JVC lease-up fell 10 points. How full is JVC really?

Here is the question I get most often from JVC owners. My unit is ready. Where is my tenant?

So I went and counted. Take the towers in Jumeirah Village Circle that began letting from 2018 to 2022. One year in, 58.0% of their homes had a tenant. Now take the towers that began letting from 2023 to 2025. They reached 47.9%.

That is a drop of 10.1 points. JVC lease-up is slower than it used to be.

Nobody publishes this number. So I built it from the registered record, and I will show you the working.

One thing before we start

Every figure on this page comes from JVC data only. I counted JVC buildings, JVC leases and JVC sales.

So please do not read it as a Dubai-wide finding. Other communities may run faster or slower. I have not measured them here, so I will not pretend I have.

What you get instead is one community, measured properly.

What you will learn on this page

Three things, and you can act on all of them.

First, you will learn how long a new JVC tower takes to fill, and how much that has changed.

Second, you will learn how long your own off-plan home may wait for a tenant. The honest answer is longer than most payment plans assume.

Third, you will learn why nobody can tell you the JVC occupancy rate. Not me, not a broker, not a portal. I will show you exactly where the measurement stops.

What JVC lease-up actually measures

Lease-up is a simple idea. It counts how many homes in a building have found a tenant, and how quickly.

The idea is simple. The measuring is not.

Where these numbers come from

Dubai runs a lease registration system called Ejari. The Dubai Land Department operates it. Every contract that goes through it lands in a public record.

I pulled that record through DXB Interact on 13 August 2026. Then I built my own clean file from the raw rows. I never trust a summary sheet.

That gives me 145,232 clean apartment leases for JVC. They go back to February 2013.

The other half of the sum is building stock. Property Monitor is a Dubai property data firm. It publishes a list of homes, building by building. Its JVC list holds 91,028 homes in 627 buildings.

So the sum is simple. Registered leases, divided by homes.

Why a registered lease is only half the story

Both halves have a problem. And both problems push the same way.

A registered lease proves a tenant. That part is solid.

But a missing lease proves nothing at all. Plenty of leases never reach the register. Many owners live in their own homes. Others run short stays, which the register does not carry.

And 25.5% of JVC leases carry no building name. So they drop out of every building-level count.

So every lease-up number on this page is a floor. The real figure sits higher. How much higher, nobody knows.

A missing lease is not an empty home

This is where most Dubai rental analysis goes wrong, and my own earlier work went wrong too.

If a home has no registered lease, the honest reading is this: we cannot see one. It is not proof the home sits empty.

That reading cuts against the scary version of the story. Good. It should.

So I will not print a vacancy rate anywhere on this page. Not in the text, and not on a single chart. The data cannot carry one.

Then why does the comparison still work?

Fair question. A floor is useless as a level. But it works fine as a comparison, as long as the floor sits at the same height over time.

So I tested it.

I took 35 buildings that were already letting by June 2018. Together they hold 4,940 homes. Then I froze that set. And I read it on eight 30 June dates in a row.

The readings run 43.9%, 40.8%, 49.8%, 48.4%, 47.9%, 49.0%, 47.3% and 45.6%.

That is a nine point band. No trend sits inside it. In short, filing habits in the same older buildings held steady for eight years.

There is a second check, and it points the same way. The share of new leases carrying a building name rose from 69.8% in 2022 to 82.9% in 2025. Naming got better, not worse.

Better naming lifts recent numbers. So the decline you are about to read is, if anything, smaller than the real one.

JVC building absorption fell across two cohort groups

Now the counting.

I grouped every JVC building by the year of its first lease. Then I followed each one for twelve months.

A building needs a full year of history to join the count. So a tower that opened last month cannot drag the number down.

That leaves 175 buildings and 30,179 homes. Here is what they did.

The eight cohorts, year by year

First let	Buildings	Homes	Let at 12 months	95% range
2018	9	1,114	56.4%	35.2 to 78.3
2019	19	1,977	66.2%	60.9 to 72.2
2020	13	2,346	47.4%	26.8 to 66.7
2021	18	3,643	57.7%	48.4 to 69.4
2022	14	1,892	64.3%	56.7 to 70.2
2023	11	1,499	51.2%	40.9 to 62.7
2024	21	4,450	50.4%	46.4 to 54.6
2025	18	4,715	44.5%	35.4 to 53.8

Read the last column first

Most people read the percentage and stop. Please do not.

