Studios or Villas: What Does AED 10 Million Really Buy?

Smaller is not automatically better—but in Dubai’s current recorded mix, the studio’s indicative area-normalised gross yield is 5.45%, versus 3.77% for a three-bedroom villa. Neither number is a realised return on a specific property.

The raw data layer

A claim that smaller Dubai homes deliver better investor returns needs a definition of “return” before it needs a ranking. The available measure is narrower: for each asset-and-layout category, it divides mean annual contracted rent per measured square metre by mean registered sale price per measured square metre. Both means come from independent Dubai Land Department-derived cohorts in public/data/market-profiles/dubai.json, reflected in data/evidence.json:investment. The unit of analysis is a recorded residential Unit or Villa sale, or a physical-unit rental contract, grouped by layout. The window is September 2025–August 2026, the last complete rolling year in these profiles. Sales include ready and off-plan homes; rental contracts concern physical premises. No sale is paired with a lease on the same property. The quotient is therefore a gross area-rate proxy, not income received on an acquired asset.

On that definition, studios record 5.45%, based on 43,994 sales and 105,030 rental contracts. Three-bedroom apartments record 3.26%, based on 7,994 sales and 35,370 contracts; three-bedroom villas record 3.77%, based on 4,666 sales and 25,971 contracts. These counts describe the two inputs separately, not a pool of properties observed both selling and renting. The villa source class also includes townhouses. None of these cells reports vacancy, service charges, acquisition costs, financing or resale proceeds.

The profile also supplies mean tickets: a studio sale of AED 755,701 and an annual studio contract of AED 42,299; a three-bedroom villa sale of AED 3,553,238 and annual contract of AED 163,590. These are useful for a budget illustration but cannot replace the area-rate definition. Dividing one mean ticket by the other would compare the prices of the sold cohort with the rents of a different leased cohort, whose unit-size and property mix need not match. Neither calculation is the yield available on a current listing.

The named-community evidence, data/evidence.json:investment_communities, applies display minimums of 50 sales and 100 leases for apartment layouts, or 30 sales and 80 leases for three-bedroom villas, then shows the highest qualifying ratios by category. It is a selected table, not a complete map of Dubai communities. The distinction matters: a conspicuous community cell can be arithmetically sound while offering little evidence about the purchase price, achievable rent or running costs of an identifiable ready home.

The deeper analysis layer

The first test is whether the apparent size pattern persists within an asset class. Across apartments, the published area-rate proxy declines from 5.45% for studios to 4.51% for one-bedroom, 3.71% for two-bedroom, 3.26% for three-bedroom and 2.66% for four-bedroom units. The studio–three-bedroom apartment gap is 2.19 percentage points, using those two category ratios as its denominator comparison, not matched properties. Sample depth also changes: the studio’s 43,994 sales are about 5.5 times the three-bedroom apartment’s 7,994, while its 105,030 leases are about 3.0 times the latter’s 35,370. A consistent ordering of five observed apartment cells is evidence of a cohort pattern, not evidence that reducing a particular flat’s size would raise its return.

Villas supply a counterexample to a universal “smaller wins” rule. Their one-, two-, three- and four-bedroom area-rate proxies are 6.45%, 4.78%, 3.77% and 4.13%, respectively. The four-bedroom figure exceeds the three-bedroom figure by 0.36 percentage points, rather than continuing downward. At the other end, the unusually high one-bedroom villa cell rests on only 127 sales and 728 leases. A ranking that leads with its 6.45% but omits that sale denominator would conceal how unlike the deep studio cohort it is.

A second test changes the calculation, without changing the supplied observations. Dividing the mean studio annual ticket, AED 42,299, by its mean sale ticket, AED 755,701, gives 5.60% after rounding—not the published 5.45% area-rate proxy. For three-bedroom villas, AED 163,590 ÷ AED 3,553,238 gives 4.60%, rather than 3.77%. The studio’s lead over that villa is thus 1.68 percentage points on the published area-rate measure, but about 0.99 percentage points on the independent-cohort mean-ticket calculation. The ticket calculation is not a correction or a better property yield. The divergence demonstrates that a conclusion about the size of the advantage depends on which denominator the question actually requires.

