Dubai ranks fourth in the UBS bubble index, ten places higher than in 2024. The price level is only part of the reason.
UBS published the 2026 edition of its Global Real Estate Bubble Index on 26 August 2026.
UBS published the 2026 edition of its Global Real Estate Bubble Index on 26 August 2026. Dubai scores 1.16 and ranks fourth of 23 cities, behind Zurich, Tokyo and Miami. Two years earlier Dubai scored 0.64 and ranked fourteenth of 25. The city moved from the moderate band into the elevated band, and the score is still rising year on year.
The headline invites a simple reading: prices ran away. The data says something more specific. Dubai's real home prices rose 0.4 % in the year to the second quarter of 2026, close to flat. What moved the score is the gap that opened between the price level and the two things that used to justify it.
How Dubai's sub-indices changed — UBS Global Real Estate Bubble Index, 2024 edition against 2026 edition
| Sub-index | 2024 | 2026 |
|---|---|---|
| Price to income | moderate | elevated |
| Price to rent | elevated | elevated |
| Real price index | moderate | high |
| Mortgage to GDP | low | low |
| Construction to GDP | low | moderate |
For Dubai, Singapore and Hong Kong, UBS replaces the city-to-country price ratio with an inflation-adjusted price index.
Score 2024: 0.64 · rank 14 of 25 · Score 2026: 1.16 · rank 4 of 23 · Bands: <0.5 low · 0.5–1.0 moderate · 1.0–1.5 elevated · >1.5 high
What the index actually measures. The score is a weighted average of five standardised sub-indices: price to income, price to rent, the change in mortgage debt relative to GDP, the change in construction activity relative to GDP, and the city-to-country price ratio. For Dubai, Singapore and Hong Kong, UBS replaces that last one with an inflation-adjusted price index, because a city-to-country comparison carries little meaning in a city state or an emirate that dominates its own national market.
UBS is explicit about what the number is not. The index does not predict whether or when a correction happens. Bubble risk means the likelihood of a substantial price correction, measured against historical bubble episodes. A score of 1.16 is a position on a distribution, not a forecast.
Only one sub-index turned red. Of the five components, exactly one moved into the high band: the inflation-adjusted price index. Price to income moved from moderate to elevated. Construction moved from low to moderate. Price to rent did not change. Mortgage debt to GDP stayed low, where it was two years ago.
That last point deserves weight. The classic bubble pattern is a price level carried by credit. Dubai does not show it. The mortgage sub-index has been in the lowest band through the entire run-up. Whatever risk sits in this market, it is not leveraged household debt.
Why 2024 looked moderate at plus 16.8 %. In the 2024 edition Dubai posted the strongest real price growth in the study, 16.8 % in a single year, and still ranked fourteenth. The reason is in the denominators. Real rents rose 16.6 % over the same year and roughly 60 % since 2020, against a price increase of about 40 %. Rental yields stood at 6 to 7 %. Household income growth in the Emirates outpaced every other city UBS measured. Price, rent and income rose together, and ratios built from three numbers moving in step barely shift, whatever the absolute pace.
What the last two years did. Prices continued. Real growth exceeded 10 % in 2025. Then, as UBS puts it, the boom came to an abrupt halt at the onset of the conflict involving Iran. Real prices have fallen back to mid-2025 levels, and in the year to the second quarter of 2026 they are 0.4 % higher than a year earlier.
Rents went the other way. Real rents are 4.0 % below their year-earlier level. The sequence matters: rents carried the price level in 2024, and in 2026 they are subtracting from it. A flat price against a falling rent widens the price to rent ratio without a single transaction at a higher price. Price to income moved a full band while prices stood still, which means incomes stopped outrunning them.
The ten-year window moved too. The 2024 edition reported a ten-year real price change of minus 1.3 % a year. The 2026 edition reports plus 2.2 % a year. That is a swing of 3.5 percentage points annually on a decade-long series, and it happened without anything new occurring in 2026. The explanation is the window: the weak years from 2014 to 2020 have rolled out of the ten-year calculation. The same market, measured the same way, now shows a decade of real appreciation where it previously showed a decade of real decline.
Part of the climb belongs to the other cities. Dubai's own score rose 0.52 points. The rank rose ten places. The difference is the field. Of the ten cities that stood above Dubai in 2024, eight scored lower in 2026. Miami fell from 1.79 to 1.41, Los Angeles from 1.17 to 0.69, Toronto from 1.03 to 0.63, Vancouver from 0.77 to 0.62, Hong Kong from 0.74 to 0.61. Two more, Boston and Tel Aviv, are not in the 2026 study at all; Seoul and Lisbon were added in their place. A rank is a relative measure, and reading it as a statement about Dubai alone overstates what happened.
The caveat that belongs with the number. UBS states in the 2026 report that the data are not comparable with previous years because of a comprehensive data revision, and that scores may differ from previous editions through revisions, updates and the dynamic calculation of index weights. Part of the move from 0.64 to 1.16 is therefore methodology rather than market. UBS does not quantify how much. Anyone presenting the two-year jump as a pure market signal is reading past the report's own footnote.
What the report says about owning versus renting. One figure cuts against the headline. UBS estimates the user cost of owner-occupied housing, which includes mortgage and maintenance payments adjusted for risk premiums and expected price growth, and compares it to the rent on a comparable property. In Dubai that estimated user cost sits below rent, alongside Madrid and Zurich, while in Hong Kong, Vancouver and Los Angeles it sits well above. Dubai also carries one of the lowest price to income ratios in the study at about 5 years of income for a 60 sqm apartment, against 15 years in Hong Kong and 11 in London, and a price to rent ratio of about 16 years. A market can be expensive relative to its own history, which is what the index measures, and still be cheap relative to renting in it. Both are true here.
How to read it. The constructive case: no credit excess, ownership cheaper than renting, price to income among the lowest of 23 cities, and a price level that paused rather than broke. UBS expects an improvement in the geopolitical environment to support a rapid recovery in sentiment. The cautious case: the price level is now carried by expectations rather than by rents or incomes, rents are falling, construction has moved up a band, and UBS names persistent concerns about structural oversupply. A market priced on expected appreciation is the market that reprices fastest when the expectation changes.
Neither case is settled by a score of 1.16. The score says the distance between price and fundamentals has widened. What closes that distance, rising rents or falling prices, the index does not say.
Sources: UBS Global Real Estate Bubble Index 2026, pages 15, 25 and 26, data collected through 26 August 2026 (Tier 2, institutional primary document); UBS Global Real Estate Bubble Index 2024, pages 19 and 22, data as of the second quarter of 2024 (Tier 2). Sub-index bands, methodology and the comparability caveat are taken from the methodology section of the 2026 edition. No Tier 1 source exists for this index; it is UBS's own construction.