Dubai’s New–Renewal Rent Gap: Pressure or Catch-Up?

A hot rental market can pull new leases away from renewals; a downturn can close that gap. But a large remaining gap is also a memory of earlier increases. Dubai’s citywide apartment gap narrowed in 2025–26 even as both rent series rose. In a fixed panel of 33 one-bedroom markets, narrowing had already begun a year earlier. For tenants and investors, the direction and the two underlying rents matter more than the headline spread.

The question: what does the gap tell us about the market?

The useful question is not simply whether New leases cost more than Renew leases. It is whether their separation tells a renter something about the cost of moving, or an investor something about the income already attached to a tenanted property. To answer it, we need to distinguish a market accelerating today from one still carrying the effects of yesterday’s boom.

The test uses DLD-derived apartment registrations in ten complete September–August windows ending 2017–2026. Each rate is mean annual contractual rent per measured square foot, with New and Renew kept separate. The gap is New rate divided by Renew rate, minus one. It compares independently registered contracts, not the same apartment before and after a tenant moves. We then repeat the test with one-bedroom apartments in a fixed group of communities to reduce changes in geography and bedroom mix.

The downturn is a clear starting point. In 2016–17, New apartments averaged AED 84.07/sqft/year, versus AED 73.25 for Renew: a 14.8% gap. By 2020–21, the rates were approximately AED 53.60 and AED 53.95, leaving a −0.6% gap. New rents had fallen 36.2%, faster than the 26.4% decline in renewals. Entering the market no longer carried a positive citywide group-average premium.

That narrowing did not start with COVID. The gap was already 2.2% in 2018–19 and approximately zero in 2019–20. The pandemic-period trough is part of the observed history, not a sufficient explanation for a decline underway beforehand. This distinction matters: attributing the entire cycle to one event would miss the earlier adjustment visible in the registrations.

The rebound reversed the relative movement. In 2021–22, New rates rose 11.0% while Renew rates fell 0.6%; the gap reopened to 11.0%. By 2024–25, it had reached 34.1%. Across the rebound from 2020–21, New rates increased 76.5%, against 30.7% for Renew. Those differences establish that entry-market rents moved much further than the renewing-contract average.

The latest window supplies the crucial qualification. In 2025–26, New rents rose another 3.0%, but renewals rose 6.3%. The gap narrowed to 30.0%, although neither rate fell. A reader looking only at the still-large 30% spread could mistake accumulated separation for continuing acceleration. A reader looking only at its decline could incorrectly conclude that new apartments had become cheaper.

The mechanism: new contracts reset; renewals catch up

A newly negotiated lease encounters the market at entry. A renewal starts from an existing contractual rent and a continuing landlord–tenant relationship. When entry prices rise sharply, a constrained or slower adjustment of continuing contracts can leave renewals behind. When entry rents weaken, the outside option improves and the gap can close. This is an economic explanation to test against the data, not a claim that every landlord or tenant behaves identically.

Dubai adds an explicit regulatory mechanism. Decree No. (43) of 2013 sets maximum renewal increases according to how far the existing rent lies below the RERA-approved rent for similar units. Those bands are not an instruction to reset every lease to whatever a new tenant might pay. For example, an existing rent of AED 80,000 against a hypothetical official comparator of AED 100,000 is 20% below that comparator. The decree’s 11–20% band specifies a maximum increase of 5% of the existing rent: AED 4,000, taking it to AED 84,000, subject to the contract and notice requirements.

That example explains why a spread can persist during rapid repricing. Even an increase up to the relevant maximum need not close the distance to the comparator in one renewal. It does not establish that regulation caused 30 percentage points of this citywide gap: we do not have the official index result, previous lease and next lease for each apartment. Different buildings can also enter the New and Renew samples.

