Competing for the title of "the First A-share Humanoid Robot Stock", will Zhiyuan "steal the show" from Unitree?
The news in the embodied intelligence industry this year is like the dog days of summer - the heat keeps rising wave after wave.
As of mid-July this year, according to incomplete statistics, several embodied intelligence companies such as DeepCloud AI, Sainade, Xingdong Jiyuan, Xiaoyu Zhizao, Tashi Zhihang, Xinghaitu, Kuawei Intelligence, Langyi Robotics, and Zhipu Robotics have carried out financing.
The most dramatic one is the fierce competition for the title of "the first A-share humanoid robot stock".
Not long ago, some media reported that Unitree Robotics planned to officially submit an IPO application to the A-share market before the end of 2025. However, unexpectedly, it was "intercepted" by Zhipu Robotics (hereinafter referred to as "Zhipu").
Recently, Shangwei New Materials issued an announcement stating that Zhipu will acquire approximately 67% of the company's shares, and the controlling shareholder will be changed to Zhipu Hengyue. Deng Taihua will become the actual controller of Shangwei New Materials.
In this acquisition, although Zhipu emphasized that it is "buying" a listed company through agreement transfer + voluntary tender offer. However, investors obviously don't think it's that simple. As of now, the stock price of Shangwei New Materials has hit the daily limit of 20% for seven consecutive days. A small company worth only 2 - 3 billion yuan originally has a market value of over 10 billion yuan in an instant.
Why do secondary market investors have such high expectations for a listed company tendered by Zhipu?
Some people think this operation is "backdoor listing". Others believe that it is Shangwei New Materials actively embracing the "Nine National Policies" and the "Six Merger Policies", and it is a benchmark case for listed companies to actively strengthen integration and transform and upgrade towards new - quality productivity.
However, no matter which view it is, it points to an expectation.
After the transaction is completed, when calculating the market value of Shangwei New Materials, Zhipu's current valuation of over 15 billion yuan should be considered. The benefits of becoming the "first A-share humanoid robot stock" are also obvious. Zhipu can get a large amount of top - tier social capital first, thus guiding subsequent investors to price companies in the same track.
In the era when the embodied intelligence industry is booming, everyone wants to get on this upward - moving elevator.
Is Zhipu really backdoor listing?
To understand why Zhipu's acquisition has caused such a stir in the capital market, we have to look at the specific operation steps.
Zhipu used a series of very complex methods to "buy" a listed company.
Generally speaking, Zhipu's shareholding platform will acquire 69.99% of the shares of Shangwei New Materials at a total price of 2.1 billion yuan. The general process is divided into three steps:
The first step is for the shareholding platforms Zhipu Hengyue and Zhiyuan Xinchuang to obtain 29.99% of the shares of the listed company from the shareholders of Shangwei New Materials.
The second step is that the two original shareholders of Shangwei New Materials (SWANCOR Samoa, STRATEGIC Samoa) promise to waive the voting rights of all the shares they hold in the listed company. Zhipu Hengyue will become the new controlling shareholder of Shangwei New Materials.
The third step is that the acquisition enters the voluntary tender offer stage. Zhipu Hengyue will continue to increase its holdings of the listed company's shares through a partial tender offer. The number of shares to be tendered for acquisition accounts for 37.00% of the total share capital of the listed company.
After the transaction is completed, Zhipu is expected to hold at least 63.62% of the shares of Shangwei New Materials. The actual controller of Shangwei New Materials will become Deng Taihua, the founder of Zhipu Robotics.
Is this whole acquisition really Zhipu's "backdoor listing"? Actually, it's not, because Zhipu does not meet the qualifications for "backdoor listing".
