Dubai sales fell in H1 2026, but the headline number counts a data outage as market behaviour
The first year-on-year decline in five years, separated from both seasonality and a month of missing records.
Dubai recorded 75,318 sales in the first half of 2026, against 98,305 in the same period of 2025.
H1 2025
98,305
sales registered
H1 2026
75,318
-23.4% year on year
Two things have to be dealt with before that percentage means anything.
The comparison everyone reaches for is the wrong one
The figure quoted most often sets H1 2026 against H2 2025, which produces a much steeper drop. It also compares two halves of the year that were never comparable. Dubai's second half runs ahead of its first every year, by a ratio between 1.17 and 1.28 over the last four years.
Measure a first half against a second half and you book the seasonal difference as decline before you have measured anything real.
| Period | Sales | Year on year |
|---|---|---|
| H1 2021 | 26,452 | |
| H1 2022 | 41,934 | +58.5% |
| H1 2023 | 60,871 | +45.2% |
| H1 2024 | 79,354 | +30.4% |
| H1 2025 | 98,305 | +23.9% |
| H1 2026 | 75,318 | -23.4% |
Read down that column and the story is not a crash. It is four years of decelerating growth, then a turn.
A month of the data is simply absent
Daily volume either side of the gap runs at 430 sales. At that rate those 31 days should hold roughly 13,342 sales. What was recorded is close to zero, which puts the shortfall at about 12,778 sales.
75,318
As recorded
-23.4% year on year
88,096
Gap estimate restored
-10.4% year on year
12,778
Sales unaccounted for
31 days at 430/day
The same dataset yields two conclusions more than twice apart, and which one you get depends entirely on whether anyone checked the daily counts.
The corrected figure is an estimate, not a correction. It assumes the missing weeks would have run at the same pace as the weeks around them. If activity was already slowing then, it overstates what is missing, and the true number sits somewhere between the two.
What the data cannot tell you
Nothing in the transaction record explains why activity turned. The registry holds what was sold, not why someone chose not to buy.
One external fact is worth putting beside these figures, because it points the opposite way to the obvious explanation. The UAE central bank cut its base rate three times between September and December 2025, from 4.15% to 3.65%, and held it there through mid-2026. The dirham's peg to the dollar means UAE policy tracks the Fed rather than local conditions. Borrowing costs were falling into the period when transactions fell.
That does not explain the decline. It rules out the tidiest explanation, which is worth more than a guess would be.
The second fact: no public notice of a registration interruption covering the gap was found when this was written. The gap is visible in the data. Its cause is not.
What can be said
The direction is settled. H1 2026 came in below H1 2025, the first turn after 4 years of growth.
The magnitude is not. -23.4% counts a data failure as market behaviour. -10.4% compensates for it with an assumption. Neither is precise, and quoting either one alone claims more than the evidence carries.
Data comes from Dubai Land Department historical sales records through 2026-07-27. Sales only, not rentals. Market reference, not valuation or investment advice.