The mechanism is not unique to Dubai. The Cleveland Fed’s Disentangling Rent Index Differences research constructs repeat-rent indices from BLS microdata and finds that new-tenant rent inflation led official BLS rent inflation by four quarters in its study. That is meaningful supporting evidence for slower adjustment across the wider tenant population, unlike comparing our spread with an unrelated foreign CPI number. It is not a four-quarter Dubai forecast: the US study’s population, index construction and institutions differ from these unpaired DLD means.

There is also an important institutional change within our observation period. DLD launched its Smart Rental Index in January 2025, describing a building-classification approach that considers structure, finishes, maintenance, location and services. The legal comparator should therefore not be treated as simply our community-average New rent. Updating the index could influence renewal adjustment; our series cannot separate that effect from market cooling or changes in the registered stock. The contemporary framework is a reason to check the actual building’s index result, not apply an old area average mechanically.

The gap also has an elementary limitation as an indicator. It widens whenever the New mean grows faster than the Renew mean; that relationship follows from the ratio itself. A correlation between gap movement and New-rent growth is not independent proof of rising demand. We need the level of both rates, their changes and evidence that the pattern survives a more consistent sample. That is what the next test supplies.

The test: does the cycle survive holding the market mix fixed?

The earlier citywide comparison could be distorted by more New registrations in expensive neighbourhoods or by a changing balance of studios and larger apartments. We therefore build a balanced panel of 33 named one-bedroom markets from the already-generated monthly files. A place qualifies only if every one of the six September–August windows ending 2021–2026 contains all twelve months and at least 100 New and 100 Renew contracts. Admission does not depend on whether its gap widens or narrows.

We give each place the same weight in every year and on both sides of the comparison: its combined New-plus-Renew contract count in 2020–21. Holding those weights and the one-bedroom layout fixed removes shifts in the relative representation of these communities from the calculation. The latest 66 place/registration groups reconcile in counts and mean rates to the frozen main extract. Buildings, finishes and individual tenants remain unpaired, so this is a geographic/layout check rather than a same-home index.

The rebound survives. All 33 places have a larger gap in 2024–25 than in 2020–21. The fixed-weight panel gap rises from −3.8% to 18.6% between those endpoints. Thus a change in which of these communities trades more heavily cannot be the sole explanation for the widening. The repeated local pattern supports the interpretation that New and Renew contracts adjusted differently through the cycle.

But the panel also changes the timing of the story. Its gap peaks at 22.2% in 2023–24, then falls to 18.6% in 2024–25 and 13.6% in 2025–26. The all-layout citywide gap peaks a year later, in 2024–25. The citywide number therefore conceals an earlier turn among these established, consistently sampled one-bedroom markets. Neither timing is a statistical error: the series cover different populations. The panel excludes newer and thinner markets rather than representing every Dubai apartment. Its earlier peak is not an accident of the chosen baseline: equal community weights and fixed pooled weights from each of the other five years all retain the 2023–24 peak.

The latest cooling is broad, not just a JVC anecdote. 26 of 33 local gaps narrow from 2024–25 to 2025–26. In 18 of those 26, New rents nevertheless increase. Renewals catching up to a more slowly rising entry market is therefore a common observed pattern. It is more accurate than saying either “the rental boom continues because the gap is high” or “rents are falling because the gap is shrinking.”

Today’s local comparison also differs from the headline. Across 129 qualifying named-community and bedroom cells in the latest window, the median gap is 13.4%, and three are negative. That median gives each cell one vote; it is not the citywide 30% ratio with a different rounding rule. The New and Renew citywide means can have very different geographic and bedroom weights. A household should not budget a 30% moving premium simply because that is the aggregate result.

The decision: staying, moving and buying a tenanted home

For a tenant, compare actual alternatives—not the headline gap

During a sharp upswing, the gap warns that re-entering the market may be expensive relative to a continuing contract. During cooling, it is a reason to revisit the alternatives rather than assume last year’s entry premium still applies. Neither signal determines the rent on a particular available home. Obtain an actual offer and compare it with the permitted renewal of the home you occupy.