According to the "Review Rules for Major Asset Reorganizations of Science and Technology Innovation Board Listed Companies on the Shanghai Stock Exchange", Zhipu's financial situation obviously does not meet the conditions for "backdoor listing" (implementing a restructuring and listing). Firstly, the net profit in the past two years should be positive, and the cumulative net profit should not be less than 50 million yuan. Secondly, the revenue in the most recent year should not be less than 300 million yuan, and the cumulative net cash flow from operating activities in the past three years should not be less than 100 million yuan.
However, from the perspective of the acquisition process, the whole process is quite delicately designed.
According to the "Measures for the Administration of the Acquisition of Listed Companies", if the agreement acquisition exceeds 30%, it will trigger a mandatory full - tender offer. At that time, in order to protect the interests of small and medium - sized shareholders, the acquirer has to acquire all the shares of all shareholders, and the acquisition cost and compliance cost of the whole process will become uncontrollable.
So, by disassembling Zhipu's acquisition process this time, it first acquires 29.99% without triggering regulatory intervention. Next, Zhipu gets the control rights. Finally, as the controlling party, Zhipu then increases its holdings of most of the remaining shares through a voluntary tender offer to achieve the goal of being able to advance or retreat. Increasing holdings can not only meet the exit requirements of the original shareholders but also maximize the control of equity to achieve its own commercial interests.
Although Zhipu mentioned in the "Detailed Report on Changes in Equity Interests" that it will not change the main business of the listed company within the next 12 months, nor will it adjust the company's assets and business through sales, mergers, joint ventures or cooperation with others.
However, after 12 months, many operations can be carried out.
For example, if there is a subsequent additional issuance, although the cost of refinancing in the secondary market is higher than that in the primary market, the definite market value can provide more negotiation space than a simple valuation. Or, if Zhipu chooses to rush for listing on the Hong Kong Stock Exchange later, having a clear market value and business within the system will also make it easier to meet the "18C" listing rules designed for special - purpose technology companies.
Did Zhipu carry out this painstaking acquisition because it is short of money?
It's unlikely. As a startup company established for more than two years, Zhipu has carried out 11 rounds of financing in total. Although most of the specific amounts are unknown, considering the situation of three consecutive rounds of financing in 2025, it's hard to say whether Zhipu is really short of money.
Did Zhipu acquire Shangwei New Materials because it saw the integration value?
It's also hard to say. According to public information, the main business of Shangwei New Materials can be roughly considered as "producing coatings and plastics". Its products mainly include materials for wind turbine blades, environmentally friendly high - performance corrosion - resistant materials, new composite materials, recyclable composite materials, etc.
Regarding the combination with the robot business, some investors asked about it in February this year. Shangwei New Materials replied that currently, the humanoid robot shell and accessory industry mainly uses thermoplastic materials, while the company's product materials are thermosetting environmentally friendly corrosion - resistant resins, and the testing direction is the industrial robot industry. In other words, there is almost no meaning of integration between the two companies.
Even considering the risk of integration failure, there are actually huge potential risks in this whole acquisition.
If the market value of Shangwei New Materials is calculated according to Zhipu's valuation, then Zhipu's valuation of over 15 billion yuan is a 5 - fold "lever". Since the pricing in the secondary market basically depends on the value of tradable shares, there is no common gambling agreement or "equity in name but debt in essence" in the primary market. If Zhipu's development fails to meet expectations, it will be a serious destruction of value for investors.
In the capital market, it's always about being bold.
Indeed, Zhipu is not "backdoor listing", and it may not necessarily inject its subsequent assets into Shangwei New Materials. However, this does not prevent extreme risk - takers from regarding Zhipu's "indirect listing" as the wild mushrooms in Yunnan - it may be risky, but it may also be extremely delicious.
Everyone knows that according to the development path of traditional hard - tech startups, it first needs to build a core technology barrier, initially verify the product and technology closed - loop, accumulate strength through financing, and finally, after initially verifying the business model or showing the scale effect, it finally goes through the strict IPO review process to enter the capital market.
