On the morning of July 27, 2026, at 9:30, ChangXin Technology opened for trading on the STAR Market.
The issue price was 8.66 yuan, the opening price was 49.50 yuan, a gain of 471.59%. Based on the 66.881 billion shares post-issuance, the total market capitalization reached 3.31 trillion yuan, surpassing Industrial and Commercial Bank of China to become the company with the highest market cap in the A-share market. At midday, the stock price reached 54.65 yuan, pushing the total market cap to nearly 3.66 trillion yuan at one point; the closing turnover was 141.187 billion yuan, setting a new record for a single stock's single-day turnover on the A-share market.
Reportedly, 9.4288 million retail investors participated in the online subscription, with a winning rate of 0.4714%. Selling one lot of 500 shares at the opening could net a profit of 20,000 yuan. ChangXin raised 57.919 billion yuan based on the initial offering size, making it the largest IPO on the STAR Market since its launch.
After the bell rang, people quickly calculated the fortunes of the old shareholders. What was the value of Hefei State-owned Assets' holding? How much was Alibaba's stake worth? And the employee stock ownership platform?
But one name was missing from the list: Country Garden.
In 2021, it invested 2 billion yuan to acquire approximately 2.24% of ChangXin's equity. After several rounds of capital increases and dilution, its stake was reduced to 1.56%. In December 2024, Country Garden sold this entire stake to Hefei State-owned Assets, recovering 2 billion yuan. On the day of ChangXin's debut, the shares Country Garden once held were worth roughly 50 billion yuan.
Five years ago, it correctly judged an extremely difficult matter. Nineteen months ago, it was forced to sell that judgment.
Country Garden assessed the industry correctly, but miscalculated how long it could wait.
Why Did a Property Seller Venture into Hard Tech Research?
Country Garden Capital (CGC) was established in 2019.
A year before its establishment, Country Garden proposed a transformation into a "high-tech comprehensive enterprise." At that time, it was still one of China's private enterprises with the most abundant cash flow. The property main business could provide large amounts of proprietary capital, and the group's brand helped this new investment arm access projects that smaller institutions couldn't reach. Back then, it certainly didn't lack money, but it lacked a path beyond real estate.
CGC's first high-profile hard tech investment was in rockets.
In 2019, the team spent nearly six months interviewing domestic commercial aerospace companies to first understand the industry's real bottlenecks. Remote sensing, communications, and navigation all had satellite launch needs, and the bottleneck lay in launch capacity. Ultimately, they selected LandSpace, which was then focusing on medium-to-large liquid-fuel rockets, and exclusively invested 500 million yuan in its Series C round in December. They later led two more consecutive rounds, boosting LandSpace's valuation from around 3 billion yuan to over 10 billion yuan.
This episode is quite illustrative of how Country Garden learned to invest.
It brought its house-building approach into venture capital. Real estate development is inherently a business highly dependent on supply chains; Country Garden was most familiar with how to break down a long chain and find its most critical nodes.
In investment, this approach was given two names. Core companies on the industrial chain were called "chain anchors," and bottlenecks were called "chain constraints." Before entering a new industry, the team had to conduct complete industry research; in principle, projects couldn't enter investment decision-making without it. Its first semiconductor investment was Unisoc, a "chain anchor" enterprise; moving towards the manufacturing end, ChangXin was the kind of unavoidable "chain constraint" enterprise.
Country Garden's capital allocation was also distinctive. Most ordinary RMB funds have a lifespan; they must exit after seven or ten years. CGC primarily used the group's proprietary capital, which managing partner Niu Ruolei then called "perpetual capital," theoretically without time constraints. It adopted a barbell strategy, investing in very early-stage technology projects on one end and super unicorns close to IPO with single investments exceeding 1 billion yuan on the other, rarely participating in the most crowded growth stage in the middle.
By early 2022, CGC had independently completed over 90 investments, backing 26 unicorns and 10 listed companies; 52% of the investment amount was placed in hard tech fields like advanced manufacturing, semiconductors, and carbon neutrality. The list included ChangXin, LandSpace, Unisoc, Biren Technology, BYD Semiconductor, SJ Semiconductor, SVOLT, Dreame Technology.
Why Invest 2 Billion in ChangXin?