Each group holds only nine to twenty one buildings. So each number carries a wide range. That range is the honest part.

Here is the test. You can pair these years in 28 ways. Only four pairs come apart cleanly. All four set an early year against a late one: 2019 or 2022, against 2024 or 2025.

The other 24 pairs share ground. So anyone selling you a clean three year slide is selling you noise.

So what does hold up?

Pool the years instead. That is the fair test.

Buildings first let from 2018 to 2022 reached 58.0% at twelve months. That is 73 buildings and 10,972 homes.

Buildings first let from 2023 to 2025 reached 47.9%, across 50 buildings and 10,664 homes.

The gap is 10.1 points. Its 95% range runs from 2.2 to 18.0 points.

Then I ran the test again 8,000 times, each time on a random draw of the same buildings. The later group came out slower in 99.4% of those runs.

Then I dropped every building one at a time. The answer did not move.

So here is the honest sentence. New JVC buildings fill about ten points slower at the one year mark. That is a real change, not a wobble.

The shape is not a smooth slide

Look at 2019 and 2020. Lease-up went from 66.2% down to 47.4%, then back up to 57.7% and 64.3%.

Nobody should read that as a trend. That is a market that stopped, then restarted.

The recent numbers are different in kind. Three cohorts in a row now sit below every pre-2023 cohort except 2020. And 2025, at 44.5%, is the lowest of the eight.

One more warning about ranges

I owe you one more caution here, because it cuts against a lazy reading of my own chart.

The two pooled ranges still touch, from 51.9% to 52.6%.

Some people will see that and say the gap is not real. They would be wrong, and it is worth knowing why.

Two ranges can share a little ground and still mean two different things. The right test is not whether they touch.

The right test is the gap itself, measured directly. That gap came out real 99.4% of the time.

Off-plan handover to tenant takes longer than most buyers plan for

Now the question owners actually ask. I bought off-plan, so when does rent start?

I matched 49,998 clean off-plan apartment sales to the first lease filed on that same home.

Then I kept only the 9,581 purchases with five years or more behind them. That way the answer is not cut short.

Two medians, and they answer different questions

This part matters, so I will be slow about it.

Take only the homes that find a tenant. Their middle wait runs 32.7 months from purchase. That is just under three years.

But that figure quietly drops every home that never lets. So count them all instead. On that basis, half have a tenant by month 45.

Two in five have one by month 36. And 61.5% have one by month 60.

The gap between 32.7 months and 45 months is the quarter that never appears. I will come to them next.

The spread is wider than the average suggests

Averages hide the thing you need. So here is the spread.

Among homes that do let, the middle half waited 18.1 to 52.7 months.

So one buyer in four had a tenant inside a year and a half. Another one in four was still waiting after four and a half years.

Same community. Same years. And the same product.

Year two does most of the work

The chance of landing a tenant is low at first. Then it jumps. Then it holds flat for years.

In the first year after purchase, 9.7% of homes find their first tenant. In year two, that rises to 18.5%.

It peaks at 20.9% in year three. Then it eases only a little, to 19.3% and 18.0% in years four and five.

So most homes land in year two. But years four and five still land homes at close to the year three rate.

That is the useful part. A home still empty after three years is not a lost cause. It sits in the slow tail, and the slow tail runs long.

The quarter that never registers a tenant

Now the number that stops people.

Of off-plan homes bought at least five years ago, 24.6% have never had a registered lease. That is 2,353 homes out of 9,581.

The number is stable

I tried hard to break it.

Run it at four years and you get 25.0%. At six years, 25.2%. At seven years, 24.9%. Change the way buildings are matched and it does not move either.

So it is not a quirk of one window or one rule.

What that quarter actually contains

Now read it properly. This is a measurement, not a verdict.

Owners live in part of that quarter. Another part runs as short stays. A third part sits on a lease nobody filed. And yes, some of it does stand empty, held by an owner who never got round to letting.

