The handover ladder: why a 2030 key costs more per sqft than a 2026 one

The pipeline is front-loaded, and 2027 is the crest.

880 projects are scheduled into 2027, 32% of everything due in the window, and the count falls away on both sides of it. 133 projects carry a 2030 date, which is 15% of the crest. That shape is a publishing artefact as much as a construction one: a developer selling today names a date two to three years out, and the far end of the window fills in later as new launches land.

Each year has its own page under the handover pipeline, with the emirate, community and developer split behind these totals.

Asking price per sqft climbs with every step of the calendar, 49% from end to end.

There is no wobble in it. Every cohort asks more per square foot than the one that hands over before it, from AED 1,673 on the 2026 book to AED 2,499 on the 2030 book. The quartiles move with the median rather than under it: the 2026 interquartile range runs AED 1,255 to 2,326, the 2030 range AED 1,768 to 3,133, so the whole distribution has slid up rather than a few trophy launches dragging an average.

The ladder is not a price forecast. It is a shelf that has been picked over.

Run the sell-through down the same table and it moves the opposite way. 57% of the 2026 cohort is recorded sold out, out of stock or close to it, against 17% of the 2030 cohort. What is still listed on a near cohort is therefore not a representative sample of what was launched into it. It is the residue: the units that did not clear at the price they were launched at, plus late releases in schemes that are already mostly gone.

Two forces run underneath the ladder and they point the same way. Launch vintage is the first: a 2030 handover was priced by a developer setting a sheet in 2026, a 2026 handover was priced years ago and the sheet has only been trimmed since. Survivorship is the second. Neither one tells a buyer that the 2030 unit will be worth more on delivery. They tell them the near-term shelf is thin, and thin shelves are where the negotiation is.

Sale status is what the developer publishes on the listing, not a registry of transactions. A project can be marked sold out and still see units come back to the market.

The Dubai-only version of the sell-through question, split by build stage rather than handover year, is in Dubai off-plan sell-through 2026.

The calendar is rounded to December, and the nearest years are rounded hardest.

Split the cohorts by quarter and Q4 takes 54% of 2026 and 38% of 2027, against a flat quarter share of 25%. Q4 is the biggest quarter, or joint biggest, in every year through 2029, and Q3 dips below Q2 in every year from 2027 out. Only 2030breaks the shape, and it breaks it because it is still filling in: 70% of that book sits in the first half of the year. Developers round a target to the end of the calendar year the way people round a delivery to the end of the month, and a date that has already slipped once gets pushed to December rather than into January. Treat a Q4 date as the year, not the quarter.

Schemes get bigger the further out they land, 14.6 priced units each against 22.7.

The 2026 cohort averages 14.6 priced unit records per project. By the middle of the window that is 22.7. Part of this is real, master-plan phases and multi-tower districts take longer and land later. Part of it is publication: a scheme still years from completion is being actively sold, so its full unit table is online, while a project handing over next quarter has retired most of its rows. The last year of the window breaks the run at 19.3, on a book of only 133 projects.

The single largest scheme anywhere in the window is Azizi Venice at 3,122 priced units, due 2026.

From 2028 out, 60% of the price before the keys is simply the standard.

The median plan on the 2026 book asks 50% of the price before the keys, and on 2027 it is 50%, the loosest year in the window. From 2028 onwards it is 60% in every single year. That is the opposite of what a buyer expects: the longer the wait, the more of the price is collected during it. The mechanism is simple enough. A developer with three or four years of build ahead has three or four years of construction milestones to bill against, and a plan that back-loads to handover leaves the build unfunded.

Read the two ladders together and the far-out cohort is dearer per square foot and asks more of that price earlier. The compensation, where it exists, is the post-handover tail, which is written into a minority of plans in every year.

Plan structures in full, including who writes a tail past handover, are in the payment plan note, and any specific plan can be run through the payment plan calculator.

The emirates change places as the calendar moves.

Dubai carries the pipeline in every year of the window, but the price order behind it does not hold still. Read across a row rather than down a column and each emirate has its own arc. Ras Al Khaimah asks more per sqft than Dubai in every single year of the window. Abu Dhabi closes on Dubai steadily, from 86 dirhams per sqft behind it on the 2026 book to -27 on 2029, and on the 2030 book it asks more than Dubai does. Two markets usually described as a tier apart are, on the far end of the published pipeline, the same price.

Emirates appearing in at least four of the 5 cohorts. A cell reads n/a where the emirate has projects in that year but too few priced medians to publish one, with the project count in brackets. Medians here are computed inside the year and the emirate, so a thin cell moves on very little.

Head to head pages for any two emirates are under compare emirates.

The map rotates: the addresses carrying 2026 are not the ones carrying 2029.

The busiest community in each cohort changes almost every year of the window. Infill districts dominate the near book, waterfront and reclaimed-island masterplans the middle, and outlying land-led districts the far end. A community that tops one year and vanishes from the next is not a market turning off, it is a masterplan finishing its sales run.

Emaar runs its own ladder, and it does not follow the market's.

Emaar appears in all 5 cohorts and carries 237 projects across the window, more than any other developer. Its own price per sqft does not climb the way the market ladder does. Its heaviest year, 2029 with 82 projects, asks AED 1,900 per sqft against a market median of AED 2,054 in the same year. Its cheapest cohort is 2028 at AED 1,566, against AED 1,801 for the market.

The point is not that one builder is dear or cheap. It is that a developer's year to year price is set by its land bank and its product mix, not by how far away the handover is, so the market ladder in Finding 02 is a property of the book as a whole and not a rule any single seller follows.

Median of the per project medians for that developer inside that year, largest book first, restricted to developers that appear in all 5 cohorts. A cell reads n/a where the developer sits outside that year's twelve largest books, with its project count in brackets where there is one. Bases here are small by construction, so read the arc rather than any single cell. Project counts per year are on the handover year pages.

Almost nothing in the published pipeline is still on paper.

Of the 2030 projects that publish a construction status, only 12 of 104 are recorded as not started. Four years out, the overwhelming majority of what is being sold is already out of the ground. That is the strongest argument in the whole dataset for taking the far end of the window seriously, and the strongest against reading it as speculative paper.

The near end carries the stranger number. 10 projects in the 2026 cohort are recorded as completed while still carrying a 2026 handover date. Some are genuinely finished and awaiting registration, some are a status field that stopped being updated. Either way, a handover date in the current year is the least reliable field in the corpus.

Construction status as published by the developer. Projects that publish no status at all are excluded from these counts, which is why the row totals sit below the cohort sizes in Finding 01.