The DRAM produced by ChangXin is the memory used when computers, phones, and servers are running. The product may seem small, but behind it lies a massive factory that needs to be sustained.
A wafer fab must continuously buy equipment, modify processes, support engineers, and endure the severe cyclical fluctuations of memory prices. By the end of 2024, ChangXin had accumulated losses of approximately 38.52 billion yuan on its books. Before the IPO, it already had three 12-inch wafer fabs, nearly 20,000 employees, including over 6,000 R&D personnel.
In 2025, Samsung, SK Hynix, and Micron still collectively held over 90% of the global DRAM market. In semiconductors, many tracks can start with a smart, small team, but DRAM cannot. Design capability, manufacturing processes, yield, equipment, and capital—fail at any one, and you cannot achieve scale.
ChangXin was founded in 2016. On September 20, 2019, it announced the production of its self-developed manufacturing project, with the first appearance of its 8Gb DDR4, proving that mainland China finally had an enterprise crossing the threshold of "whether it can be made" for DRAM mass manufacturing.
Therefore, when Country Garden entered in 2021, ChangXin's earliest technical risks had decreased, while the most expensive phases of capacity expansion, iteration, and market competition were still ahead. This position precisely fell on the large-investment end of Country Garden's "barbell."
On July 5, 2021, Country Garden Capital (Haikou) signed a Series B capital increase agreement with ChangXin and other shareholders. The unified price for the Series B was 2.219 yuan per registered capital unit, with an initial fundraising cap of 26 billion yuan. Country Garden invested 2 billion yuan, subscribing to approximately 901.3 million registered capital units, holding 2.24% post-investment.
The same round also included National Integrated Circuit Industry Investment Fund Phase II, Anhui Provincial Investment Group, China Structure Reform Fund, China Merchants Bank, Xiaomi, Midea, and other institutions and companies; in December of that year, the fundraising cap was raised again to 36 billion yuan.
From Unisoc's chip design to ChangXin's wafer manufacturing to SJ Semiconductor's packaging, Country Garden Capital pieced together the semiconductor chain segment by segment. Investing in ChangXin was because they believed manufacturing was the unavoidable "chain constraint" for China's semiconductors.
China was a major global demand market for DRAM, and at that time, mainland规模化 supply was almost starting from zero. As long as ChangXin survived, expansion itself would allow it to grow rapidly.
This judgment later proved quite accurate. By the fourth quarter of 2025, ChangXin's global market share based on sales rose to 7.67%, becoming China's No. 1 and the world's No. 4; in the first quarter of 2026, revenue was 50.8 billion yuan, with net profit attributable to the parent company reaching 24.762 billion yuan. It took many years to turn "can manufacture" into "can sell," and it coincided with AI servers pushing memory demand to a new height.
"Perpetual Capital" Suddenly Had a Deadline
2021 was the year of ChangXin's Series B funding, and also the peak year for China's real estate sales.
That year, 1.794 billion square meters of commercial housing were sold nationwide, with sales value reaching 18.19 trillion yuan. By 2025, the sales area of newly built commercial housing was only 881 million square meters, almost halving in four years. The national population was 1.4126 billion at the end of 2021, and negative growth began the following year.
In the ebbing tide, Country Garden was among the first to feel the impact. In 2021, 68% of its sales came from third- and fourth-tier cities. These cities once gave Country Garden its broadest market but also made it harder to pivot when demand contracted.
At its peak, Country Garden's equity contract sales were 558 billion yuan, with cash collection of 502.2 billion yuan, and available cash of 181.3 billion yuan. Three years later, equity contract sales dropped to 47.2 billion yuan, with a net loss attributable to the parent of 32.8 billion yuan, and total borrowings of 253.5 billion yuan. At year-end, total cash was 29.9 billion yuan, of which 23.5 billion was restricted, leaving only 6.362 billion yuan listed as cash and cash equivalents.
Two billion yuan on Country Garden's books in 2021 was an industrial investment that could wait ten years. By 2024, it was equivalent to nearly one-third of the year-end free cash.
Pre-sale payments received by sales offices do not truly belong to the developer. Behind each pre-sale payment is a house yet to be delivered; the money must eventually become steel, concrete, elevators, and keys.
