Skip to main content
All broker guides

Market data · Broker guides

Accessing ADREC Transaction Data in Abu Dhabi: A Broker's Working Guide

What the Abu Dhabi registry record actually contains, what it cannot see, why recorded sales and portal listings disagree, how long registration lag runs, and how to turn registry evidence into pricing verdicts your client can check.

  • What is ADREC transaction data?
  • What does the record not contain?
  • How do recorded sales differ from portal listings?
  • How long is the lag between sale and registration?
  • How do you build a comp set from registry evidence?
  • How do you quote registry figures without overstating them?

Updated · 7 min read

ADREC transaction data is the Abu Dhabi Real Estate Centre's record of registered property transfers — the price, date, property identifiers and transaction type recorded when ownership actually changed hands. It is settled evidence rather than asking prices, and it is the one source a broker can put in front of a sceptical client and say: check it yourself, month by month.

The golden rule of the whole guide: a registered transaction answers "what was paid," a listing answers "what was hoped for," and confusing the two is the single most expensive mistake in a pricing conversation.

What is ADREC transaction data?

The registry exists because ownership in Abu Dhabi has to be recorded to be real. Every completed transfer produces a record, and the aggregate of those records is the closest thing the emirate has to a ledger of what property is worth. For a broker, the useful shape of that record is short: a date, a price, an identified property with its area and location, and a classification of what kind of transaction it was.

Three attributes matter more than the rest.

  • It is settled. A record exists because a transfer completed. Deals that collapsed at valuation or died at NOC never appear, which is exactly why the file is credible.
  • It is standardised. Price and area come from the same fields across the whole emirate, so a rate computed in one community is comparable to a rate computed in another. Portal data has no such guarantee.
  • It is retrospective. The registry is a record of the past, not a forecast. It tells you where the market has been with high confidence and where it is going with none.

Abu Dhabi is the live market here. Brokers working other emirates deal with their own registries and their own conventions, and none of the mechanics below should be assumed to transfer.

What does the record not contain?

Knowing the boundaries is what lets you speak with confidence inside them. The registry is silent on more than most brokers realise, and volunteering the silence before a client's advisor finds it is the whole credibility play.

  • Motivation. Nothing in the record says whether a seller was relocating, refinancing, or simply testing a number. A distressed sale and a patient sale look identical.
  • Condition and finish. A renovated unit and a tired one at the same size register as a size. Fit-out, view, floor level and outlook are your job, not the registry's.
  • Off-record incentives. Furniture packages, absorbed fees, service-charge holidays and post-handover payment plans sit behind clean-looking prices, particularly on primary stock.
  • Rents. Leasing sits in a different system — Tawtheeq — and rental assumptions built from live listings must always be labelled as assumptions, never treated as transaction fact.
  • The near present. The most recent weeks are structurally incomplete, for the reason set out below.

How do recorded sales differ from portal listings?

They are different species of evidence, and the gap between them is not a flaw to be explained away but a number to be used. A portal shows the market's ambition; the registry shows the market's behaviour.

QuestionRegistered transactionPortal listing
What does the price mean?What a buyer paid and a registry recordedWhat a seller is asking today
Does it survive a sceptic?Yes — it can be checked against the recordOnly as evidence of asking levels
Can it be withdrawn or revised?Rarely, and visiblyConstantly and silently
What is it best for?Valuation, negotiation, trend readingCompetitive positioning, supply visibility
Its main blind spotCondition, incentives, motivationEverything about whether the price is real

The practical rule for daily work: listings set your marketing position, transactions set your value opinion. A seller who quotes three neighbours' asking prices at you is quoting three hypotheses. Bring the recorded sales in the same building over the same period and the conversation changes register entirely — from whose opinion is bigger to what the building has actually done.

How long is the lag between sale and registration?

Longer than most brokers assume, and the gap is where quoting errors are born. A deal enters the registry when the transfer completes, not when the MOU is signed. Between those two moments sits the developer NOC, the buyer's financing, the appointment booking and ordinary human delay — commonly a few weeks, and longer where mortgages, powers of attorney or complicated title sit in the corridor.

StageWhat existsWhat you can honestly claim
Offer acceptedA price and a handshakeNothing yet — verbal offers are not evidence
MOU signed, deposit heldA contract and a timeline"A deal is agreed at X" — clearly labelled as pending
Transfer completedA recorded transaction"This sold at X"
Appears in an extractA comparable"The registry shows X on this date"

Two disciplines follow. First, the newest month is provisional — it will fill in as late registrations arrive, so read trend on settled months and say the word "provisional" out loud before anyone else has to. Second, your own pipeline is early intelligence the registry does not have yet. The deals you and your colleagues agreed this month are next quarter's comparables. Keeping your own dated log of agreed prices in your farm buildings gives you a genuine lead over anyone reading extracts alone.

How do you build a comp set from registry evidence?

