---
author: Imad A. Adel
---

# Dubai gross rental yield history

We measure the annual rent on a registered lease relative to the price of a ready home in the same community, dwelling class and bedroom group. The headline series uses new contracts. The all-contract series includes renewals and supplies the rent numerator for the published comparison.

## Sources and dates

Dubai Land Department publishes both registers on data.dubai under the Open Data Licence. The manifest lists 11 Ejari rent files dated 20 November 2026 at 18:02:26 and two transaction files dated 20 November 2026 at 17:50:42. Retrieval took place on 21 November UTC. All rent rows carry the load timestamp `2026-11-20T17:50:31.000Z` and all transaction rows carry `2026-11-20T17:47:24.000Z`. [Rent Contracts](https://data.dubai/en/l/468586) · [Real Estate Transactions](https://data.dubai/en/l/470061)

**Attribution.** Source data: Dubai Land Department, Rent Contracts dataset (https://data.dubai/en/l/468586) and Real Estate Transactions dataset (https://data.dubai/en/l/470061), published as open data on data.dubai under the Dubai Open Data Licence, snapshot of 20 November 2026. The medians, counts and yields in this folder are calculated by Adel Labs and are made available on the terms of the same licence. This work is not affiliated with or endorsed by the Dubai Land Department or any Dubai Government entity. No raw row of either register is reproduced here.

The benchmark is **Dubai Data (datadubai.ae), from Dubai Land Department open data**, licensed under CC BY 4.0. Its saved pages and methods are documented in the [benchmark extract](../README.md). Individual community page addresses are in `../yield_segments.csv`.

A window runs from 1 November to 31 October and takes its closing year. We assign leases by `contract_start_date` and transactions by `instance_date`. Window 2026 therefore ends on 31 October 2026. For the first published year we require eligible records in each of the twelve months in both registers and at least one Dubai dwelling segment with 15 new leases and 15 ready sales.

## Record selection

Before reading the raw CSVs we check each part's byte size against the manifest, then reconcile the number of lines read. A residential lease must have a mapped Ejari dwelling type and bedroom label, one property, and a nonblank contract ID occurring on exactly one line across the full delivery. Repeated IDs are checked across files and include lines outside the study period. Annual rent must fall between AED 5,000 and AED 5,000,000 inclusive. The term, end date less start date, must be 300–400 days inclusive. The term sensitivity repeats the 2026 rent calculation with that last restriction lifted.

A sale must be in the Sales group, registered under Existing Properties, and have residential usage, a Unit or Villa property type, a mapped dwelling subtype and room label, and an eligible priced-sale procedure. Actual worth must fall between AED 50,000 and AED 5,000,000,000, inclusive. The procedure inventory below covers every raw Sales line before other exclusions.

Within each register we retain only explicit mappings. A studio has zero bedrooms. Five or more bedrooms share the code `5`. Missing and ambiguous labels, including penthouses without a bedroom count, have no bedroom assignment. Community rows use the English master-project field with surrounding whitespace removed. Dubai-wide estimates also include eligible addresses with no master project.

## Statistics and sample sizes

For each rent basis and sales cohort, we report the 25th percentile, median and 75th percentile by linear interpolation. Money is shown in whole AED. Gross yield is 100 times median annual rent divided by median ready price, using medians before display rounding. We publish a yield when both sides contain at least 15 observations. A minimum of 15–29 on either side gives a thin result; with 30 or more on both sides the yield carries no flag. A market median can be published on its own when its sample reaches 15. Positive sample counts below 15 are blank. Filter counts, segment counts and inventories describe coverage.

A community's yield is the unweighted median of its eligible flat and villa segment yields, calculated before rounding. Segment membership may change between windows. A Dubai-wide bedroom yield pools eligible records across Dubai. Each change in percentage points uses the published two-decimal yields. Changes above +0.50 pp are widened, changes below −0.50 pp are compressed, and the boundary values are held.

The comparison with the peak includes 2026 among potential peak years and requires a 2026 yield and coverage in at least five of the six preceding windows from 2020. The 2019 and 2021 comparisons each require observations at both endpoints. The samples can therefore differ between comparisons.

