The stabilization of total housing demand and the halt in rental price declines have further solidified the long-term support for the property market. The core issue confronting the real estate sector is no longer whether the aggregate market can bottom out, but rather whether the valuation premiums on certain properties can be sustained. In cities where total demand has stabilized, prices for essential housing are secure, yet the trajectory of upgrade property prices remains contingent on future expectations and confidence.

In our late-January report, we highlighted that based on long-term indicators such as cumulative price declines, rental yields, and price-to-income ratios, the property markets in most cities are approaching valuation floors. This, combined with the stabilization of total housing demand in core cities during 2025, suggests the market has preliminarily established the conditions for a halt in decline. These long-term supports explain why the market has maintained resilience amidst multiple headwinds. Since the start of the year, these stabilizing factors have gained further confirmation.

Total Housing Demand Stabilizes as New Supply Contracts

In the first half of this year, national commercial housing sales area fell 12.4% year-on-year, yet this contraction stems not from weak demand but from a rapid shift in transaction composition. The share of existing home sales nationally has surged from 46% in 2025 to 50.4%, a 5-percentage-point increase in just six months. Given this compositional shift, new home sales figures have become less representative of true housing demand. The combined sales area of new and existing homes serves as a superior gauge of total demand. In H1, the combined total declined merely 1.9% year-on-year, against a relatively high base boosted by the September 24 policy. We project total housing demand in 2026 to hold steady at 13.0-13.3 billion square meters, roughly flat with 2025, maintaining a three-year range of 13-14 billion. This indicates current market pressures are not demand-driven; structural oversupply and weak confidence are the true challenges.

Notably, the accelerated rise in existing home transaction share this year differs markedly from prior years. Between 2022 and 2025, the increase was demand-led. On one hand, after price caps on new homes were lifted, their low-price advantage eroded. Existing homes, with more flexible pricing, underwent deeper price correction during the downturn, attracting buyers through price-for-volume strategies. On the other hand, new homes dominated pre-2021 because urban expansion followed an outward-growth model, where non-core new developments offered strong investment value. Now that urban development has shifted to an intensive-growth approach, with slower industrial sprawl, occupancy value has become the dominant pricing factor. In this context, existing homes have gained traction due to superior surrounding amenities and the certainty of immediate occupancy, steadily increasing their transaction share.

This year's rapid growth in existing home share is not driven by price-for-volume dynamics but by supply-side factors. Unlike the accelerated price declines seen in the same period last year, existing home prices have shown limited overall decline, yet the transaction share jumped 5 percentage points in six months. The underlying cause is developers' proactive strategic retrenchment, focusing land acquisition on core tier-1 and tier-2 cities. The number of cities where typical developers acquire land is shrinking; for instance, China Merchants Shekou acquired land in only 4 core cities in H1, down from 20 last year. Concentration has intensified, with the top 20 cities accounting for 62% of national residential land transfer fees in H1, up 10 percentage points from 2025. Shanghai, Hangzhou, Beijing, Guangzhou, and Shenzhen lead nationally. While overall land market turnover has fallen significantly, bidding for quality parcels in core cities remains intense, with high premiums and multiple rounds of bidding. Conversely, non-core city land markets stay depressed; tier-2 residential land transaction volume plunged 37.6% year-on-year in H1, far exceeding the 23.7% decline across 300 cities. The market consensus reflected in these land trends is that real estate has entered an era of stock, and as the total population declines, cities lacking population appeal will struggle to revive their property markets, while quality land in core cities retains higher safety margins.

The supply-side contraction impacts two fronts. First, the absence of new quality supply in some cities pushes more buyers toward existing homes, accelerating the compositional shift. Notably, tier-1 cities' existing home transaction share reached a stable range of 60%-65% in 2025 and has since remained steady without further upside. The national increase this year is primarily driven by tier-2 cities. Among our 18 sample cities, tier-2 existing home transaction share surged in 2026, even overtaking tier-1 cities in Q2. Second, short-term new home inventory is clearing, and mid-term inventory has fallen to low levels. Since March, the national unsold commercial housing area has ended its continuous rise since July 2021, standing at 760 million square meters by end-July, down 0.8% year-on-year. Of this, 560 million square meters unsold for under three years declined 3.6%, while the remaining 200 million square meters of over-three-year inventory is essentially ineffective. Additionally, the mid-term inventory metric—unsold started area of new residential units (with a sellable ratio of 98%)—recorded 1.22 billion square meters at end-July, returning to early-2008 levels, with monthly declines of roughly 30 million square meters. If new supply continues at current pace, some lower-tier cities will see new supply dry up first.

Overall, the stabilization of national total housing demand reinforces the long-term foundation for a market bottom. Shanghai's experience of price stabilization confirms the logic that volume precedes price. When demand stops contracting and supply gradually clears, prices will find their floor. Shanghai's total housing demand stabilized in 2025, with combined new and existing home sales reaching 34.748 million square meters, marginally above 2024's 34.747 million. Inventory clearing for existing homes began in May 2025, aided by both transactions and landlords actively delisting listings. With improved supply-demand dynamics and confidence boosted by the "Shanghai Seven Measures," existing home listing prices stabilized in late March 2026, followed by five months of "overall stability with structural gains."