The registry is raw material. A comp set is the finished product, and it is built the same way every time so that it survives a lawyer, a bank valuer and a hostile advisor.

  • Match in tiers. Same project first, then sub-community, then community. Never skip a tier upward without saying you did.
  • Match the property, not just the location. Bedrooms matched, a size band of roughly plus or minus a quarter, and a consistent area basis across every comp in the table.
  • Use enough of them. Three to four comparable transactions is the floor. Two deals are a coincidence, however precisely you write them up.
  • Weight by value, not by count. Total dirhams divided by total area, so one small outlier cannot drag the rate. Say "value-weighted" aloud — the phrase is a credential.
  • Exclude your own listings and flag the odd ones. Court-ordered sales, partial shares and related-party transfers are real records and misleading comparables.
  • Rebuild it the week you use it. One new registration in the tower can move the band.

Products like Knownable exist to compress that work — turning the raw record into value-weighted community and project rates a broker can pull in a minute rather than an afternoon. The method above still matters, because a rate you cannot rebuild by hand is a rate you cannot defend when someone asks how it was computed.

How do you quote registry figures without overstating them?

The sentence that separates evidence from opinion is short: source, method, period. "Registered transactions, value-weighted, Reem Island ready one-bedrooms, the twelve months to last settled month." Anyone can check it, and that is the point.

Around that, four habits protect you.

  • Give bands, not points. An indicative range with its logic survives negotiation; a single number dies in the first counter-offer.
  • Quote two windows. A figure measured from a trough reads as a boom and from a peak as a crash. Full period plus recent leg, always.
  • Ask what changed in the mix. A rate can rise because larger, newer or better-located units happened to transact, with nothing appreciating at all.
  • Refuse thin cells. Six registered sales cannot carry a headline rate for a whole quarter. Report the count and a band instead, rather than quoting to two decimals.

Nothing here is investment, legal or tax advice, and any range you build from a registry extract is indicative — verify current figures against the ADREC record before you put them in front of a client.

The bottom line

The registry does not win rooms; brokers who can read it do. Know what the record contains and admit what it cannot see. Treat listings as ambition and registrations as behaviour, and use the gap between them as negotiating room rather than an embarrassment. Respect the lag — the newest month is provisional, and your own agreed pipeline is intelligence the record does not have yet. Build every comp set the same way so it survives inspection, and quote with source, method and period attached. Do that consistently and the evidence, not your volume, becomes the most persuasive thing in the meeting.

Knownable is building this into a working tool for Abu Dhabi brokers. It sits in closed beta today, with Explore-tier access carrying no charge and a waitlist for everyone else.

frequently asked questions

Can I get ADREC transaction data for a single tower, or only for a whole community?

Registered transactions carry property identifiers, so tower-level and even layout-level views are possible in principle — the constraint is depth, not access. A single tower may only produce a handful of registered sales in a quarter, and a rate built on three or four deals is a range with error bars, not a figure. The working method is to build the tower view for narrative and the sub-community or community view for the number you actually quote, then say plainly which one you are standing on when a client presses.

Why is the transaction I closed last month not showing in the data yet?

Because agreement and registration are different events. A deal reaches the registry when the transfer is completed and recorded, which follows the MOU by a corridor of developer NOC time, bank time and appointment scheduling. That means the most recent month in any registry extract is always partial and will be revised upward as late registrations land. Read trends on settled months, treat the newest month as provisional, and say so out loud before you quote it — the caution is what protects you when the corrected figure arrives.

A client says the portal average is higher than my registry figure. Which one is wrong?

Neither — they measure different things. A portal average reflects asking prices, meaning what sellers hope to achieve, including listings that will never sell at that number. A registered transaction reflects what a buyer actually paid and a registry actually recorded. The gap between the two is not an error in your data; it is the measured distance between asking and reality, and it is the room you negotiate in. Explain the two species first, then offer the source and invite the client to check any month themselves.

Does a registered price tell me what the buyer really paid?

It tells you what was recorded on the transfer, which is a much stronger fact than an asking price but still not the full commercial picture. Off-record incentives sit behind clean-looking numbers: furniture packages, service-charge holidays, absorbed fees, post-handover payment plans on primary stock. Two units at an identical recorded price can represent materially different deals. Use the recorded figure as the anchor and treat known incentives as an annotation you keep yourself, clearly labelled as your observation rather than registry fact.

How do I use registry evidence without breaking advertising rules?

Keep the permit discipline and the data discipline separate but simultaneous. Every advertisement still needs its ADREC ad permit number, whatever evidence you cite inside it. When you quote a transaction figure in a just-sold campaign or a seller letter, quote it as a rate and a period rather than as identifying detail about a private party, name the source and the month, and avoid dressing an indicative range as a precise valuation. Evidence-led marketing is fine; implying a guaranteed price or a personal endorsement from the registry is not.