The final off-plan instalment is often the biggest, and it lands exactly when your cash is tied up elsewhere. What the gap costs, what happens if you miss it, and the asset-backed way to cover it at handover without selling.Abu Dhabi sales rose 163.7 per cent in the first half of 2026 while registered bank finance grew 33 per cent. The two lines sit a paragraph apart in the same release, and the gap between them is the story neither emirate's headline tells
In the first half of 2026, Abu Dhabi recorded AED 117 billion in property transactions, up 112 per cent in value against the same period a year earlier.1 Over the same six months, Dubai's residential sales came to AED 221.3 billion, down 15.7 per cent.5 That single comparison is the headline everywhere, and it is the least interesting thing in the Abu Dhabi vs Dubai real estate market data.
Dubai is still, by a wide margin, the larger market. But the direction of travel reversed, and the two emirates spent the half-year moving away from each other.
The number worth reading sits inside Abu Dhabi's own release, in two lines printed a paragraph apart: sales value rose 163.7 per cent, and mortgage volume rose 40 per cent.1
Buying grew close to threefold. Borrowing grew by a third.
The half-year in full, as published by the Abu Dhabi Real Estate Centre:12
| H1 2026, Abu Dhabi | Value | Transactions | Against H1 2025 |
|---|---|---|---|
| Sales | AED 86.1 bn | 16,838 | +163.7% by value |
| Mortgages | AED 26.7 bn | 8,876 | +40% by value |
| Musataha and long lease | AED 4 bn | ||
| Gifts | AED 311.5 m | ||
| Total | AED 117 bn | +112% value, +61.7% volume |
One note on the mortgage line, because the release and its own arithmetic do not phrase it the same way. ADREC describes the increase as more than 33 per cent. Against the published H1 2025 figure of AED 19.03 billion, AED 26.7 billion is a rise of 40 per cent, and the transaction count moves from 6,204 to 8,876, which is 43 per cent. The percentages here are calculated from the published totals rather than repeated from the summary line.3
Foreign direct investment reached AED 13.8 billion, up 309 per cent, which is more than the whole of 2025 delivered.1 Investors came from 116 nationalities against 82 a year earlier, with the United Kingdom, China, Russia, the United States, Germany and France among the largest sources.1
The access story is real. ADREC approved eight new investment zones during the half-year, bringing the total to 50, and defines them as areas open to ownership by investors of all nationalities. Those zones drew AED 75 billion, against AED 26.7 billion in H1 2025.1
Figure 3
Abu Dhabi: four consecutive half-years of growth
All property transactions, AED billion. ADREC publishes first halves and full years, so second halves are obtained by subtraction and are hatched here.
Full year 2025 closed at AED 142 billion across 42,814 transactions. This is a run rather than a single spike. Sources 1, 3, 4.
Set against the longer run, this is the fourth consecutive half-year of growth rather than a single spike. H1 2024 came in at AED 37.2 billion, H1 2025 at AED 51.7 billion, and the full year 2025 at AED 142 billion across 42,814 transactions.34 The intervening second halves are not published separately, so any H2 figure here is arithmetic rather than a reported number, and it is treated that way throughout.
Dubai spent the same six months in a different part of the cycle.
Residential sales reached AED 221.3 billion across roughly 79,200 transactions in H1 2026, down 14 per cent by count and 15.7 per cent by value against H1 2025, when the market did AED 262.1 billion across 91,900 transactions.57 For scale, the full year 2025 closed at AED 541.5 billion and more than 200,000 transactions, so H1 2026 is a step back from a record rather than a market in trouble.7
The composition tells you why. New launches in Q2 2026 came to about 5,335 units, against more than 45,000 in Q1, while H1 added 24,800 completed homes and analysts described the market as moving from launch-led to delivery-driven.6 Dubai is digesting what it already sold, which is a slower activity than selling it. That handover pressure, and the gap it opens between what buyers owe and what they have, is the delivery wave and the capital gap behind it.