However, Zhipu's operation is equivalent to "ripening" the whole traditional model from the capital level. Perhaps, in the eyes of Zhipu and investors, the embodied intelligence industry should tolerate more innovation. Both technology and finance can be "innovated".
Why does Zhipu choose "financial innovation"?
All the contradictions ultimately boil down to one question.
Why does Zhipu go to so much trouble to engage in "financial innovation"? Perhaps it is because while there are some consensuses in the embodied intelligence industry, there are also many contradictions that are difficult to reconcile.
It should be admitted that embodied intelligence is one of the hottest tracks globally.
In February 2024, OpenAI participated in a $745 million financing round of the robot company Figure. Subsequently, on the Figure 01 robot, people saw that the robots powered by large models had a significant upgrade in intelligent capabilities, and the intelligence of robots officially entered the next generation.
This year, we have successively seen numerous large, medium, and small factories join the embodied intelligence track. The endless robot sports "challenges" (such as marathons, fighting, etc.) and even the "price war" at robot exhibitions can prove that the ever - expanding popularity of the embodied intelligence industry has reached a new height.
"The next wave of artificial intelligence is to understand the physical world, and within 10 years, there will be robot systems that can perform tasks. It will completely change the existing factory model, with software and robot teams working together."
As Jensen Huang, the founder and CEO of NVIDIA, mentioned at the China International Supply - Chain Promotion Expo, AI and embodied intelligence are expected to change all industries.
Lightcone Intelligence deeply felt this view when participating in the 2024 World Robot Conference. Countless manufacturers present firmly believed that embodied intelligence has a series of application scenarios and a vast market space in household, industrial, educational, and other fields. And this clear product concept is unique in the current AI hardware track where new products are emerging in an endless stream but the positioning is basically not very clear.
So today, capital is willing to increase investment in the embodied intelligence industry. According to statistics from IT Juzi, as of mid - 2025, both the number of investments and the investment amount in the industry have exceeded the whole of 2024.
Data source: IT Juzi
On the other hand, the embodied intelligence industry is still in a very early stage.
In February this year, after Goldman Sachs "urgently" investigated the humanoid robot industry after the Spring Festival, it concluded that humanoid robots still have a long way to go before they can truly take up jobs. In the next 2 - 3 years, humanoid robots cannot reach the same work efficiency as human workers, and practical applications may have to wait 5 - 10 years. In March this year, Zhu Xiaohu, the managing partner of GSR Ventures, also posted an article reflecting that currently, the consensus on embodied intelligence is highly concentrated, but the commercialization path is not clear.
Currently, the largest domestic order for humanoid robots is the procurement project of humanoid biped robot OEM services of China Mobile (Hangzhou) Information Technology Co., Ltd. won by Zhipu Robotics and Unitree Robotics on July 11th.
The total budget of this procurement project is 124.05 million yuan (including tax). Procurement Package 1 is for full - size humanoid biped robots, with a budget of 78 million yuan (including tax), and the winning bidder is Zhipu Robotics; Procurement Package 2 is for small - size humanoid biped robots, computing power backpacks, and five - finger dexterous hands, with a budget of 46.05 million yuan (including tax), and the winning bidder is Unitree Robotics.
Source: "Technical Specification" of the Procurement Project of Humanoid Biped Robot OEM Services of China Mobile (Hangzhou) Information Technology Co., Ltd.
As for what the robots are bought for, it can be seen from the tender documents that they are mainly for "greeting guests".
It should be noted that it's still hard to say how much profit the current "greeting" robots can bring to customers. Referring to the humanoid robot Mornine arranged by Chery to "work" in car 4S stores this year. In this scenario, it can only be said that the display significance of humanoid robots is greater.
If customers cannot generate actual value or directly solve business pain points, it means that the commercialization of these humanoid robots has not reached the stage of a commercial closed - loop.
The current situation of commercialization shows that the humanoid robot industry still faces quite a lot of technical obstacles.
When Tesla decided to "strategically suspend"