Summary of Key Points
Recently, several small and medium-sized real estate companies (including those listed on the Hong Kong Stock Exchange and established real estate firms) have ventured into the AI industry due to the downward pressure on their core real estate business (reduced profits and increased losses), with a particular focus on the computing power sector. They have quickly entered this field by establishing joint ventures, investing in AI companies, changing their names for transformation, and securing state-owned capital investment. This has led to significant short-term increases in their stock prices. However, the transition faces numerous barriers such as power supply, licensing requirements, and operational capabilities, and whether they will be successful remains to be seen over time.
Why Are Real Estate Companies Flocking to AI? —— Their Core Business Is Struggling, So They Need to Find New Revenue Streams
The real estate industry is currently facing tough times. Moving away from the past model of high leverage and fast turnover, more than 60% of A-share real estate companies are losing money as of 2025, with residential development profit margins ranging only between 3% and 8%. For small and medium-sized real estate firms, they either have to wait for doom or find new ways to generate revenue (a second growth curve). AI is currently a hot topic in the capital market, which can not only offset the decline in their core business but also present a compelling story to investors, boosting stock prices. For example, Yuegangwan Holdings has stated that since 2021, real estate sales have declined, making it difficult to obtain funds. Apart from debt restructuring, diversification and transformation are necessary, and the AI computing power sector, with its established customers and orders, represents a good opportunity.
What Methods Are Real Estate Companies Using for Their AI Transformation? —— From Building Houses to Managing Computing Power
Different real estate companies are adopting various approaches to transform, but they all target the most capital-intensive aspect of AI: computing power:
- Zhongao Dajia: Started with a small trial (partnering with Dongyu Technology in January to use AI for property management optimization) and then made a major leap into new areas (established a new company in July to engage in car computing power enhancement, big data for ride-hailing services, and AI screening for cross-border e-commerce, completely unrelated to their core real estate business).
- Dexiang Real Estate: Completely changed direction—invested $2 million in an AI infrastructure company in March, partnered with Beijing Qiyuan Zhisuan in May to obtain IDC/ISP licenses, hired industry experts as senior managers, and quickly signed its first computing power service contract, also collaborating with Unicom Hong Kong.
- Yuegangwan Zhisuan: Changed its name from a real estate firm focused on urban renewal in the Greater Bay Area. In October 2025, it acquired AI computing power company Tiandun Data, received an 800 million RMB investment from Futian state-owned capital in January this year, and renamed itself Yuegangwan Zhisuan in May. In June, it raised 277 million HKD through a private placement, with 85% of the funds going towards computing power cloud services.
- Jiazhaoye Capital: Entered the AI sector through partnerships, collaborating with domestic AI computing power providers to expand the application of AI chips and products in industrial scenarios.
What Are the Short-Term Effects of These Transformations? —— Stock Prices Soared
Investors are keen on the “AI + real estate” story, and companies that transform quickly have seen substantial stock price increases:
- Dexiang Real Estate: Opening price on May 15 was HK$0.96, rising to a high of HK$2.92, with a market value exceeding HK$3 billion.
- Yuegangwan Zhisuan: Stock price rose from the HK$5–6 range to HK$19.85, an increase of over 300% since the beginning of the year.
This indicates that temporarily, leveraging the AI trend can indeed revive stagnant stock prices.
What Are the Obstacles on the Transformation Path? —— Having Land Is Not Enough
Although real estate companies have land and factories, they still face many challenges:
1. Insufficient Power Supply: Computing power centers require a large amount of electricity (for example, a medium-sized center consumes as much as several tens of thousands of households). Many real estate firms do not have stable green energy sources.
2. Difficulties in Obtaining Licenses: Establishing IDCs (data centers) and ISPs (internet service providers) requires special licenses that are not easily obtained.
3. Lack of Operational Knowledge: Real estate companies are good at building, but operating AI computing power centers requires technical teams and customer networks, which are areas of weakness for them.
4. Incompatible Facilities: Traditional factory and commercial properties do not meet the requirements of computing power centers, and renovations are costly.
Expert Wu Rui pointed out that real estate companies' transition to AI is essentially arbitrage between a “low point in the real estate market” and a “peak in the AI computing power sector.” In the short term, it’s about capital storytelling; in the long run, success depends on whether they can shift from a “building-focused mindset” to an “AI operations mindset”—these two industries have fundamentally different approaches.
What Do Experts Think of This Transformation? —— Opportunities Exist, but Challenges Are Greater
- Liu Shui from China Index Academy: Real estate companies can use their own parks and green energy to establish smart computing centers, renting out cabinets and providing maintenance services as a viable new business.
- Kuang Yuqing from Lens Research: AI computing power is a concept favored by investors, and real estate companies are adept at managing heavy assets, which aligns with the high investment required for AI infrastructure.
- Wu Rui from First Pacific Davies: The bottlenecks in the computing power sector are power supply, licenses, operations, and customers, all of which are areas where real estate companies struggle.
In summary, real estate companies' entry into the AI industry is a result of necessity. While it can boost stock prices in the short term, whether they can truly make money in the long run depends on their ability to overcome technical and operational challenges. After all, AI is not a panacea, and simply riding on a trend will not guarantee success.