's price-trend component uses Propefy's quality-adjusted price index, not the raw median-price-per-sqft change shown above as the 12-month headline — the two can move differently when the mix of units sold changes.
In Dubai, yield is computed from the registered annual rent ÷ registered sale price for comparable units in the same area. It is a historical measure of what the market actually recorded — not a prediction, and never labelled "estimated".
Gross yield is the median registered annual rent divided by the median registered sale price for the same community and property type over the trailing 12 months — both figures are shown beside it with the number of contracts and sales behind each, so the ratio can be checked by hand. It deducts nothing; the service-charge-adjusted figure is reported separately as net yield below. A second line reports the same ratio computed per square foot, which is immune to differences in unit size between the rent and sale samples.
Where rent or price registrations are too sparse to compute a defensible figure, the yield is left blank rather than inferred from a wider, less comparable pool.
Abu Dhabi's separate area-level yield reading is not a registered figure — there is no per-contract ADREC rental register to compute one from. It is a Propefy-computed ratio of ADREC's own published period rent average ÷ published period sale-price average, pooled across every property type and layout in the district-year rather than matched to comparable units.
It is computed only for the district-year periods where ADREC publishes both a rent and a price average; periods missing either average carry no yield figure rather than one inferred from a wider pool.
Net yield is gross yield minus the registered recurring service-charge budget, from DLD owners-association (Mollak) filings — operating, cooling, and shared charges where filed. Under Dubai Law No. 6 of 2019 (Art. 25), the service charge is the owner's obligation. District-cooling consumption billed directly to occupants is regulated separately (Dubai RSB) and is not part of the registered Mollak budget figure used here.
Where a net-yield range is shown, it reflects actual variation across building clusters filing a budget in the same community and year — not a scenario or what-if band. The major-works levy, when filed, is shown separately and is not subtracted from net yield, since it is an irregular charge rather than an annually recurring one.
When a value cannot be sourced from registered data, the cell stays blank. We never fill a gap with a guess, an interpolation, or a placeholder dressed up as a real number. A blank means "we don't have it" — nothing more, nothing less.
Every community, building, or project median is shown regardless of sample size, together with the number of sales or contracts behind it — a median backed by 2 registered sales is not hidden, only labelled thin ("from 2 sales"). The statistical confidence band drawn around a median is withheld below 10 sales, since a bootstrap interval built on fewer than 10 points is not a meaningful range; the median itself still shows. Ranked lists and cross-community comparisons still require at least 30 eligible transactions, per the suppression rule below.
Any community, building, or project metric backed by fewer than 30 eligible transactions is suppressed from rankings and comparisons. A handful of deals can swing a median wildly, so thin samples are shown for context only and never placed on a leaderboard.
One named exception: the new-supply pressure ratio (pipeline units vs. delivered units) needs only 3 registered projects, not 30 — but also needs the area to have at least 100 delivered units on record; below either floor the ratio is left blank. It is a ratio of exact project counts rather than a median of sale prices, so a handful of projects doesn't carry the same sampling noise a handful of sale prices would, but a very small delivered base would still make the ratio swing on one or two projects.
A valuation anchors on registered comparable sales in the same community and applies published, standard transaction costs (DLD transfer fee, agency, trustee/admin). We surface sourced anchors separately from your own input assumptions.
We do not forecast future value or attach High/Medium/Low investment ratings. The output is a transparent calculation you can audit line by line, not an opinion.
The calculator does show an AVM read tag (Solid / Indicative / Rough) next to the estimated price range — a statistical note on how tight that range is and how many comparable transactions it was fitted on, not a judgement on the property. The model is fitted on registered sales with a recorded floor area in the same community and property type over the last 24 months, leaving out any sale flagged as non-market (priced under a quarter of the segment median), as the medians do. See the glossary below.
On the home overview, each Dubai community's median price/sqft is compared to its same-property-class peers using a statistical read (a z-score on the log-scaled median, so a handful of very high-priced communities don't skew the comparison). Communities read "Below cohort", "Near cohort" or "Above cohort" — never "undervalued", "cheap", or "bargain". A property class needs at least 20 non-suppressed communities before the comparison is shown; below that floor it reads blank rather than being compared against a different property type.
A separate "Own trend" line compares a community only to its own recent registered prices — a different, complementary axis from the cross-community comparison, and the two are never combined into one number.