I cannot split that quarter four ways. Nobody can, from public records.

What it is not

Here is what I can say. About one off-plan home in four goes five years without once appearing in the lease register.

Anyone who quotes that as a vacancy rate has misread it. Please do not let them.

Thinking about an off-plan purchase? Test the rent side before you sign, not after handover. That one hour is the cheapest hour in the whole process.

My guide on what to know before buying off-plan in Dubai covers the rest of that checklist.

The JVC occupancy rate nobody can publish

This is the question I get most, and this is the most honest answer I have.

One number, and what it means

On 30 June 2026, JVC held 24,264 live filed leases on apartments. Property Monitor counts 44,210 ready apartments in JVC.

So 54.9% of ready JVC apartments had a live filed lease that day.

That is a floor. It is the only occupancy figure on this page. The real number sits higher, and I do not know by how much.

New building occupancy in JVC is measured, not counted

Split that same floor by when a building started letting. The pattern is flatter than you might expect.

Building age	Buildings	Homes	Live registered leases	Floor
2017 or earlier	28	4,048	1,982	49.0%
2018 to 2019	28	3,091	1,623	52.5%
2020 to 2021	31	5,989	2,363	39.5%
2022 to 2023	25	3,391	1,731	51.0%
2024 to 2026	61	13,382	5,143	38.4%

The newest band sits lowest, at 38.4%. A lease-up story predicts exactly that.

But look at the 2020 to 2021 band. It sits almost as low, after five full years. So age explains part of this. It does not explain all of it.

I publish this by age band only, and never by building. A building list would rank towers.

And a rank built on a floor punishes any tower whose owners file less paperwork. That is not a quality score. It is a paperwork score.

The bottom of the sum is shaky too

The top of the sum is a floor. Now look underneath. That half is a range.

Property Monitor does not agree with itself. Its JVC tiles add up to 83,331 homes once offices and shops come out. Its own building list holds 91,028.

That is a gap of 7,697 homes inside one company.

Say which base you use, because it changes the number. Against the tile total, 7,697 homes is 9.2%. Against the building list, the same gap is 8.5%.

Either way, it beats the gap between that company and the government register.

A third of JVC homes never appear at all

Our own evidence is worse still, and I would rather you heard it from me.

Property Monitor reports 47,823 ready homes in JVC. Only 33,589 homes ever show up in the record at all. That means either a matched lease or a finished sale.

So 29.8% of delivered JVC homes never surface in the data. Not once, in fourteen years of records.

That is the real reason nobody can publish a JVC occupancy rate. The answer is not shameful. It is simply that a third of the homes stay invisible. And the size of JVC itself carries a nine percent margin.

Article 02 in this series hits the same wall on the supply side. There, the JVC pipeline holds 35,120 homes and the published counts still disagree.

JVC rental take-up: what the register can measure about empty homes

So the register cannot tell you what share of homes sit empty. Fine. But it can tell you something close, and useful.

It can measure the gap between two tenants in the same home. Both ends of that gap are filed events. So the wait between them is real.

I ran that test in two other Dubai districts, not in JVC. The middle wait between tenants runs from 97 days in Dubai Creek Harbour to 152 days for a Business Bay one-bed. Those records run to 13 May 2026.

So treat those two numbers as a method example, not as a JVC figure.

You can read that work in how long a Dubai apartment sits empty.

That is the difference worth holding on to. How long a home waits between tenants is measurable. What share of homes are empty right now is not.

The escrow break, and what the register cannot date

One more thing the register shows. And this one is good news.

Where the cancelled projects sit

The Dubai Land Department register holds 541 JVC projects. Of those, 251 finished. Another 139 are still being built. And 151 carry a cancelled mark.

So 27.9% were cancelled, counting all three cancelled marks.

A tighter count uses only the projects marked Cancelled. That gives 149, or 27.5%. Both stand up, so I will tell you which one I use.

The cliff is real

Now sort those 541 projects by registration number.

Of the first 250, 149 were cancelled. That is 59.6%. Of the last 250, beginning at project number 2011, exactly one was.

Project numbers work as a running order. I checked them against sale dates.