From 2022 to November 2025, Country Garden delivered approximately 1.8 million housing units cumulatively. To deliver these units, starting from 2022, it sold equity stakes, hotels, major assets, and even corporate vehicles, cumulatively recovering over 65 billion yuan. The 2 billion yuan from ChangXin was part of this.
On May 31, 2024, Bloomberg reported that Country Garden Capital was seeking a buyer for its ChangXin stake, asking for about 2 billion yuan. The transaction was still under review and might not be completed. Country Garden later responded that the group was assessing its asset portfolio and potential disposal opportunities to optimize its asset-liability structure.
In June, ChangXin completed a new round of capital increase, with 12 investors subscribing to 10.8 billion yuan at 2.61 yuan per share. Based on this price, Country Garden's holding was already worth over 2.3 billion yuan.
On December 27, the seller Huibi Fund V, the buyer Hefei Jianchang, and ChangXin Technology signed a share transfer agreement, with the final price still at 2 billion yuan, approximately 2.22 yuan per share. Hefei Jianchang is directly held 87.45% by Hefei Construction Investment Holding Group, with the ultimate beneficial owner being the Hefei State-owned Assets Supervision and Administration Commission. The entity taking over Country Garden's stake was the state-owned capital of the city where ChangXin is based.
This was no longer a composed investment institution choosing the optimal exit window. Both buyer and seller knew the seller needed cash, and the last semblance of dignity Country Garden could maintain was to get its principal back intact.
The payment arrangement in the agreement also reflected this urgency. ChangXin was to provide a sealed new shareholder register on the tenth business day; Hefei Jianchang would pay in full upon receiving the register and payment notice. Hefei Construction Investment also provided joint and several guarantees for 99.985% of the consideration, up to 1.9997 billion yuan. If not completed within 180 days, the party not in breach could terminate the agreement. Other old shareholders explicitly or implicitly waived their pre-emptive rights.
Country Garden's announcement specified the use of this 2 billion yuan as general working capital, mainly for project construction such as completing pre-sold projects. ChangXin's equity was turned into construction payments for the work sites.
Judging from the changes in shareholding during the reporting period disclosed in ChangXin's prospectus, Country Garden was also the only early-stage external investor that completely sold off its holding, not waiting for the IPO. Some made small transfers, some moved shares to related platforms; only it truly cleared its position.
The Tale of Twenty-Four Cities
At the end of 1958, a large number of cadres, workers, and equipment from Shenyang's Factory 111 migrated southwest. Some sold their houses and furniture, taking their families by train and ship across thousands of kilometers to Chengdu. In January 1959, Factory 420 was formally established. It later became an important aviation engine factory in southwestern China, with workshops, dormitories, schools, and canteens all connected, the lives of generations revolving around the machinery.
Half a century later, Chengdu's eastern suburbs adjusted their industrial layout. The old site of Factory 420 was handed over for real estate development, the original factory area becoming the residential complex "Twenty-Four Cities."
Upon hearing that a state-owned factory with tens of thousands of workers was to be turned into residential housing within a year, Jia Zhangke felt there was "too much that could be said" about this event and thus filmed "24 City." In the movie, people sit in front of the old factory buildings, talking about production lines, collective dormitories, and a vanished way of life.
Country Garden's bell-ringing in Hong Kong, China, was on April 20, 2007. The issue price was HK$5.38, and it closed at HK$7.27 on its first trading day. The public offering was oversubscribed 255.7 times, freezing approximately HK$330 billion. Yang Huiyan, not yet thirty, held 58.19% of the shares; real estate created China's new richest person that morning.
Factory 420 becoming Twenty-Four Cities and Country Garden becoming a rising star in the capital market happened in the same era. There was a strong sense of direction in cities then; old factories represented the past, commodity housing represented the future; the land under the production lines was vacated, built into residential properties, and only then was its value rediscovered.
Country Garden grew precisely within this direction. It connected land, population mobility, and pre-sale funds into a massive machine, moving from Shunde to over two hundred cities. By 2019, it was attempting to channel the private capital accumulated from real estate towards rockets, chips, and new energy.
The money earned from the old industry should naturally fund the new industries not yet mature. It's just that the old cycle receded too quickly, and the new cycle arrived too slowly.
On July 27, 2026, the bell rang again. This time, standing on the podium was a wafer manufacturing enterprise.
