Roboter - Investment: Wer sind diejenigen, die nur aus Neugier zuschauen und so das Publikumswachstum antreiben? Die Risikokapitalgeber (VC) sind die, die Geld investieren. Aber wer ist es, der Verluste erleidet?
If one says that the year 2023 marked the beginning of the listings of robotics companies on the Hong Kong Stock Exchange, then the year 2025 is the peak.
On December 29, 2023, an "unprecedented" bell rang on the Hong Kong Stock Exchange – not by a person, but by a robot.
It was the Walker S, manufactured by Ubtech. Standing at 1.7 meters tall, it rang the bell together with Zhou Jian, the founder of Ubtech. This ceremony was the world's first event where a humanoid robot participated in the stock - ringing ceremony. It was full of tech elements and camera flashes and caught the attention of the capital markets.
This show in 2023 marked the beginning of the capital fairytale of Embodied Intelligence. Two years later, in 2025, the most frequent question from the media was not whether one could build robots, but to whom one could sell them and whether they could function sustainably.
As the waves of applause faded, the stock prices fluctuated violently, and institutions rushed to sell their stocks, a problem suddenly became clear: We always say that the ultimate form of a robot is to be like a human. But when such a robot actually appears, what can it really do?
2025: The Hong Kong Stock Exchange experiences the "Robot Season"
In the first half of 2025, the Hong Kong Stock Exchange witnessed an unprecedented "robot wave". By July of this year, a total of three robot - related companies went public (Geek+, Borayton, Sanhua Intelligent Control), and eleven robot - related companies submitted applications (Stanley Robotics, SEER, Estun Automation, Woan Robotics, Yunji Technology, LeDong Robotics, Yushi Technology, Kylas, Yifei Technology, Zhaowei Machinery & Electronics, Roborock).
Among them, Geek+, the market leader in warehouse robots, successfully went public on July 9, raised over HK$ 2.7 billion in capital, and had a market value of over HK$ 22 billion at its IPO. However, it had a net loss of over 3.5 billion yuan from 2022 to 2024.
Mech - Mind has submitted an IPO application. It is mainly specialized in industrial and medical automation hardware as well as AI platforms. From 2022 to 2024, it had a cumulative loss of 2.281 billion yuan, and the research and development expenses accounted for about 40% of the sales revenue.
Yunji Technology focuses on hotel and business delivery robots. In 2024, it had a sales revenue of about 245 million yuan and a gross margin of 43%. At the same time, it had a cumulative loss of over 800 million yuan from 2022 to 2024 and a convertible bond consideration of 1.8 billion yuan.
Stanley Robotics and Kylas Technology have also successively submitted applications. They position themselves as providers of AGV and AMR platforms. The former had a cumulative net loss of 273 million yuan from 2022 to 2024, and the latter had an annual loss of about 170 million yuan.
The data shows that with the exception of some companies that are already profitable on the Chinese stock market (A - shares) and raise capital again through a "return IPO" on the Hong Kong Stock Exchange (e.g., Quectel Wireless Solutions, Zhaowei Machinery & Electronics), most robotics companies are in a long - term loss state. Under Rule 18C of the Hong Kong Stock Exchange, there are not a few technology companies that go public "with problems". In other words, this wave of IPOs is more of a capital race among non - profitable companies: Those who can tell a good story and raise capital have more chances to survive the tough industry reshuffle.
In 2023, the Hong Kong Stock Exchange introduced the new IPO rule 18C, which allows non - profitable "specialized technology companies" to go public. Together with the "special route for technology companies" launched in 2025, which supports technology companies to submit applications confidentially and accelerates the approval process, this has opened a low - barrier access for robotics companies. Many non - profitable robotics companies in urgent need of capital decide to list on the Hong Kong Stock Exchange.
The enthusiasm of investors is shown not only in the number of applications but also in the performance on the secondary markets. Thanks to the artificial intelligence wave, policy advantages, and the popularity of the "humanoid robot" concept, the stock prices of already - listed robot stocks have been rising steadily. Since 2025, the stock price of Ubtech, the "first stock in the humanoid robot field", has risen by about 60%. The stock price of Dobot (which mainly manufactures robot arms and educational robots), which was listed on the Hong Kong Stock Exchange at the end of 2024, has risen by over 170% this year. The Hong Kong market has a high acceptance of new technology companies such as robotics companies and is willing to give a valuation premium, which provides valuable capital support for the companies' development.
It's not only the Hong Kong Stock Exchange that is hot. The Chinese stock market (A - shares) and the primary markets are also active. Tash Robotics raised the largest angel round of $120 million in March 2025. Unitree Robotics announced on July 18 that it was preparing for an IPO on the Chinese stock market (A - shares). On July 21, JD.com made a series of investments in three robot start - ups: Qianxun Intelligence, Zhujidongli, and Zhongqing Robotics. The investments cover the fields of warehousing, logistics, and humanoid robots, which almost cover all the current narrative hotspots of the robotics industry. Sichuan Tianlian Robotics announced on July 30 that it was preparing for an IPO on the Chinese stock market (A - shares).
We can see that although many robotics companies have heavy losses and tight cash flows, this does not prevent them from becoming the new favorites on the Hong Kong Stock Exchange, with surprisingly high valuations. The investors' trust in this industry is based on the "future view" rather than the "balance - sheet view".
Capital investments and industry portrait – Who is investing? Who is making the most money?
These companies that have submitted applications have attracted a series of established institutions and new industry investors.
Many large investors participated in the IPO of Geek+, such as Volcanics Venture, Banyan Capital, Warburg Pincus, CPE Yuanfeng, Granite Asia, Ant Group. They are mainly state - owned sovereign funds and other sovereign funds.