It is also the emirate where the bank side moved first: through June and July 2026, three developer and bank partnerships pushed mortgage access earlier into construction instead of waiting for handover. That is a separate story, covered in where banks moved earlier into the build cycle, and it matters here for one reason. Abu Dhabi's growth is not being driven by an equivalent easing on the borrowing side.
Most comparisons of these two markets are broken before they start, because the two numbers are not the same kind of number.
ADREC reports all property transactions in Abu Dhabi: sales, mortgages, musataha and long lease, and gifts. Cavendish Maxwell reports residential sales in Dubai. Putting AED 117 billion next to AED 221.3 billion and dividing one by the other produces a ratio that means nothing, because a mortgage registration is counted on one side and not on the other.
The honest way to compare them is to let each market be measured against itself. Index both series to their own H1 2025 and watch the shape, not the size.
Figure 1
Two emirates indexed to their own H1 2025
Each market measured against itself, H1 2025 = 100. The shape is comparable, the size is not: ADREC counts all transaction types, Cavendish Maxwell counts residential sales only. Hatched bars are calculated from published full-year totals rather than reported directly.
Dubai H1 2024 is excluded: two accounts of the same Cavendish Maxwell report give incompatible growth against H1 2025, so the base is unverified. Sources 1, 3, 4, 5, 7.
Read that way, 2025 was a year when both emirates moved together. 2026 is when they parted. And nothing in the index requires anyone to pretend that a market roughly a third of Dubai's size, tripling from a smaller base, is the same event as a larger market pausing after a record.
Now the arithmetic that no one else runs. Take Abu Dhabi's two published lines and divide one into the other.
Figure 2
Abu Dhabi: buying tripled, registered bank finance grew by a third
AED billion per half-year, from the same ADREC release. The two lines are printed a paragraph apart and are rarely read together.
58%
Registered mortgage value as a share of sales value, H1 2025
31%
The same ratio, H1 2026
ADREC's mortgage line records charge registrations, which include refinancing and borrowing against property already owned. The precise claim is that registered bank finance fell behind purchase activity, not that the share of mortgage buyers collapsed. Sources 1, 3.
In twelve months, the share of purchase value matched by registered bank finance fell by nearly half. Sales grew 163.7 per cent. Mortgage registrations grew 40 per cent. The gap between what people bought and what banks wrote against it widened by a factor of two.
One caveat belongs here, and it cuts against the drama rather than for it. ADREC's mortgage line records charge registrations, which include refinancing and borrowing against property someone already owns, not only purchases made with a mortgage. So the precise claim is that registered bank finance fell behind purchase activity, not that the share of mortgage buyers collapsed. The direction survives the caveat. Even counting refinancing on the generous side, the banking system wrote a third more while the market bought nearly three times as much.
Dubai has its own version of this, measured differently and covered separately in what the cash vs bank finance numbers show. The pattern rhymes across both emirates: at the top of a market, purchases run ahead of what banks will carry.
Here is the part that surprises people who read the access story and assume the rest followed. Abu Dhabi opened 50 investment zones and welcomed buyers from 116 countries. It did not, and could not, change what a bank is allowed to advance, because those rules are federal and identical in both emirates.
Three ceilings do most of the work.
The off-plan ceiling. For a purchase before completion, the Central Bank caps advances at 50 per cent of value, regardless of the buyer's category or the price of the property.8 Half the purchase has to exist as cash before a bank participates at all.
The income ceiling. Total debt service is capped at 50 per cent of gross income, and the calculation is unforgiving: card exposure counts against the limit rather than the balance actually owed. A buyer with a clean record and a strong asset can fail this test on paper.
The non-resident ceiling. For buyers who live elsewhere, banks typically advance 50 to 65 per cent of value. This one is not a Central Bank rule. It is bank policy, which is why it varies between institutions and why the same file can get two different answers.
LTV is the industry shorthand for the share of value a bank will advance, and every one of these ceilings is a rule about the borrower rather than about the property. That is the whole point. A market can triple the number of nationalities allowed to buy, and the file each of those buyers has to produce stays exactly as demanding as before. Whole categories of solvent people sit outside it for reasons that have nothing to do with what they own, which is the same mechanism at work in why banks underserve whole segments.