Where DLD has published enough official property valuations for a segment (at least 30), registered sale prices (trailing 12 months) are compared with DLD's own assessed value (trailing 24 months — a longer window because valuations are recorded far less often than sales) for the same area and property type. Both raw figures are shown side by side, never just the derived percentage. The published median price is winsorized within its cell before it enters this comparison, which narrows its statistical confidence band but does not move the median itself — a median is unaffected by clipping the tail values around it.
The two windows are different lengths, so part of any gap can reflect market movement between the two periods rather than a difference from the assessed value alone — that is disclosed alongside the figure. Villas, townhouses and studios rarely show this comparison today, since DLD's official valuations data has almost no non-apartment residential coverage.
RentCase compares a tenant's current rent per sqft against the area's registered-rent (Ejari) median for the same property type, then maps the result onto Decree No. (43) of 2013's rent-increase slabs. It is explicitly indicative, not a legal determination — RERA's official Rental Increase Calculator is the authoritative source, linked directly alongside every result.
The registered-rent median is a proxy for RERA's own rental index, which Propefy Invest does not have access to. The comparison uses rent per sqft, not absolute rent, so units of different sizes are compared fairly — but RERA's real index bands by number of bedrooms rather than floor area, so the two figures can diverge. RentCase is Dubai-only: Decree 43/2013 does not apply in Abu Dhabi.
Data is refreshed from the source open-data feeds on a regular batch schedule. Each view carries an "as of" date so you always know how current the underlying registrations are. Every model — the price index, momentum, and every figure derived from them — refits from scratch on each refresh, so the whole historical series can move, not only the most recent points; the last two index points are marked provisional as a stronger caveat, not as the boundary of what can revise.
Abu Dhabi's "as of" date reflects the date the sale application was filed with ADREC, not a confirmed transaction date — a sale can still be pending, amended, or withdrawn after its application is filed.
Abu Dhabi class labels follow ADREC's unit-level register. Since 28 September 2026 ADREC's published aggregate reports two families only, villas and apartments, so district villa and townhouse splits cannot be checked against it. By family, district-year totals match ADREC's aggregate page except where ADREC's aggregate has since moved away from its own row feed, which this site still serves: at the current load, Al Layyan, Al Reem Island and Bani Yas apartments and Al Bahyah and Ghadeer Al Tayr villas, in some years.
A registered floor area under 100 sq ft (about 9 m²) is treated as not recorded: the sale still counts and its price still feeds the median price, but it never prices a square foot. Dubai applies the same 100 sq ft floor.
Developer and project sales metrics cover registered sales under a DLD project number, every property type and usage. A registered sale with no project number (villas and land outside any registered project) is attributed to no project; a sale under a project number the DLD projects register links to no developer counts for its project but for no developer. Both count in the market totals.
Alongside official government sources, some data that would be useful is not part of the terminal today. This is a complete list, not a sample:
KHDA school ratings and locations are not tracked or available on the terminal today.
Area-level population/demographic figures are not tracked or available.
A complete parks and beaches register is not tracked. Building-level car-park bay counts (DLD register) are unrelated and are tracked.
Needs developer-delivery data that isn't tracked — not available.
Deliberately excluded by design — see "What we never do" below, not a coverage gap.
Abu Dhabi's sales/mortgage split is a cash-vs-mortgage percentage of AD sale transactions in the area-period, computed from ADREC's resi-financing-by-period cash/mortgage transaction counts (see Sources).
To keep the platform trustworthy, some things are off the table by design:
Key terms used across the terminal, in one place.
Dubai · DLD (1,351,424 sales records, 8,845,759 Ejari contracts) · Abu Dhabi · ADREC registered sale applications
Abu Dhabi figures are computed from ADREC open data, licensed under the Abu Dhabi Government Open Data License. Dubai figures are computed from Dubai Land Department open data. Propefy is an independent service and is not affiliated with, endorsed by, or sponsored by any government authority.
The dashboard's top 25 ranks communities by registered sales in the selected period. A community with fewer than 30 registered sales in that period is not ranked, and a price change beyond ±50% carries a "Large move" marker. Every figure a community has is still shown on its own page, with no minimum.
Yield is the median registered rent per sqft divided by the median registered sale price per sqft for the same community and property type over the last 12 months — gross, nothing deducted. Abu Dhabi's district yield is a rent-to-price ratio computed from ADREC's published period averages, not a per-contract figure.
Location maps show the community point with the nearest metro stations (RTA) and DHA-licensed health facilities within 5 km; Abu Dhabi maps show the district centre point. Community and district points are derived from OpenStreetMap data — © OpenStreetMap contributors, Open Database License (ODbL) — and the site's API returns the same points. Map tiles: Mapbox and OpenStreetMap contributors.