Behind Mech - Mind are shareholders such as Innovation Works, AIC, Joy Capital, etc., including both state - owned capital and market - oriented institutions.
Yunji Technology has received investments from companies such as Tencent, Alibaba, Ctrip, iFlytek, Lenovo, both state - owned and Internet companies in the past.
Internet giants are entering the industry. Tencent is betting on the ecosystem, has released an Embodied Intelligence Open Platform, and has invested in Zhipu Robotics. Meituan Longzhu Capital has invested in nearly 30 robotics companies and is the second - largest shareholder of Unitree Robotics. JD.com has made a series of investments in three robotics companies such as Zhujidongli and integrated them into its Supply Chain Technology strategy. Alibaba has been more quietly involved in algorithms and perception systems. On the one hand, the giants act as "water suppliers" (chips, AI platforms), and on the other hand, they are after scenarios and hardware, which restricts the living space of small and medium - sized enterprises.
One can recognize that capital is constantly "marking the territory". Its core concept is: Robots + Artificial Intelligence+ industrial manufacturing = capital landscape. In particular, the investors from large platform companies are not only looking at the future valuation but also strengthening their own ecosystems.
We can clearly see the differences between two main robotics industries:
The first type of robotics companies has a clear profit path and stable market demand.
Floor - cleaning robots (e.g., Roborock, Ecovacs, Xiaomi, Dreame): In 2024, Roborock had a net profit of about 1.98 billion yuan and has already established a profitable operation and closed an ecosystem. Dreame had a sales revenue of about 20 billion yuan in 2024 and a net profit of about 3 billion yuan, which is above the industry average. Its premiumization strategy was successful – its market share of floor - cleaning robots over 5000 yuan in China is about 30%, and the share of premium products overseas is over 50%.
Collaborative robots (Cobot): Dobot had a sales revenue of 374 million yuan. JAKA and AUBO are also actively focusing on robot arms. The technology is mature, and the ROI is quantifiable. Business customers make repeat purchases.
The second type is the "humanoid + platform" sector, which still depends on the capital flow.
Geek+, Mech - Mind, Yunji Technology, etc. have a revenue increase, but their profit models depend on continuous capital raising to maintain research and development and market expansion. The construction of their own cloud services, system integration, and robot - solution ecosystems still requires high investments.
Humanoid robot companies (Ubtech, Zhujidongli, Zhongqing Robotics) are still in the research or trial - production phase. The commercial path is still unclear. Their "star products" trigger imitations worldwide, but the sales volume is extremely low, and the market reaction is weak.
In short: The more one talks about universal ability, the fewer scenarios and orders there are. The clearer the requirements are, the easier it is to amortize the cost of a single robot.
The more humanoid, the greater the dilemma – Is the "humanoid development" a step backward?
After Unitree's appearance on the Chinese New Year's Eve Gala, the robot was very impressive, but it is the best product in the humanoid robot category, and only a few tens of thousands of units are sold annually.
New Chinese "humanoid - like" companies such as Zhujidongli and Zhongqing Robotics have a flexible structure and a human - like appearance, but they are not yet in mass production. Investors are "burning money" in the hope of the future.
The Walker S of Ubtech was once the symbol of Chinese technology in the humanoid robot field.
It can walk, open doors, carry things, dance, go downstairs, and was even seen on the stage of the Chinese New Year's Eve Gala several times. In 2023, it "rang the bell itself" for a company listed on the Hong Kong Stock Exchange.
It can do almost everything except actually being sold.
Overall, from 2021 to 2023, only 10 units of the Walker series of Ubtech were sold, and the cumulative sales revenue was 59.8 million yuan. In 2024, only 10 units of humanoid robots from Ubtech were delivered, and the sales revenue was about 35 million yuan.
You may ask: What are the tens of thousands of robots sold by Ubtech? They are mainly small products for education, presentation, or customized scenarios. They have a low price and a low margin. Although there is a certain quantity, it cannot cover the high research and development costs at all.
The problem of the Walker is almost the problem that all humanoid robot companies have: We can make robots more and more like humans, but we can't find useful applications for their "being like humans". In places where there is actually a need – such as care, patrol, security, logistics – people are not interested in "being like humans" at all. Floor - cleaning robots don't need legs, robot arms don't need emotions, and warehousing and transportation don't care whether one can dance. It is also cheaper to buy so many audio devices in a museum that one can carry around than to buy a humanoid robot as a guide.
The more humanoid, the more expensive; the more expensive, the less useful.
So these robots become "show artists" in media exhibitions and celebrities on financing PPTs, but they never really enter production lines, hospitals, or communities.
Ubtech had a market value of up to HK$ 140 billion at its IPO. Four months after the IPO, Tencent, the largest shareholder, sold stocks worth HK$ 1.034 billion and made a profit. Minyin Capital even liquidated all its stocks. To maintain its operation, Ubtech had to repeatedly issue new stocks to raise capital again within a year after its IPO.
Ubtech has not been profitable in recent years. In 2024, it had a net loss of 1.124 billion yuan. Although the sales revenue has shown a certain increase in recent years (from 740 million yuan in 2020 to 1.056 billion yuan in 2023), it was hindered from reducing the net loss due to high research and market - expansion costs. On the contrary, the loss increased from 707 million yuan in 2020 to 918 million yuan in 2021, 987 million yuan in 2022, and 1.266 billion yuan in 2023, rising year by year.
You can say that the Walker is a great technical product, but you also have to admit that currently it looks more like an over - valued industrial art piece.
The humanoid development is a way to attract attention, but not a logical approach to selling products. In the real world, no one cares whether a robot can shake hands. People care more about whether it can bring tea to the hotel room, whether it can help the elderly get up, whether it can perform medical operations precisely.