Dubai rental yields from registered leases, and why the published figure is too low
Yields by area from two registers, the correction that completed-stock pricing makes, and what explains its size.
Most Dubai yield figures divide an asking rent by an asking price. Both sides here come from registers instead: Ejari lease contracts over Land Department sales, for the same year, with both sides reduced to a rate per square metre so they can be divided at all.
Yields by area
| Area | Gross yield | Rent | Sale price | Lease contracts | Sales |
|---|---|---|---|---|---|
| Al Warsan First | 8.47% | 597 AED/sqm/yr | 7,047 AED/sqm | 25,905 | 2,167 |
| Jabal Ali Industrial Second | 7.38% | 1,046 AED/sqm/yr | 14,167 AED/sqm | 898 | 786 |
| Al Merkadh | 6.65% | 1,478 AED/sqm/yr | 22,210 AED/sqm | 11,346 | 3,166 |
| Al Hebiah Second | 6.51% | 1,021 AED/sqm/yr | 15,672 AED/sqm | 1,404 | 1,685 |
| Al Hebiah Third | 6.44% | 1,022 AED/sqm/yr | 15,869 AED/sqm | 4,324 | 1,466 |
| Warsan Fourth | 6.36% | 691 AED/sqm/yr | 10,866 AED/sqm | 9,969 | 1,661 |
| Area | Gross yield | Rent | Sale price | Lease contracts | Sales |
|---|---|---|---|---|---|
| Trade Center Second | 1.83% | 930 AED/sqm/yr | 50,839 AED/sqm | 2,866 | 765 |
| Madinat Dubai Almelaheyah | 2.04% | 656 AED/sqm/yr | 32,160 AED/sqm | 1,786 | 5,495 |
| Al Hebiah First | 3.54% | 727 AED/sqm/yr | 20,559 AED/sqm | 2,383 | 6,114 |
| Al Wasl | 3.73% | 1,249 AED/sqm/yr | 33,520 AED/sqm | 1,964 | 2,064 |
| Marsa Dubai | 4.01% | 1,165 AED/sqm/yr | 29,064 AED/sqm | 19,729 | 7,237 |
| Al Satwa | 4.03% | 921 AED/sqm/yr | 22,852 AED/sqm | 5,800 | 1,292 |
The ordering is the one anybody familiar with the city would predict. Outer districts return four times what the central towers return, and the reason is visible in the two price columns: rents vary across the city by far less than sale prices do.
The correction
A lease is signed on property that exists. A sale record in an area of active launches is mostly a price for property that does not exist yet, agreed with a developer rather than negotiated between two owners.
So the standard yield divides a real rent by a denominator that describes different stock. Restricting the sale side to completed property puts both sides on the same buildings.
33
Areas with enough completed sales
33
Where the correction raises the yield
+0.9pt
Median correction
+3.3pt
Largest correction
| Area | Published basis | Completed stock only | Difference | Off-plan share | Completed sales |
|---|---|---|---|---|---|
| Al Hebiah First | 3.54% | 6.86% | +3.3pt | 82% | 1,103 |
| Al Hebiah Fifth | 4.29% | 7.60% | +3.3pt | 59% | 475 |
| Wadi Al Safa 5 | 4.84% | 8.00% | +3.2pt | 87% | 1,143 |
| Wadi Al Safa 2 | 4.67% | 7.49% | +2.8pt | 73% | 534 |
| Madinat Al Mataar | 4.39% | 7.16% | +2.8pt | 88% | 973 |
| Al Hebiah Fourth | 6.02% | 8.78% | +2.8pt | 57% | 1,544 |
Al Hebiah First moves from 3.54% to 6.86%, close to double. A landlord there comparing the published figure against a bank deposit is working from the wrong number.
What sets the size of it
The obvious guess is that the correction is biggest where off-plan is a bigger share of sales. That holds loosely and explains less than half of it.
What explains the rest is how much more off-plan costs than completed stock in that particular area. Across the areas compared, off-plan sells at a median of 18% above completed property, reaching 94% at the extreme, and that premium tracks the correction at 0.97.
How these figures are built
Both sides are medians of price per square metre. A total-rent over total-price ratio would compare a studio's rent against a townhouse's price wherever the two datasets differ in mix, and they differ in every area.
Four filters do the real work, and each was added after the number came out wrong without it.
| Filter | What it removes | What it prevented |
|---|---|---|
| Sales only | Mortgage registrations | One area returned a sale rate of 2,886 AED/sqm because all 400 of its rows were loans, where the amount recorded is what was borrowed |
| Single-property leases | Bulk leases over many units | A block lease carries the total for every unit against one unit's size, which produced 147,400 AED/sqm/yr against a real figure near 800 |
| Residential use | Warehouses, shops, storage | Industrial districts entered the residential table at forty times the citywide rate |
| Size and amount ranges | Parking bays, per-bed staff accommodation, whole floors | Contracts under 30 sqm returned rates an order of magnitude above the rest |
Gross, not net. Service charges, agency fees, void periods and maintenance all come out of these figures before a landlord sees anything, and none of them are in either register. Service charges alone commonly run a fifth of gross rent in Dubai towers, and they vary by building rather than by area, so no citywide adjustment would be honest.
Using this
Read the corrected column where an area sells a lot of off-plan. It is the one describing property a landlord can let today.
Check both sample sizes. Every row here clears 200 completed sales and several hundred lease contracts, and the counts are printed because a yield built on thin data moves with a handful of transactions.
Then stop at the area. Service charges decide whether a strong gross yield survives to a net one, and those are set per building.
Related reading: the off-plan share by area covers the split this correction depends on, and what this data covers explains why the two registers have to be joined at the area rather than at the building.
Rent data comes from Ejari registered lease contracts for 2025. Sale data comes from Dubai Land Department records through 2026-08-13. These are gross yields and market reference, not valuation or investment advice.