A reading of the UBS Global Real Estate Bubble Index 2026: where bubble risk is high, where the bubble has already burst, and which cities may be next to start crashing.
Zurich and Tokyo are the cities UBS still classifies as being at the highest housing-bubble risk. Miami, Dubai, Seoul, Lisbon, and Geneva sit in elevated-risk territory, the second highest risk category. London, Paris, New York, San Francisco, and São Paulo are next, according to the UBS Global Real Estate Bubble Index 2026, published by UBS Chief Investment Office of Global Wealth Management, with an editorial deadline of 21 September 2026 and data collected through 26 August 2026.
UBS uses fairly crude analysis, looking at rent versus price and income versus price. Many investors know the 1% rule of thumb, that states that if you can buy a home where the rent is above 1% of the price, you will have a decent cash flow. More seasoned investors realize that property taxes are higher in places like Chicago than Spokane, that insurance is more expensive on the coast than inland, and that you really need to calculate cap rate, not just rent versus price. In the editorial, Claudio Saputelli and Matthias Holzhey write:
“As housing affordability has declined for many households, accumulated financial wealth has become an increasingly dominant driver of housing demand. Strong gains in global equity markets have boosted the purchasing power of affluent households, enabling larger down payments and, in some cases, outright cash purchases.”
They add that “the rapid expansion of AI-related investment is generating substantial yet highly concentrated wealth,” and that “prime residential segments are outperforming broader housing markets across many cities,” with the split “already evident in local housing markets, particularly in San Francisco and Seoul.”
That is the frame for the whole report: a market increasingly split between wealth-driven districts and income-driven citywide housing, less tied to the global interest-rate cycle than it used to be.
What the index actually measures
UBS is careful about the word “bubble.” The index “gauges the risk of a property bubble” from recurring patterns — prices decoupling from incomes and rents, plus real-economy imbalances such as excessive lending and construction — but it “does not predict whether or when a correction will occur.” A shift in macro conditions, investor sentiment, or housing supply “could trigger a decline in house prices.”
The score is a weighted average of five standardized sub-indices: price-to-income, price-to-rent, the change in mortgage-to-GDP, the change in construction-to-GDP, and the city-to-country price ratio (replaced by an inflation-adjusted price index in Singapore, Hong Kong, and Dubai). Bands are high (above 1.5), elevated (1.0 to 1.5), moderate (0.5 to 1.0), and low (below 0.5).
Scores can differ from prior editions because of data revisions and because the weights are recalculated. UBS notes that the indexes for New York, Los Angeles, San Francisco, and Milan were adjusted for source revisions, and that the price-to-income and price-to-rent comparisons “are not comparable with previous years due to a comprehensive data revision.”
The 2026 scoreboard
Risk levels are “broadly stable” versus 2025. Inflation-adjusted global house prices, rents, and incomes were “largely unchanged” over the past year. Financing costs rose slightly, which hurt affordability and the incentive to buy, while shortages in many cities continued to support prices.
| Rank | City | Score | Band | Real price change, last year | Real rent change, last year |
|---|---|---|---|---|---|
| 1 | Zurich | 1.69 | High | +4.6% | −0.6% |
| 2 | Tokyo | 1.54 | High | +6.0% | +6.0% |
| 3 | Miami | 1.41 | Elevated | −1.6% | −3.1% |
| 4 | Dubai | 1.16 | Elevated | +0.4% | −4.0% |
| 5 | Seoul | 1.13 | Elevated | +11.0% | +4.9% |
| 6 | Geneva | 1.12 | Elevated | +2.7% | 0.0% |
| 7 | Lisbon | 1.04 | Elevated | +10.2% | +1.4% |
| 8 | Amsterdam | 0.95 | Moderate | −2.0% | +2.0% |
| 9 | Madrid | 0.86 | Moderate | +9.0% | +2.2% |
| 10 | Los Angeles | 0.69 | Moderate | −3.2% | −3.7% |
| 11 | Sydney | 0.68 | Moderate | −4.4% | −0.6% |
| 12 | Frankfurt | 0.64 | Moderate | −3.2% | +0.5% |
| 13 | Toronto | 0.63 | Moderate | −10.7% | −3.8% |
| 14 | Vancouver | 0.62 | Moderate | −9.7% | −5.0% |
| 15 | Munich | 0.61 | Moderate | −3.7% | +0.4% |
| 16 | Hong Kong | 0.61 | Moderate | +8.7% | +3.6% |
| 17 | Singapore | 0.54 | Moderate | +1.0% | −0.1% |
| 18 | Milan | 0.50 | Moderate | −0.4% | −6.3% |
| 19 | Paris | 0.33 | Low | −1.6% | −0.9% |
| 20 | London | 0.32 | Low | −1.1% | −0.5% |
| 21 | New York | 0.28 | Low | −0.9% | +1.2% |
| 22 | San Francisco | −0.02 | Low | +1.4% | +8.3% |
| 23 | São Paulo | −0.24 | Low | −0.7% | +1.1% |
Source: UBS Global Real Estate Bubble Index 2026, overview table. Real changes are annualized, as of the report’s data cut.