Suppose the household’s property-specific renewal outcome is AED 84,000 and a move entails AED 6,000 of nonrecoverable one-off costs. A New lease at AED 80,000 is cheaper in annual rent but costs AED 86,000 in the first year including the move—AED 2,000 more than staying. The first-year break-even New rent is AED 78,000. These are explicit household assumptions, not measured Dubai fees or current listings. Refundable deposits belong in the upfront cash budget, not automatically in expenses consumed by the move.

Local direction helps decide when to investigate such offers. JVC one-bedrooms’ gap fell from 28.2% to 17.4% between 2024–25 and 2025–26; Marina’s fell from 14.1% to 7.4%. Business Bay’s gap had peaked earlier, at 40.9% in 2022–23. These places have different histories, which is precisely why a universal “hot above 20%” threshold would mislead. The four-place chart illustrates the broader panel test; those four were not selected to determine its outcome.

The market comparison is not the legal calculation

Use DLD’s current Rental Index service with the actual tenancy and property details. Its published procedure lists expiry date, property type, region, number of rooms and existing annual rent. Use the official output for the property, including any current building-specific requirements. Our community statistics are neither the RERA comparator nor an alternative calculator.

The statutory bands are shown below as written in Decree No. (43), Article (1). They use the rent’s shortfall relative to the official similar-unit comparator; the allowed percentage then applies to the existing rent. Do not infer a fractional boundary rule from the chart—use DLD’s calculation for the particular case. Under Law No. (26) of 2007 as amended by Law No. (33) of 2008, Article (14) requires at least 90 days’ notice before expiry to propose changed terms, unless otherwise agreed. Article (13) provides for reviewing rent upward or downward at renewal and for adjudication where agreement fails.

A larger potential New rent is not itself a statutory eviction ground. Amended Article (25)(2) lists specified end-of-term grounds including qualifying reconstruction or major works, qualifying personal/family use, and sale. It requires notice of those reasons at least twelve months before eviction, through a notary public or registered post. That is separate from the 90-day term-change notice; before-expiry breaches have their own provisions. A personal-use possession award also carries re-letting restrictions under amended Article (26). An actual dispute requires the relevant documents and case-specific advice, not a decision based on a graph.

For an investor, the spread is an underwriting risk

A buyer of a tenanted apartment acquires the income and obligations of its actual tenancy, not the New mean from a different building mix. A large gap should prompt scrutiny of an agent’s “market-rent upside” calculation: when could that rent lawfully be obtained, what would the property actually achieve, and what costs or vacancy would intervene? The narrowing observed in 2025–26 is a warning against assuming a boom-period spread is a permanent gain waiting to be collected.

A seller can support an income-based valuation with the actual lease, verified renewal position and genuinely comparable lettings. Calling the citywide spread guaranteed future income overstates the asset. Conversely, an occupied unit with a lower rent is not automatically worth less by the same percentage: valuation also depends on purchase price, lawful income timing, expenses and the buyer’s holding period.

Conclusion: read the gap with its direction, not alone

The evidence supports the market-cycle interpretation, with an important refinement. Faster New-rent increases widened the gap after the trough; faster New-rent declines had closed it during the downturn. But a high gap can persist after the entry market slows because continuing contracts are still catching up. In 2025–26 the gap narrowed in 26 of 33 consistently sampled one-bedroom places, including 18 where New rents still rose.

The decision rule is therefore to read three things together: the gap, New-rent growth and renewal growth, then check the relevant local property class. A widening spread in a rising entry market signals increasing separation between movers and continuing contracts. A narrowing spread with rising New rents signals catch-up rather than an outright decline. Falling New rents add evidence of genuine entry-price contraction. The gap alone cannot distinguish those states or predict the next turning point.

For a tenant, that changes how to evaluate staying against moving. For an investor, it changes whether a proposed future rent is credible. For neither does the spread override the actual tenancy or RERA framework. This is what makes the ratio useful: not a label of “hot” or “cold,” but a way to distinguish current entry pressure from the delayed adjustment left by the previous cycle.