By city tier, tier-1 cities have seen total demand recover first this year. From January to July, combined new and existing home sales area in tier-1 cities grew 3.1% year-on-year, reversing a 6.2% decline in 2025. Shanghai, Shenzhen, and Beijing posted positive growth of 9.0%, 2.6%, and 0.002%, respectively, while Guangzhou's total demand continues to contract at -3.4%. In contrast, tier-2 cities face greater contraction pressure; among our 18 sample cities, their combined sales area fell 8.0% year-on-year in the January-July period, the steepest decline. Tier-3 and tier-4 cities, having adjusted earlier, see total demand near historical lows with smaller declines. Within tier-2 cities, divergence is pronounced. Cities contracting this year are primarily domestic-demand-oriented, lacking new growth drivers amid weak traditional growth; Chengdu, Changsha, and Shenyang have experienced catch-down price adjustments. Conversely, cities benefiting from emerging industries, such as Suzhou and Hefei, have seen firm prices, with transaction prices up 2.2% and 2.1% year-to-date, respectively. For a market at cyclical bottom with national demand stabilizing, the strength of economic growth momentum will become the decisive factor in cross-city price performance.

Rents Halt Decline as Rental Yield Rises to 2.5%

Comparing real estate to equities, rental yield is akin to a house's P/E ratio, while rental prices represent its earnings per share (EPS). If rental prices—the EPS—continue falling, property prices may overshoot despite valuations being at bottom. Conversely, if rents stabilize, prices could rebound even if the yield hasn't bottomed. From a yield perspective, housing prices are broadly at valuation floors. When U.S., Japanese, and Hong Kong property markets bottomed, net rental yields traded within roughly 20 basis points of mortgage rates. In July, the 100-city rental yield reached 2.48%, approaching the 2.6% provident fund loan rate and returning to the reasonable 2.4%-2.8% range. Furthermore, optimized provident fund policies allow more buyers to anchor mortgage rates at 2.6%.

First, multiple cities have raised provident fund loan caps this year; for example, Beijing and Shanghai now cap family loans at 2.4 million yuan, sufficient to cover loans for median-priced homes. Second, the recent State Council decision to amend the Provident Fund Management Regulations allows flexible-employment workers to voluntarily contribute, broadening coverage. More critically, national rents have turned upward this year. After a Q1 stabilization with fluctuations, Q2 trends clarified, with multiple metrics showing positive month-on-month growth. The Iceberg 80 Rent Index posted non-negative monthly changes in May, June, and July, whereas it had been negative for four consecutive years from 2022-2025, ruling out seasonal factors. Additionally, the China Index Academy 50-City Rent Index rose 0.13% month-on-month in July, widening from June, while the National Monthly Rent Index gained 0.9% in July, up 4.1% from its yearly low.

The genuine significance of rent stabilization is that essential housing prices now enjoy fundamental support. From a valuation standpoint, stabilized rents combined with yields back in reasonable territory imply that occupancy value and prices for most urban properties are broadly aligned, limiting downside. However, by property type, upgrade and high-end homes retain some financial attributes, whose value hinges on risk appetite and future expectations, making rent stabilization less impactful for them. Valuation differences across property types are mirrored in tiered rental yields. In Shanghai, essential micro-units command higher yields, whereas high-total-price, large-unit properties offer lower yields—a pattern not unique to the mainland. In Hong Kong, Tokyo, and New York, yield distributions follow similar rules. In Hong Kong, homes under 40 square meters yield 3.4%, but those over 160 square meters yield just 2.3%. The low rental yield on mid-to-large units does not equate to valuation bubbles; it reflects financial value derived from prime locations, product quality, supply scarcity, and appreciation expectations. Yet this value lacks current rental income support and depends on future expectations. Consequently, the market will continue to exhibit divergence: essential housing in cities with stabilized demand faces no price worries, while upgrade property prices hinge on future expectations and confidence.

Hong Kong's relatively higher rental yields stem from elevated holding costs, including rates, government rent, maintenance reserves, and management fees. Excluding these costs, comparable unit yields in Hong Kong exceed Shanghai's by roughly 30-50 basis points, while Hong Kong's mortgage rate (P-loan at 3.25%) is about 40 basis points above Shanghai's (weighted commercial and provident fund rate around 2.85%). Thus, after accounting for holding costs and borrowing rates, Shanghai's rental yields are broadly on par with Hong Kong's. Hong Kong's housing prices stabilized and recovered in July 2025, and Shanghai has largely achieved listing price stabilization and transaction price recovery since March 2026. The market's central question is no longer whether the aggregate can bottom, but whether valuation premiums on select properties can endure. Since the start of the year, cities where existing home transaction prices have bottomed and rebounded fall into two categories: those with robust new growth drivers supporting structural premiums, such as Shanghai, Beijing, Suzhou, and Hefei; and those with fully deflated price bubbles where premiums have largely normalized, including weaker tier-2 cities like Dalian, Tianjin, Nanchang, and Guiyang, along with tier-3 cities such as Langfang and Xuzhou. Meanwhile, ongoing adjustments in "new first-tier" cities like Hangzhou and Chengdu fundamentally represent catch-down moves after premium removal.

Risk warnings: Macroeconomic improvement may fall short of expectations; property market stabilization could lag forecasts; new home supply bottlenecks may intensify.