Four-bedroom villas sharpen this warning. Their mean sale ticket, AED 3,648,308, is only AED 95,070, or about 2.7%, above the three-bedroom villa mean of AED 3,553,238. Their mean annual contract, AED 255,093, is AED 91,503, or about 55.9%, above the three-bedroom mean of AED 163,590. Yet the published area-rate proxies differ by just 0.36 percentage points. Different calculations and independently selected sold and leased populations can tell substantially different stories about “size”; the aggregates do not identify which feature of those populations produced the difference.

The familiar AED 10 million allocation thought experiment fails if it silently compares ten studios with three average three-bedroom villas. Ten studios at their observed mean price total AED 7,557,010; three villas total AED 10,659,714, exceeding the stated budget by AED 659,714, before any purchase costs. A feasible count-based illustration is ten studios versus two villas. The two villas total AED 7,106,476, so the baskets deploy different amounts and leave AED 2,442,990 and AED 2,893,524 of the budget unused, respectively. Multiplying the independent lease means gives AED 422,990 and AED 327,180 in illustrative annual gross rent. Their AED 95,810 rent difference is not a portfolio-profit difference: the studio basket deploys AED 450,534 more capital, and neither basket has been observed to earn those rents.

Sensitivity can be calculated without pretending to observe an operating record. If every home in each illustrative basket lost one month of its assumed annual rent, and all other inputs stayed fixed, studio gross rent would be about AED 387,741 and two-villa gross rent AED 299,915. On their respective mean-price baskets, the hypothetical ticket ratios would fall from 5.60% to 5.13% and from 4.60% to 4.22%. This is an identical one-month-per-property assumption, not a measured vacancy rate or a forecast of how often either layout turns over.

Costs offer a different break-even test. The unadjusted mean-ticket studio ratio exceeds the unadjusted three-bedroom villa ratio by about 0.993 percentage points. An additional annual studio expense equal to roughly AED 7,507 per average-priced studio, or AED 75,068 across ten, would erase that ticket-ratio edge if the comparator villa incurred no corresponding additional expense. That deliberately asymmetric scenario identifies how much unmeasured cost could matter; it does not estimate either asset’s service charges, repairs or net yield. For an actual jointly owned apartment, Dubai Law No. (6) of 2019, Articles (16), (25) and (27) puts service/usage charges on the owner unless the unit lease says otherwise (the owner remains liable if the tenant fails to pay), defines an area-based share of approved common-parts charges, and requires RERA approval of the budget. These are not optional deductions from an investable income analysis: obtain the candidate unit's approved actual charges and arrears rather than fill this scenario with a guessed Dubai-wide amount. Equal assumed costs as a percentage of price would not, by themselves, establish the same reversal.

Geography tests the thesis from another direction. Dubai Investment Park Second studios lead the displayed studio communities at 9.44%, but the cell has 85 sales and 1,177 leases. International City Phase 1 studios show 7.97% with 631 sales and 10,605 leases. The first ratio is 1.47 percentage points higher while relying on fewer sale observations. Those counts do not measure precision by themselves—transactions may be concentrated in particular projects—but they prevent the two percentages from being presented as equally broad observations.

International City Phase 1 permits a more focused, if still unpaired, comparison: its studio proxy is 7.97% on 631 sales and 10,605 leases, versus 7.26% for two-bedroom apartments on 119 sales and 2,610 leases. The 0.71-percentage-point studio lead survives holding the named place and broad apartment asset class fixed. It does not hold building, age, sale status or property quality fixed. Conversely, comparing that studio with a villa elsewhere would change both geography and asset type as well as size. The displayed community rows are selected for high ratios; their absence of lower-ranking cells cannot establish a citywide place-by-size rule.

Impact on renters, sellers and investors

For an investor, the consequential distinction is between a screening quotient and a cash-flow budget. The studio profile identifies a relatively high recorded area-rate ratio; it does not say that a purchasable studio at AED 755,701 will obtain an AED 42,299 contract. The sold sample includes off-plan property that may not be capable of earning immediate rent, whereas a rental contract describes premises that can be leased. Before comparing candidates, an investor would need ready-property sale comparables, credible lease evidence for the same building and specification, the date income could start, and unit-specific operating and acquisition costs. For a tenanted candidate, the assumed income path also needs the existing rent, lease expiry and an official DLD Rental Index result: Decree No. (43) of 2013, Articles (1) and (3) sets a banded renewal cap by the existing rent's shortfall from the RERA-approved comparator, not by the studio New-contract mean. The decree specifies 0%, 5%, 10%, 15% or 20% maximum increases in its defined bands. This article has no candidate unit's current rent or official calculation. None of these missing inputs can be replaced by the frozen aggregates, so the article cannot turn its gross ratios into observed net returns or an investment recommendation.