One live example shows how carefully this has to be read. In July 2026, Modon and ADIB announced an off-plan programme covering up to 75 per cent of value, structured with roughly 15 per cent paid during construction. The Central Bank's off-plan cap is 50 per cent. Both statements are on the record, and no public source reconciles them. They are presented here as they stand, without an explanation invented to join them.
Where the ceiling stops short of the price, the difference has to come from somewhere other than a bank. NEMAX reviews the property itself and confirms what it can support, in either emirate.
Put the two halves together and the position is specific rather than gloomy.
Abu Dhabi has genuinely opened. Fifty zones, AED 75 billion into them in six months, foreign direct investment up more than fourfold. If the question is whether a foreign buyer can own property there, the answer changed, and the numbers show people acting on it.
What did not change is the paperwork behind the purchase. A non-resident buying off-plan in an investment zone meets the 50 per cent construction-stage ceiling like everyone else, then meets a bank policy that advances 50 to 65 per cent of value once the property is complete, then has an income test applied to income earned in another currency and another jurisdiction. None of those tests looks at the property.
That is what the 58 to 31 per cent shift looks like on the ground: thousands of purchases completed with money that came from somewhere other than a mortgage, out of savings, out of assets sold elsewhere, out of capital raised against property the buyer already owned. The market grew into the space the banking system did not cover.
NEMAX works from the asset rather than the borrower's file. A property in the UAE with clear value and clean title can support capital against it, structured through an SPV, which is simply a separate company set up to hold the asset, with a first charge registered against the property. That mechanism is set out end to end in raising capital against a property you already own.
It fits a specific shape of problem. A deadline exists. A property exists. The two do not meet, because the bank's ceiling or the bank's timetable falls short. Deal sizes start around AED 1 million, at up to 70 per cent of valuation, against a target of thirty days to closing.
It does not fit everything, and saying so plainly is more useful than a longer list of features. This is short-term capital priced as short-term capital. It is not a cheaper mortgage, and anyone comparing it to a twenty-five-year product on price alone is comparing two different instruments. If the timeline is comfortable and the file is bankable, the bank is the right answer, and NEMAX will say so. It also does not apply to property outside the UAE, and it is not a route to buy an asset with no equity in it at all.
No. Dubai's residential sales alone reached AED 221.3 billion in H1 2026, against AED 117 billion across all Abu Dhabi transaction types.15 Abu Dhabi grew more quickly from a smaller base, which is a different statement from being larger.
Sales value rose 163.7 per cent, foreign direct investment rose 309 per cent, and eight new investment zones opened, taking the total to 50 and drawing AED 75 billion.1 Growth came mainly from new buyers entering, rather than from cheaper access to bank finance.
Transactions fell 14 per cent by count and 15.7 per cent by value in H1 2026 against a record H1 2025, while completions ran at their highest level in years.56 The phase changed from launch-led to delivery-driven. A step back from a record is not the same as a decline.
Yes, subject to the same federal rules that apply in Dubai. Off-plan purchases are capped at 50 per cent of value, debt service at 50 per cent of gross income, and non-residents are typically advanced 50 to 65 per cent by bank policy rather than by regulation.8
Areas designated by ADREC as open to ownership by investors of all nationalities. Eight were approved in H1 2026, bringing the total to 50.1 The lists of named freehold districts that circulate on agency sites are popular summaries, not the official register.
Because the ceilings are set on the borrower rather than the asset, and they did not move while the market did. Off-plan advances are capped at half of value, income tests are applied to gross income, and non-resident policy tightens the share further.8
The federal ceilings are the same in Abu Dhabi and Dubai, and so is the arithmetic they leave behind. NEMAX looks at the asset, its value and the timeline of the deal, then confirms what property-backed co-financing can carry. One review, one answer, no change of ownership.
Apply for capital →
.png){kind=logo}
We use cookies to improve website performance and analytics. By continuing to use the site, you agree to their use. Learn more