The regional shifts matter as much as the levels. U.S. and Canadian risk scores fell versus last year. In Asia, Seoul and Hong Kong posted the largest increases; Singapore was broadly stable; Tokyo and Sydney eased. In the eurozone, imbalances widened in Lisbon, Madrid, and Milan, and Paris ticked up, while Frankfurt, Munich, and Amsterdam receded. In Switzerland, “persistently low financing costs have contributed to growing imbalances in both Zurich and Geneva.”
UBS also draws a five-year line under the elevated- and high-risk cities: inflation-adjusted home prices up nearly 30 percent on average, rents up 15 percent, incomes up only about 8 percent. In moderate- and low-risk cities, home prices fell nearly 10 percent. “Historically, deteriorating affordability and widening gaps between home prices and rents have been leading indicators of housing market corrections and financial crises.”
Prices barely moved. A few cities did not get the memo.
Across the cities in the study, real residential prices rose an average of just 0.5 percent year over year, down from 1.4 percent in mid-2025.
Seoul was the outlier, with real price growth “exceeding 10%,” “supported by the AI-driven economic expansion.” Lisbon and Madrid posted similarly strong gains, though “demand has shifted toward more affordable suburban areas.” Hong Kong rebounded sharply. Tokyo recorded real price growth of 6 percent; higher Japanese interest rates “have yet to materially affect the housing market.” Zurich and Geneva kept appreciating on low mortgage rates.
At the other end, “Vancouver and Toronto were the weakest-performing markets, with real house prices declining by around 10% year over year.” Frankfurt and Munich were still adjusting to higher mortgage rates, with real declines of almost 4 percent. Most U.S. cities also saw negative real price growth. San Francisco was the exception.
UBS’s inflation post-mortem is useful. Cities flagged as high bubble risk in 2021 later recorded the sharpest declines, averaging roughly 3 percent a year. Cities with above-average inflation saw real prices fall about 1.5 percent a year; cities with below-average inflation saw real prices rise about 2.5 percent a year.
Three valuation cuts: income, rent, and user cost
Price-to-income. Buying a 60-square-meter (650-square-foot) apartment near the city center is “beyond the reach of the average skilled service worker in most global cities.” Hong Kong remains the least affordable, “requiring around 15 years of average income.” Tokyo, Paris, London, and Seoul are above 10 years. The stretched list also includes Singapore, Lisbon, Zurich, Geneva, São Paulo, Munich, Sydney, Milan, and New York. Compared with 2021, before the rate surge, that skilled worker “can now afford about one-third less living space.”
Price-to-rent. These ratios fell in 2022 and 2023 as financing costs rose, then stabilized as prices and rents moved together from 2024. Zurich now has the highest ratio at 46, Geneva 40. Munich, Frankfurt, and Hong Kong are above 30 “despite elevated borrowing costs.” UBS reads high multiples as “speculative expectations of strong future home-price gains, as well as expectations of lower future interest rates.” São Paulo, Dubai, and the U.S. cities rank among the lowest, helped by less regulated rental markets, higher rates, and, in Dubai and São Paulo, higher risk premiums.
User cost. Direct ownership costs — mortgage interest, the opportunity cost of equity, maintenance, taxes, and depreciation — exceed the rent of a comparable property “in every city covered by the study.” On that measure alone, “homeownership currently appears less financially attractive than renting.” Once UBS folds in a risk premium and expected capital gains, the picture splits. Estimated user costs are “substantially higher than rents” in Hong Kong, Vancouver, and Los Angeles, which “may weigh on future house-price performance.” They are lower than rents in Madrid, Dubai, and Zurich, which “makes homeownership relatively attractive compared with renting.” The punch line: “Expectations of future house-price appreciation appear to be providing important support to housing demand in some cities. Without such expectations, housing-market weakness would likely be considerably more widespread.”