On the 258 projects that carry both a number and a first off-plan sale, the two line up at 0.94 out of 1. So higher numbers came later.

The cancelled share then collapses as the numbers climb. Projects numbered under 500 were cancelled 62.0% of the time. The 500 to 999 band did worse, at 78.1%. Then 38.5%, then 9.4%. After that, almost nothing. One later band shows 1.9%. Every other band shows zero.

What the law says

Dubai’s escrow regime is the obvious candidate.

The Ruler of Dubai issued Law No. 8 of 2007 on 6 May 2007. It covers escrow accounts for property developers.

It forces off-plan buyer money into a project account. A licensed bank or agent holds it.

Two articles carry the weight. Article 9 keeps that money away from anyone the developer owes. Article 14 makes the agent hold back 5% of the account. It stays held for a year after the homes are signed over.

Two later rules built the clean-up machine. Decree No. 21 of 2013 set up a special court for cancelled projects.

Decree No. 33 of 2020 replaced it on 24 November 2020. The new court also covers half-built projects. It can order refunds from escrow. It can also hand a stalled project to a new developer.

What I still cannot prove

Three reasons, and I will not pretend otherwise.

First, the register carries no registration date. It carries project numbers only. So the cliff sits somewhere between two number bands, not on a calendar day.

Second, the failed projects left almost no trace. Of 151 cancelled projects, only 18 ever filed a single sale. The other 133 never sold a home here. So nothing exists to date them by.

Third, 2008 happened. The escrow law and the crash land in the same window. The crash alone would have wiped out the weak developers. Two causes, one cliff, and no way to split them here.

One more caution, and it favours the good news

Recent projects have had less time to fail.

A project registered in 2024 cannot yet show a ten year cancellation. So read the near-zero recent rate as encouraging rather than settled.

What could make me wrong

Five things, in the order they worry me.

These numbers are JVC only

I measured one community. I did not measure Dubai.

So nothing here tells you how a tower in Business Bay, JLT or Dubai Marina fills. Those markets have their own supply, their own tenants and their own timing. If you want the same test run on your community, ask me and I will run it on that community’s own records.

The cohorts are small

Nine to twenty one buildings per year. One unusual tower moves a whole year. That is exactly why I pooled the years and published the ranges.

The building frame is my own

Earlier work in this series used a tighter set of 149 buildings. It landed within 1.6 points of these numbers.

I could not rebuild that set from any written rule. So I built my own and said so. Rebuild it another way and your number will shift a little. Your answer will not.

Coverage could shift later

My fixed panel test covers 2019 to 2026. It says nothing about 2027. If filing habits change, the comparison weakens.

The newest cohort is the thinnest

The 2025 cohort has the least history and the widest range of them all. Treat it gently.

What to do with this

Three readers, three different answers.

If you are buying off-plan

Plan for 32.7 months to your first tenant. Then remember that figure counts only the homes that get one. Across every purchase, half wait 45 months.

So do not build a cash flow that starts at handover. The register says rent does not start there. Match your payment plan to the letting clock, not to the brochure.

If you already own

A home still empty after three years lands tenants at close to the third year rate.

That argues for patience and for pricing. It does not argue for panic.

If you are letting

Recent buildings fill about ten points slower at twelve months than buildings did before 2023.

So price the first tenant carefully. That first rent sets the starting point for every renewal after it.

Buying a home that already has a tenant? Read what you actually inherit first.

Frequently asked questions

Where this sits in the series

This article closes the supply act.

Article 01 covers the 2026 JVC sales decline. Article 02 covers the JVC supply pipeline and the project register.

Want the wider picture first? The market insights index carries every piece I have published. And how to tell if a Dubai property is a good deal is the shortest place to start.

The same lesson, in another market

One more piece is worth your time. It is the same idea somewhere else.

In Mina Rashid, a quarter of the resale record carries no price at all. Different district, same lesson. Check what the record can actually see before you trust an average.

Work with me

I am Fahad Al Kuwari, a buyer’s consultant for Dubai property. I work from the filed record, not the marketing deck.

And I will tell you plainly when the record cannot answer your question. I have done that three times on this page.

Want a JVC purchase or a rent roll tested against these numbers? Reach me at fahadalkuwari.com.