The basket arithmetic exposes two distinct operational exposures. Ten studios create ten possible intervals between tenancies and ten potential sets of property-level obligations; two villas put more of the illustrative AED 327,180 annual gross rent behind each individual tenancy. Losing one month on one mean-rent studio would remove about AED 3,525 from the ten-studio illustration. Losing one month on one mean-rent three-bedroom villa would remove AED 13,633 from the two-villa illustration. These are arithmetic shocks to assumed rents, not empirical probabilities. Which portfolio has greater expected vacancy loss, management expense or resale risk cannot be settled from sale and contract means.

For a seller, relevance is indirect. The buyer’s area-rate screen may frame an asking-price discussion, but it is not a valuation of that seller’s home. In particular, a ready studio seller should not be priced mechanically against a citywide sale mean that also includes off-plan transactions. Nor does a high community ratio prove a seller should expect a quick sale: the community table counts completed transactions within its cohorts, not current competing listings or time on market. The practical implication is about the evidence needed to substantiate a price claim, not a prediction that a particular layout will appreciate.

For a renter, the investment ranking is also indirect. The annual contract means describe what was registered among leased homes, not the current asking rent for an available unit. A landlord’s gross ratio cannot establish what a renewing tenant paid previously or what a new tenant will be offered. Renters are nevertheless affected by the interpretation of this evidence: treating the studio’s 5.45% as a guaranteed landlord return would conflate sale prices, signed contracts and the costs of supplying a rentable home. The supplied data support a description of recorded contract cohorts, not a justified increase for an individual tenancy.

Data in context with international markets

An international comparison needs the same distinction between a component of an investment calculation and an investment return. Eurostat’s official EU-27 House Price Index, prc_hpi_q, records 158.5 in 2025-Q1 and 166.63 in 2026-Q1 in its I15_Q index unit—a 5.13% change calculated against the 2025-Q1 value of 158.5. The supplied official API snapshot was updated 2 July 2026: Eurostat HPI, EU-27, total purchases. It is relevant to the purchase-price side of an investor’s problem: changes in that side can alter an eventual return even when rent is unchanged.

It is not, however, a foreign counterpart to Dubai’s 5.45% studio gross area-rate proxy. The Eurostat number is a year-on-year movement in a broad house-price index; the Dubai number divides independently observed annual rent and sale levels for one layout across September 2025–August 2026. The populations, windows, units and mathematical operations differ. The supplied Eurostat actual-rent HICP observations stop at August 2025, before the Dubai window begins, and are consumer-price indices rather than paired property income. Combining either foreign index with the Dubai contract or sale mean would manufacture a yield denominator. No numerical Dubai-versus-EU investor-return rank follows from these sources.

The comparison is useful precisely because it identifies the missing bridge. A cross-market investor ranking would require consistently defined purchase cohorts, rental income on comparable investable stock, timing between purchase and letting, expenses and an agreed treatment of currency and resale value. The available official international index supplies none of those as a matched return. Its 5.13% price-index change and Dubai’s 5.45% rent-to-price proxy may be numerically close, but numerical proximity is not comparability.

Summary and clear conclusion

For the September 2025–August 2026 Dubai cohorts, the falsifiable finding is that the published gross area-rate proxy declines across all five apartment layouts, from 5.45% for studios to 2.66% for four-bedroom apartments. It does not decline at every step across villas: four-bedroom villas record 4.13%, above three-bedroom villas’ 3.77%. A subsequent complete-window profile using the same sale and rental eligibility rules could test whether either ordering persists.

That is the supported unit-size conclusion. The stronger claim—that smaller purchases deliver better realised returns—is not established by independent sale and lease means, particularly when the current sale denominator includes off-plan homes. The community and budget counterexamples show how geography, sample composition, deployed capital and unobserved costs can change the answer being asked. No observed property-paired return, net-income comparison or international investor-return ranking can be extracted from these aggregates.