City notes
Zurich (1.69, high). Real prices are up nearly 140 percent over two decades, against rental growth of 40 percent and income growth of 30 percent. Vacancy is near zero, and listings of owner-occupied homes are about two-thirds below their level 20 years ago. UBS still thinks a significant near-term correction is unlikely, given competitiveness and appeal to international tech and AI talent — but “the market is increasingly dependent on persistently low interest rates,” and Zurich has the study’s highest price-to-rent ratio.
Tokyo (1.54, high). Real prices are about 50 percent higher than seven years ago, twice the rise in real rents and roughly five times the rise in real incomes. The boom “may be approaching its limits.” Local buyers are being priced out toward suburbs and rentals, and “any further increase in financing costs would significantly reduce the investment appeal of homeownership.”
Miami (1.41, elevated, score down). Over 15 years Miami had the strongest inflation-adjusted price growth in the study: prices tripled, incomes rose 25 percent. Real prices fell last year and lagged the national average. Single-family inventory is tight; condo supply is not, and reserve rules, inspections, and insurance are raising ownership costs. A broad correction still looks unlikely near term because of relocation demand from higher-cost, higher-tax states and a cash-heavy luxury segment.
Dubai (1.16, elevated). After real price growth above 10 percent in 2025, “Dubai’s housing boom came to an abrupt halt at the onset of the conflict involving Iran.” Real prices fell back to mid-2025 levels; real rents are below a year earlier. Ownership is still relatively attractive versus high rents, but the premium segment depends on high-income inflows, and “persistent concerns about structural oversupply” remain.
Seoul (1.13, elevated, new to the city set). Prices rose 50 percent from 2017 to 2021, fell 20 percent into the 2023 correction, then rose 11 percent over the past four quarters, with rents up 5 percent. “The city is a key AI hub, making future price appreciation closely tied to AI-led economic growth.”
Lisbon (1.04, elevated, also new). Real prices rose nearly 7 percent a year over the past decade, the fastest among the cities analyzed, and another 10 percent since mid-2025. Policies meant to attract foreign capital and residents fueled the boom; a turn toward more selective immigration has produced negative population growth, rental growth has stalled, and demand is shifting outside the city.
London (0.32, low). Weakness is in its fifth year. Real prices are more than 15 percent below 2021. London topped this index in 2016; the score has mostly fallen since. Rents are near record highs, construction is constrained, and 2025 brought the first population decline since the late 1980s, excluding the pandemic years.
San Francisco (−0.02, low). The metro is tightening on an AI-driven tech employment upswing and concentrated equity wealth. Several segments are posting double-digit gains, and real rents are up 8 percent since mid-2025. Weak population growth and affordability-driven out-migration cap the upside; thin inventory limits correction risk. This is the editorial’s exhibit for prime-versus-broader divergence inside a city whose aggregate score is not flashing bubble.
Toronto and Vancouver (0.63 and 0.62). Toronto prices are nearly 30 percent below the peak; Vancouver’s inflation-adjusted prices are 20 percent below the 2022 peak after a near-10 percent drop since mid-2025. Foreign-buyer rules, stress tests, taxes, and, in Toronto, increased supply did the work that the 2022 rate shock started.
What an investor should take from it
UBS is not calling a crash calendar. It is saying three things at once.
First, average exuberance is “well below the levels observed before interest rates began rising in 2021,” and cities that stayed hot have become more disconnected from local incomes and rents.
Second, overheating is local. Prime districts can diverge from the citywide print, and wealth and cash buyers can hold up trophy segments even when the leveraged buyer is gone. Miami’s luxury market and San Francisco’s AI pockets are the examples in this edition. That doesn’t mean that it is safe to buy in what they consider to be lower risk cities. I would argue that all these cities are in a bubble, priced well above what the 1% rule would allow an investor to consider.
Third, the buy-versus-rent math, on direct costs, favors renting almost everywhere in the study. Ownership only “works” where buyers are underwriting future appreciation (if any), or where user-cost estimates (Madrid, Dubai, Zurich) still sit below rents. There are some cities where it almost never makes sense to buy versus rent, such as in the Bay Area.
The index is a risk gauge, not a timing tool, and the publication is not a personal recommendation. For anyone underwriting a city, the 2026 edition is a warning that investors in most cities should not be buying based on an expectation of appreciation.