After the IPO threshold for robotics companies was raised, primary-market investment expectations for embodied AI diverged

Main text

Since regulators issued opinions raising the IPO threshold for robotics companies, capital developments related to such projects have continued to emerge thick and fast.

This week, market reports said Songyan Dynamics has confidentially submitted a Hong Kong IPO application, planning to raise several hundred million US dollars and potentially list as early as 2027. Songyan Dynamics has not yet responded to the report.

A week earlier, Zhiyuan was reported to have launched a new funding round of about 1 billion to 1.5 billion yuan, at a latest valuation of about 50 billion yuan. The report also cited investors' judgment that the new funding would affect its plan to complete an IPO within the year; Zhiyuan has also not responded to questions about the funding, valuation and IPO arrangements.

It is clear that companies are still pushing ahead with capital moves, but investors are already split on whether to keep writing checks.

One investor who has already backed multiple embodied intelligence projects told the Science and Technology Innovation Board Daily that they will continue to follow the sector and look for companies with innovative points and differentiation.

Another AI-sector investment professional at a state-owned-backed institution told the Science and Technology Innovation Board Daily that they had basically stopped looking at embodied intelligence projects since June this year, with the main concerns being project homogenization and a lack of substantive technological progress.

Embodied intelligence: still investing?

On the next moves in the embodied intelligence sector, caution is the shared attitude of the investors interviewed, though their considerations are not entirely the same.

Zhang Zhikuan, a director in the investment department at Cornerstone Capital, told the Science and Technology Innovation Board Daily that being cautious about high-valuation projects is a consistent investment stance of investment institutions and is not specific to embodied intelligence. As the sector gradually forms a leading tier, many companies' valuations are already at relatively high levels, and further investment requires more prudent judgment.

Cornerstone Capital has previously invested in multiple embodied intelligence projects, including Zhiyuan, Zhiyuan Square, Zibianliang, Zhongqing Robotics and Zhuji Dynamics, all of which are now in the 10-billion-yuan valuation club. Regarding these portfolio companies, Zhang Zhikuan said the firm remains bullish on their intrinsic value and, as an investor, will not pressure them toward capitalization because of short-term market fluctuations.

This long-term bullish attitude also extends to attention on new projects. Zhang Zhikuan said Cornerstone Capital will continue to track the development of embodied intelligence and look for companies with innovative points and differentiation, "for example, those that can truly achieve major breakthroughs in the direction of physical AI, or those that can make major innovations in the body field."

If Zhang Zhikuan remains cautiously optimistic about the sector, the above-mentioned anonymous investment professional at a state-owned-backed institution has already grown doubtful about the quality of some current embodied intelligence projects.

The person explicitly told the Science and Technology Innovation Board Daily that they had "stopped looking much at embodied intelligence projects" as early as June this year.

They said that at the time they had already noticed some problems in the embodied intelligence industry: "On the one hand, homogenization is quite serious, and many projects are just tinkering on top of open source and making demos; apart from fundraising, some teams are not genuinely running the company, and no substantive progress or breakthroughs can be seen in frontier research."

Will M&A integration be the next step?

Beyond continued fundraising and pushing through IPOs, some embodied intelligence companies have begun seeking control of listed companies by participating in restructurings.

According to a September announcement by ST Meizhi, a consortium led by Shenpu Intelligence has been confirmed as the selected investor for the company's out-of-court restructuring and subsequent judicial reorganization. Under the plan disclosed at the time, Shenpu Intelligence plans to contribute about 810 million yuan to subscribe to converted shares, and if the restructuring is completed smoothly, it will become the controlling shareholder of ST Meizhi.

Relevant materials show that Shenpu Intelligence was founded in October 2024 and mainly develops general-purpose embodied intelligent robots for life scenarios such as hotels and health care. As of September this year, the company had completed five funding rounds, with investors including Shunwei Capital, BV Baidu Ventures, Didi and Linear Capital.

The Shenpu Intelligence-ST Meizhi deal involves a listed company restructuring and a potential change of control, and is not itself equivalent to Shenpu Intelligence completing a listing. However, against the backdrop of a slower IPO pace, whether acquiring a listed company can become another capitalization path for embodied intelligence companies has also become a topic worth discussing.

Regarding the idea of accelerating capitalization through M&A, Zhang Zhikuan said that if direct listing applications face obstacles, taking another path is not a reasonable way to bypass relevant review requirements. "If it is purely a capitalization maneuver, it is not necessarily a good thing. At this stage, companies should still put more energy into R&D, product capabilities and application scenarios."

But M&A integration centered on business and technology could happen, he believes. "Subsequent integration is more likely to occur between teams with complementary capabilities, for example, a physical AI company acquiring a body team, or a company strong in bodies acquiring a physical AI team, to fill gaps in each other's technical capabilities."

For companies that have already secured large funding, M&A may also become another way to leverage their capital advantage.

The above-mentioned anonymous investment professional told the Science and Technology Innovation Board Daily that high-valuation companies with ample cash on their books can, after the industry cools, acquire some embodied intelligence teams at lower prices and then wait for a technology inflection point. The capital accumulated earlier then becomes a later-stage survival advantage, leaving room for waiting and adjustment.

A possible investment cooling-off period

Earlier this year, the Science and Technology Innovation Board Daily reporter interviewed Zhang Zhikuan on the topic of embodied intelligence companies collectively moving toward 10-billion-yuan valuations. Looking back at the judgment then, Zhang Zhikuan believes that the business and technological development of the embodied intelligence industry is broadly still advancing along the direction he previously expected.

"The leading landscape in the body field has gradually become clear, and the maturity of hardware and motion control is relatively higher. Next, it will increasingly take on the attributes of high-end manufacturing and advanced manufacturing, and the competitive landscape will tend to stabilize. Physical AI, meanwhile, is still in a process of differentiation. Within the next one to two years, the number of leading players may be relatively large, and whether companies can widen the gap will depend on technical capabilities and their willingness and ability to continuously invest in R&D such as model training."

What needs adjustment is the expectation for IPOs. He admitted that he had previously been more optimistic about the listing process, "and now I may be a bit more patient."

In fact, this kind of optimism about IPOs for embodied intelligence projects was once widespread among investors. In interviews at the beginning of the year, multiple investors regarded the relatively clear IPO exit path for embodied intelligence projects as one of the important reasons for the rapid rise in valuations in this round.

For institutions, changes in the listing pace affect exit arrangements and the timing of realizing investment returns. For embodied intelligence companies still in the early stages of commercialization, what they face is a direct operational test of how to sustain large-scale investment when public-market fundraising is delayed.

It can be seen that over the past three years, embodied intelligence has experienced a wave of intensive capital inflows and rapidly rising valuations. Now, the expected pace of IPOs has been adjusted, and the industry is still waiting for technological breakthroughs and commercialization to scale up. How the waiting time of capital matches the development progress of companies will become a new test.

China's red-hot primary market for embodied intelligence may be about to enter a temporary cooling-off period.

(Source: Cailian Press) Middle of the main text Embedded advertisement