A developer premium does not travel

The address moves a builder 2.05x. The builder moves an address 1.38x.

Both spreads are computed the same way, dearest divided by cheapest over medians that each rest on 25 or more records. Held to one name, the range across addresses is wider than the range across names at one address, for the same reason a plot on the Palm costs more than a plot in Majan whoever pours the concrete. The extreme case is Binghatti, at 4.8x between Majan and Downtown Dubai.

A builder qualifies for the first table at 2 or more published addresses, an address for the second at 2 or more published builders. Neither table is like for like on product: a name selling duplexes on a waterfront and studios in the suburbs will read wide on any measure.

11 of 13 builders are dearer than one address and cheaper than another.

If a premium belonged to the name it would show up wherever the name builds. Of the 13 builders published at 3 or more addresses, 1 sit above the local rate everywhere, 1 sit below it everywhere, and 11 do both. The widest of them is Binghatti, 18% under the rate at Business Bay and 113% over it at Downtown Dubai. A developer page can tell a buyer what a builder charges on average. Only the address can say whether that is a premium.

The median builder sits 11% from the address it is selling in.

A community median is what the market quotes when it has nothing more specific, and for 31 of the 155 pairs here it is more than a quarter out. 74 builders ask above their address and 81 ask below it, so the error has no default direction to correct for: it has to be read one name at a time.

The address rate is the median over every priced record that community publishes, this builder included, which makes the gap a conservative reading: a name carrying most of the book is being compared largely with itself.

Inside one address, booking deposits run 10 points apart.

Hold the postcode still and the terms still move. Across the 9 addresses with three or more builders publishing plans, the median gap between the softest and the hardest booking deposit is 10 points, against a market median deposit of 20%. The widest is Dubai Land Residence Complex, from 15% at AG to 50% at Tarrad. Where the earlier read found bedroom count barely touches a deposit, the name on the brochure moves it hard.

A builder enters with 5 or more published plans at that address, counted once per project and only where the stages sum to between 90 and 110 percent. Schemes taking the whole price on booking are full payment plans rather than deposits, and are left out of this table. Model a deposit against a real ticket in the payment plan calculator.

The builder asking under its address is the one selling out.

Split the pairs by which side of the local rate they sit on and the sale statuses split with them: builders asking below their address have a median 42% of projects marked sold out, against 29% for the builders asking above it. That is a wide gap for a market that describes almost everything as selling fast, and it is the clearest signal in this data that the premium half of an address is where stock actually sits.

Only pairs with 4 or more projects carrying a published sale status are scored. Sold out is a developer claim on a project page, not a registered transaction, and a scheme can carry it while units return to the market.

The median builder is 11% of the address, but the biggest is most of it.

Concentration decides how much a community median can be trusted at all. The median pair here carries 11% of the priced records its address publishes, which leaves the headline free to describe the other nine tenths. At the top of the table the relationship inverts: where one name carries four fifths of the book, the community median is that builder's price list wearing a postcode.

A builder over nine tenths of a community book is not published as a pair at all: at that point the community page already is the builder page, and a second surface would only repeat it.

What this means with two brochures on the table.

Read the address first and the name second. A builder's own average is drawn from wherever it happens to be building, so it travels badly: the 2.05x median spread in finding one is the size of the error a buyer inherits by quoting it at a specific address. What survives the move is the local comparison, which is what every one of these 155 pages is: this name, this postcode, every rival benchmarked beside it, and the rate gap between them stated rather than implied.

The other two axes of the same argument: address beats bedroom count and the community median is priced for the flat next door. Every builder benchmarked here is in the developer index.