## Rule tested and dropped

The former rule, which trimmed rents and prices at the Dubai-wide 1st and 99th percentiles and applied no absolute rent bounds, retained 726 Dubai Marina one-bedroom ready sales in 2026 at a median of AED 1,550,000, compared with 762 at AED 1,545,856 under absolute bounds alone. For DownTown Dubai studios, the former rule retained 330 new leases at AED 65,000 in 2026, compared with 405 at AED 69,000, and 272 at AED 40,500 in 2021, compared with 332 at AED 43,000.

## Registry geography

The five exact master-project strings are `Dubai Marina`, `Business Bay`, `DownTown Dubai`, `Jumeirah Village Circle` and `DMCC Master Community`. We display the fifth as **Jumeirah Lakes Towers (registered as DMCC Master Community)**, with slug `dmcc-master-community`. This registry name may include addresses beyond the towers’ district where the registers assign them to this master project. Across the complete raw delivery, 3,911 of 164,088 rent lines (2.38%) and 1,473 of 84,228 transaction lines (1.75%) fall outside Al Thanyah Fifth, leaving 97.62% and 98.25% inside, respectively.

A case-insensitive search for “DMCC” or “Lakes Towers” in the full English master-project field returned the labels below. Neither register contained a “Lakes Towers” label. The separate `DMCC-EZ2` label contributes no rows to the named series. The geographic audit uses all raw lines, including contracts dated beyond the last published window.

| Window | DMCC Master Community rent lines | Of which Al Thanyah Fifth | DMCC Master Community transaction lines | Of which Al Thanyah Fifth | DMCC-EZ2 transaction lines |
| --- | --- | --- | --- | --- | --- |
| 2003 | 0 | 0 | 1 | 1 | 0 |
| 2004 | 0 | 0 | 1 | 1 | 0 |
| 2005 | 0 | 0 | 2 | 2 | 0 |
| 2007 | 0 | 0 | 3 | 3 | 0 |
| 2008 | 0 | 0 | 302 | 302 | 0 |
| 2009 | 1 | 1 | 8,491 | 7,970 | 0 |
| 2010 | 12 | 12 | 6,085 | 6,073 | 0 |
| 2011 | 145 | 145 | 4,707 | 4,614 | 0 |
| 2012 | 469 | 469 | 6,613 | 6,511 | 0 |
| 2013 | 3,627 | 3,626 | 8,006 | 7,744 | 0 |
| 2014 | 5,897 | 5,885 | 4,212 | 4,138 | 0 |
| 2015 | 6,249 | 6,238 | 3,060 | 3,012 | 2 |
| 2016 | 6,770 | 6,751 | 2,899 | 2,873 | 0 |
| 2017 | 8,463 | 8,449 | 2,777 | 2,712 | 0 |
| 2018 | 12,308 | 12,007 | 2,007 | 1,992 | 0 |
| 2019 | 13,400 | 12,878 | 2,310 | 2,227 | 0 |
| 2020 | 13,247 | 12,770 | 1,578 | 1,562 | 0 |
| 2021 | 15,124 | 14,702 | 3,970 | 3,954 | 0 |
| 2022 | 15,547 | 15,124 | 3,456 | 3,418 | 0 |
| 2023 | 15,778 | 15,424 | 6,063 | 6,045 | 0 |
| 2024 | 15,227 | 14,868 | 6,918 | 6,891 | 0 |
| 2025 | 15,675 | 15,205 | 6,640 | 6,606 | 898 |
| 2026 | 15,127 | 14,657 | 3,878 | 3,856 | 1,757 |
| 2027 | 1,019 | 963 | 249 | 248 | 46 |
| 2028 | 3 | 3 | 0 | 0 | 0 |

Dubai Data retains both names. Its calendar-year table records 3,417 sales in 2025 and 109 in 2026 on the Jumeirah Lakes Towers page. The DMCC Master Community page records one in 2023, one in 2025 and 1,652 in 2026. Both pages list Al Thanyah Fifth in their area breakdown. The DMCC page also lists Jabal Ali First. We compare the registry series with the DMCC page alone. Both publisher rows retain their original keys in the validation file. [Jumeirah Lakes Towers](../source-pages/dubai-data/communities/jumeirah-lakes-towers.json) · [DMCC Master Community](../source-pages/dubai-data/communities/dmcc-master-community.json)

