The headline price is a number attached to a schedule

The distribution is narrow and it starts at zero, which is what it should do. A plan taking the entire price on booking has nothing to defer and returns exactly 0.00%; 117 of the 1,937 ranked projects come in under one point. At the other end 167 projects clear ten points, and the deepest in the corpus is Azha Views Residences at 23.57%. Nothing goes negative, because no instalment can land before the day we value it from.

In cash, the median ranked project asks AED 1,690,745 and is worth AED 1,587,116 once its own schedule is discounted. Measured project by project rather than as a ratio of medians, the gap runs to a median of AED 82,436 and an upper quartile of AED 177,135. That is not a discount anyone will give you at the counter. It is the part of the asking price you have not been asked for yet.

Every metric on this site is only as wide as the data behind it, and this one is the narrowest we publish. Of 2,924 tracked off-plan projects, 1,937 carry a plan discount. The other 987 break into three groups, and none of them is a rounding error.

661 publish no payment schedule at all, or publish one whose stated steps do not sum to anything near a whole price. A schedule totalling under 95 percent has lost a line somewhere between the developer and us, so it is dropped rather than rescaled: rescaling would invent the missing instalment and put it wherever the arithmetic wanted it. 222 publish a plan and a price but carry a handover date that has already gone by, which collapses the timeline to nothing: there is no deferral left to value, and the arithmetic would return a discount near zero that means “this project is late” rather than “this plan is tough”. A further 104 publish a usable schedule and no price to apply it to.

Counted against the whole corpus rather than the live book the share is lower still, 48% of 4,011 records. That figure is the less useful of the two: the corpus also holds 1,087 delivered buildings, and a building that has already handed over has no schedule left to discount by construction. Every one of them grades out, and none of them could ever have ranked.

113 of the 1,937 that do rank needed one assumption: the plan named a post-handover tail without saying how long it runs, so it is priced over 24 months, the shorter of the two lengths the corpus actually states. That understates their discount, which is the safe direction for a number a buyer is going to argue with. Those rows are labelled wherever they appear.

Sort the ranked projects by the handover year they publish and the discount climbs at every step, from 1.06% on the 2026 book to 10.47% on the 2030 one. That is the whole mechanism in one column. A 2030 handover means a schedule stretched across 4 more years of instalments, and at 5.0 percent money each of those years is worth something.

This is worth stating because the ladder is the metric’s own sanity check rather than a finding about the market. If the numbers did not rise with the horizon, the computation would be wrong. It also warns against reading a deep discount as a soft developer: a builder selling a 2030 handover collects a deeper discount than a builder selling a 2026 one on identical terms, purely because the money sits further out. The comparison that means something is between projects handing over at the same time.

Across the 45 Dubai communities with at least 10 ranked projects, the community medians run from 10.15% down to 1.89%, a range of 8.26 points. Now look inside a single one of them: the median community’s own projects spread 9.41 points from shallowest to deepest, and 31 of the 45 spread further than the entire gap between the softest address and the hardest.

Dubai Production City is the extreme case. Its 25 ranked projects run from 0.92% to 21.39%. Two buyers at the same address, in the same quarter, at the same rate per square foot, are being offered schedules that differ by 20.47 points of present value.

Which is the practical reading of this whole article. A community median tells you what to expect before you shortlist. It does not tell you what you have been offered. The plan discount is the one number on a project page that is set by the developer’s finance department rather than by the postcode, and it is the one a buyer can actually negotiate.

Jumeirah Islands carries a median plan discount of 10.15% across 11 ranked projects, 5.4x the 1.89% at DAMAC Hills 2. On a AED 1.69M ticket that difference is worth AED 139,656 in present value, for the same headline price.

Read the base count next to every median. Jumeirah Islands is ranked on 11 of the 13 projects the corpus holds there, and communities where the two numbers diverge are describing the schedules we could price rather than the address as a whole. Below 10 ranked projects a community is absent from the table entirely, because one unusually deep tail moves a median of six and there is no honest way to show that in a column.

Cut the same 1,937 projects by developer rather than by address and the range widens. Dugasta writes the deepest terms of any builder with 8 or more ranked projects, a median 15.19% across 8, against Eagle Hills at 2.32%. That is 6.5x, against 5.4x across communities.

The gap is not a small-sample artefact either. The two largest books in the table, Emaar on 165 ranked projects and DAMAC on 103, sit 2.13 points apart at 6.25% and 4.12%. Two builders of comparable scale, selling into the same city, are financing their buyers on materially different terms.

One column deserves attention before anyone leans on a row. “Tail length stated” counts how many of a builder’s ranked projects publish how long their post-handover tail runs. Where it reads short of the base count, that builder’s median rests partly on the assumed 24 month tail, and the true figure is likely deeper than shown.

The 5.0 percent rate is the one judgement call in the whole computation, so it is worth showing what it does and does not carry. Run the same schedules at 3 percent and the median discount falls to 3.23%. Run them at 8 percent and it rises to 8.25%. The magnitude scales almost linearly with the rate, which is exactly what discounting does.

The ranking does not move with it. The project ordering correlates 0.99998 between 3 and 5.0 percent, and 0.99996 between 5.0 and 8 percent. 20 of the top 20 projects are identical between 3 and 5.0 percent, and 19 of 20 between 5.0 and 8 percent.

So the defensible claim is narrower than the number suggests, and we would rather say it than let a reader assume more. Who offers the better terms is rate invariant. How much those terms are worth is not. Anyone who thinks 5.0 percent is the wrong cost of money can move it, and every league table on this page survives.

The obvious next question is whether pricing the plan changes which addresses are cheap. Mostly it does not. Rank the 45 communities on their median asking rate per square foot, then rank them again on the same rate after each project’s own plan discount: 22 do not move a single place and 33 move one place or less. The largest move in the table is Dubai Industrial City, from 12 to 6.

A plan discount of a few points cannot overturn a rate spread that runs several times over, so anyone expecting this metric to reveal a hidden bargain address will be disappointed. What it does is separate two projects that a price per square foot cannot tell apart, which is the decision a buyer with a shortlist is actually making.

Four things it does not count, all of which we could have added and all of which would have made the number larger and less defensible. It gives no credit for rent: a post-handover tail lets you live in or let the unit while still paying, so the real advantage of a tail is bigger than we report, but pricing it needs a yield assumption we would then have to defend. It gives no credit for capital growth, and takes no view on where prices go. It does not price handover delay, which cuts the other way: a long tail is also a long exposure to a builder slipping. And it compares plans at the same headline price, so where a developer quietly discounts for cash our figure flatters the deferred plan. Plans that name a cash discount are excluded from the comparison, but the corpus can only exclude the ones that say so.