## Table keys and fields

| File | Observations and fields |
|---|---|
| `yield_history_by_segment.csv` | One row per window, registry community, home class, bedroom group and rent basis. The `without_term` columns report the count and median for the five named communities in 2026 with the term restriction lifted. |
| `yield_history_dubai.csv` | Same dwelling and rent-basis key across Dubai. Window-wide diagnostics repeat across rows and must be read once per window. Raw counts precede eligibility tests. The missing-master count is for raw rent lines. Kept rents and sales and exclusions at the term step are separate columns. The all-window term exclusion includes prehistory. The August, September and October fields count distinct nonblank contract IDs among all raw starts in each month in 2025 and 2026. |
| `yield_history_by_community.csv` | One row per window, registry community and basis. `n_segments` counts eligible components and `n_thin_segments` counts thin components. The flag is `insufficient` for zero components, `includes-thin` if any are thin, and `ok` otherwise. |
| `validation_vs_published.csv` | One row per benchmark community, home class and bedroom group in 2026. The `our_` and `published_` columns give all-contract rent, ready price, yield and counts. Signed differences are ours less published in AED, percent of the published median, percentage points of yield or records. The sample flag describes our cohort. |
| `percentile_rule_test.csv` | One row per window, registry community and flat bedroom group for new leases. Prefixes `old_` and `new_` identify sample counts and medians for new-contract rent and ready price under each rule. |
| `label_maps.csv` | Register, field and raw label identify a row. The mapped value and decision describe the label in isolation. The count covers the delivery, with first and last occurrences shown within the published windows. A blank raw label denotes missing text. Procedure rows cover the Sales group. |
| `RAW_MANIFEST.csv` | Dataset name, ID, file name, byte size, SHA-256, rows, columns, retrieval time and portal addresses for each source part. |

On segment and Dubai rows, `basis` is all, new or renew. `n_rents` counts the corresponding leases. `n_ready_sales` repeats for each rent basis of a dwelling segment. The `rent_` and `ready_price_` fields ending in `q1_aed`, `median_aed` and `q3_aed` describe their respective distributions. `yield_new_pct` and `yield_all_pct` appear on the corresponding basis rows. The renewal row reports the rent distribution.

Run the two scripts offline from the project root. The first uses pandas and NumPy and takes the local directory containing `rent_contracts/` and `transactions/`. The second uses the Python standard library and reads the aggregate tables.

```text
python -I -B source/analysis/build_yield_history.py <raw-root>
python -I -B source/analysis/analyse_history.py
```

The analysis tables and their figure register are in `source/data/analysis-history.json` and `source/data/number-bank-history.json`. Each figure records its unit, table, row and column. Calculated figures also identify their formula and inputs.

The first Dubai record carries `rent_filter_ledger` and `sale_filter_ledger`, JSON-encoded copies of their respective diagnostics in `filter_counts.json`. Each records sequential exclusions, counts read and kept, and the raw master-project audit.

## Filter ledger

The counts are sequential. Each exclusion uses the records remaining after the preceding step.

| Ejari lines | Removed at step | Remaining |
| --- | --- | --- |
| Read | — | 10,745,998 |
| Missing date or later than window 2026 | 74,094 | 10,671,904 |
| Usage other than residential | 3,121,593 | 7,550,311 |
| Unmapped dwelling class | 1,327,561 | 6,222,750 |
| Unmapped bedrooms | 13,923 | 6,208,827 |
| Property count other than one | 305,136 | 5,903,691 |
| Missing or repeated contract ID | 9,254 | 5,894,437 |
| Missing or nonfinite annual amount | 0 | 5,894,437 |
| Term outside 300–400 days or missing | 317,788 | 5,576,649 |
| Annual amount outside absolute bounds | 1,535 | 5,575,114 |
| Before first complete publishable window | 4,622 | 5,570,492 |
| Kept | — | 5,570,492 |

| Transaction lines | Removed at step | Remaining |
| --- | --- | --- |
| Read | — | 1,822,024 |
| Missing date or later than window 2026 | 10,907 | 1,811,117 |
| Group other than Sales | 426,208 | 1,384,909 |
| Registration other than Existing Properties | 648,617 | 736,292 |
| Usage other than residential | 162,855 | 573,437 |
| Ineligible dwelling type or subtype | 111,550 | 461,887 |
| Unmapped bedrooms | 1,194 | 460,693 |
| Procedure outside priced-sale list | 34,012 | 426,681 |
| Missing price or outside absolute bounds | 878 | 425,803 |
| Before first complete publishable window | 51,357 | 374,446 |
| Kept | — | 374,446 |

The preceding window, 2010, contains eligible rents in 11 months and eligible sales in all 12, with eight Dubai bedroom and dwelling groups at the new-lease publication floor. The 16 complete windows are shown below. The filter ledger also records eligible rows before the first complete window, by year.

| Window | Raw rent lines | Kept leases | Raw transaction lines | Kept ready sales | Raw rents without master project |
| --- | --- | --- | --- | --- | --- |
| 2011 | 48,124 | 29,479 | 37,564 | 15,805 | 90.78% |
| 2012 | 143,298 | 89,789 | 43,198 | 17,767 | 88.19% |
| 2013 | 305,281 | 188,447 | 72,818 | 31,120 | 77.24% |
| 2014 | 436,537 | 246,130 | 72,292 | 23,731 | 73.10% |
| 2015 | 512,995 | 266,892 | 59,036 | 15,638 | 72.85% |
| 2016 | 545,887 | 276,354 | 61,402 | 11,713 | 71.86% |
| 2017 | 579,976 | 298,828 | 64,205 | 12,262 | 69.75% |
| 2018 | 632,442 | 328,120 | 52,305 | 9,353 | 67.38% |
| 2019 | 651,037 | 347,887 | 51,520 | 9,507 | 65.68% |
| 2020 | 649,811 | 348,995 | 48,771 | 10,448 | 62.38% |
| 2021 | 741,678 | 403,029 | 76,961 | 20,244 | 60.66% |
| 2022 | 878,486 | 472,982 | 111,355 | 28,305 | 62.48% |
| 2023 | 976,057 | 524,848 | 160,474 | 39,001 | 63.44% |
| 2024 | 1,132,789 | 566,236 | 215,712 | 45,384 | 66.01% |
| 2025 | 1,229,242 | 589,840 | 258,466 | 47,359 | 65.94% |
| 2026 | 1,199,425 | 592,636 | 223,798 | 36,809 | 64.57% |

### Sale procedures

| Raw procedure in Sales group | Lines | Decision |
| --- | --- | --- |
| Adding Land By Sell | 471 | dropped |
| Delayed Development | 1,343 | dropped |
| Delayed Lease to Own Registration | 1,089 | dropped |
| Delayed Sell | 156,977 | kept |
| Delayed Sell Development | 165 | kept |
| Delayed Sell Lease to Own Registration | 370 | dropped |
| Development Registration | 12,879 | dropped |
| Development Registration Pre-Registration | 6,254 | dropped |
| Lease Development Registration | 471 | dropped |
| Lease to Own Registration | 29,915 | dropped |
| Lease to Own Registration Pre-Registration | 3,903 | dropped |
| Lease to Own on Development Registration | 310 | dropped |
| Portfolio Development Registration | 6 | dropped |
| Sale On Payment Plan | 2,411 | kept |
| Sell | 519,982 | kept |
| Sell - Pre registration | 642,382 | kept |
| Sell Development | 13,058 | kept |
| Sell Development - Pre Registration | 284 | kept |

## Benchmark agreement

Dubai Data removes values outside its 1st–99th percentiles, describes area and unit-price bounds for price statistics, counts transactions before trimming, and limits displayed yields to 2–15%. Our validation uses its ready-home segment table. Its other bedroom-level tables include off-plan sales. We compare its five-bedroom category with our five-or-more category. The publisher does not supply record IDs for its calculation, so a single difference cannot be assigned to a specific selection step. [Saved methodology](../source-pages/dubai-data/methodology.md)

Of 228 benchmark segments, 220 support a rent comparison, 216 a price comparison and 216 a yield comparison. Percent errors use the published median as denominator. Each tolerance share uses the segments with two available medians.

| Measure | Within 2% | Within 5% | Median absolute percentage difference |
| --- | --- | --- | --- |
| All-contract annual rent | 191/220 (86.82%) | 216/220 (98.18%) | 0.00% |
| Ready sale price | 180/216 (83.33%) | 205/216 (94.91%) | 0.22% |

The median absolute yield difference was 0.065 pp (0.07 pp to two decimals). The rent and price median absolute percentage differences both remain below the 5% review threshold.

### Named communities in the benchmark

Each pair gives our figure followed by the publisher’s. Complete counts and differences are in `validation_vs_published.csv`.

| Community | Class | Bedrooms | Median rent (AED) | Median price (AED) | All-contract yield (%) |
| --- | --- | --- | --- | --- | --- |
| Business Bay | flat | 0 | 65,000 / 65,000 | 1,000,000 / 1,011,468 | 6.50 / 6.43 |
| Business Bay | flat | 1 | 85,000 / 85,000 | 1,443,592 / 1,450,000 | 5.89 / 5.86 |
| Business Bay | flat | 2 | 128,350 / 129,555 | 2,260,000 / 2,250,000 | 5.68 / 5.76 |
| Business Bay | flat | 3 | 185,000 / 190,000 | 3,500,000 / 3,345,000 | 5.29 / 5.68 |
| Business Bay | flat | 4 | 300,000 / 300,000 | 5,400,000 / 5,349,550 | 5.56 / 5.61 |
| Jumeirah Lakes Towers (registered as DMCC Master Community) | flat | 0 | 57,750 / 56,000 | 770,000 / 765,000 | 7.50 / 7.32 |
| Jumeirah Lakes Towers (registered as DMCC Master Community) | flat | 1 | 80,000 / 80,000 | 1,150,000 / 1,127,500 | 6.96 / 7.10 |
| Jumeirah Lakes Towers (registered as DMCC Master Community) | flat | 2 | 120,000 / 120,000 | 1,950,000 / 1,900,000 | 6.15 / 6.32 |
| Jumeirah Lakes Towers (registered as DMCC Master Community) | flat | 3 | 140,000 / 140,000 | 2,600,000 / 2,600,000 | 5.38 / 5.38 |
| DownTown Dubai | flat | 0 | 70,000 / 70,000 | 1,150,000 / 1,200,000 | 6.09 / 5.83 |
| DownTown Dubai | flat | 1 | 112,000 / 112,697 | 2,000,000 / 2,120,000 | 5.60 / 5.32 |
| DownTown Dubai | flat | 2 | 175,000 / 175,000 | 3,380,000 / 3,652,735 | 5.18 / 4.79 |
| DownTown Dubai | flat | 3 | 295,000 / 295,000 | 5,500,000 / 5,605,000 | 5.36 / 5.26 |
| Dubai Marina | flat | 0 | 60,000 / 60,000 | 1,050,000 / 1,050,000 | 5.71 / 5.71 |
| Dubai Marina | flat | 1 | 89,759 / 90,000 | 1,545,856 / 1,550,000 | 5.81 / 5.81 |
| Dubai Marina | flat | 2 | 137,000 / 138,000 | 2,510,000 / 2,525,000 | 5.46 / 5.47 |
| Dubai Marina | flat | 3 | 210,000 / 210,000 | 4,000,000 / 3,988,346 | 5.25 / 5.27 |
| Dubai Marina | flat | 4 | 285,125 / 290,000 | 4,500,000 / 3,700,000 | 6.34 / 7.84 |
| Jumeirah Village Circle | flat | 0 | 47,000 / 48,000 | 610,000 / 615,000 | 7.70 / 7.80 |
| Jumeirah Village Circle | flat | 1 | 69,000 / 70,000 | 1,000,000 / 1,000,000 | 6.90 / 7.00 |
| Jumeirah Village Circle | flat | 2 | 93,856 / 95,000 | 1,500,000 / 1,500,000 | 6.26 / 6.33 |
| Jumeirah Village Circle | flat | 3 | 140,000 / 140,000 | 2,261,475 / 2,261,475 | 6.19 / 6.19 |
| Jumeirah Village Circle | villa | 3 | 175,000 / 175,000 | 3,100,000 / 3,100,000 | 5.65 / 5.65 |
| Jumeirah Village Circle | villa | 4 | 200,000 / 200,000 | 3,325,000 / 3,300,000 | 6.02 / 6.06 |

## Independent checks and limits

Across all five named communities, a separate raw-row calculation reproduced the counts and published medians for new rent, all-contract rent and ready price for one-bedroom flats in 2019, 2021 and 2026. We also compared the four communities covered by the lead’s narrower Sell and Delayed Sell check. All 36 medians agreed to whole AED and all 24 rent counts agreed. The wider procedure list added two ready sales in Dubai Marina in 2021 (820 versus 818) and 14 in Jumeirah Village Circle in 2026 (2,380 versus 2,366). The other ten ready-sale counts agreed.

The registers describe different homes sold and let. They cannot identify a matched property return, service charges, vacancy costs, financing costs or the effect of regulation on a renewal. The lease register gives the contract start date and supplies no registration date. The monthly start counts measure reporting coverage at the end of the extract.
