---
title: "Change of label before IPO: International capital made a 414 times oversubscription in Momenta"
type: "News"
locale: "en"
url: "https://longbridge.com/en/news/292234424.md"
description: "Momenta went public on the Hong Kong Stock Exchange on July 8, 2026, opening with a rise of over 6% before retreating, ultimately closing at the issue price of HKD 295.6. Despite receiving 414 times oversubscription and support from luxury cornerstone investors such as GIC and Mercedes-Benz, retail investors still faced losses on the first day of trading due to fees. This trend has been pointed out as having obvious signs of market protection, reflecting the market game under international capital operations"
datetime: "2026-07-09T16:02:09.000Z"
locales:
  - [zh-CN](https://longbridge.com/zh-CN/news/292234424.md)
  - [en](https://longbridge.com/en/news/292234424.md)
  - [zh-HK](https://longbridge.com/zh-HK/news/292234424.md)
---

# Change of label before IPO: International capital made a 414 times oversubscription in Momenta

Written by: Xia Yang

On July 8, 2026, the Hong Kong Stock Exchange rang the listing bell for five companies in one day. The largest by market capitalization is called Momenta (06880.HK), claiming to be the "first stock of physical AI."

It opened with a rise of over 6%, reaching a market capitalization of HKD 70 billion, then fell back, dipping below the issue price during trading, and closed exactly at HKD 295.6—precisely the issue price.

This closing price itself is the best footnote for the entire article.

**A "precisely eerie" IPO debut**

Momenta's first day of listing resembled a meticulously choreographed drama.

![Image](https://imageproxy.pbkrs.com/https://inews.gtimg.com/om_bt/OxSWZKghaXtjuXIywJGXhyvwcTEyxPxGkUU-NIHBGP_iQAA/641?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg)

Source: Daily Economic News

It opened at HKD 301, about 1.83% higher than the issue price of HKD 295.6, peaking at HKD 314.8 during trading, with a rise of over 6%, and the total market capitalization briefly exceeding HKD 70 billion. However, the situation changed dramatically in the afternoon, with the stock price rapidly falling, briefly dipping below the issue price, and ultimately closing precisely back at HKD 295.6.

For retail investors participating in the IPO, this was almost the worst outcome. With 20 shares per lot, the entry fee was HKD 5,971.6. Selling at the closing price resulted in zero profit or loss, with a loss of about HKD 60 for the subscription fee and selling commission; plus interest if margin trading was used. The 414 times oversubscription for this stock turned out to be a futile effort.

Some observers bluntly stated: "The trend of 'precisely closing at the issue price' shows that the skills in supporting the stock are more impressive than the market itself."

And indeed, the lineup supporting the stock was quite luxurious.

14 cornerstone investors locked in 49.95% of the issuance. The Government of Singapore Investment Corporation (GIC) and Fidelity International each invested USD 100 million as lead investors, BlackRock invested USD 25 million, Oak Tree Capital invested USD 20 million, and Franklin Templeton invested USD 10 million. Among the industry participants, Mercedes-Benz and BYD invested USD 25 million and USD 15 million, respectively. Chinese institutions such as Gao Yi, Boyu, Huaxia Fund, Guangfa Fund, and Pacific Insurance each invested USD 10 million.

The public offering portion was oversubscribed by 414 times, and the international offering received over HKD 100 billion in institutional orders, totaling an oversubscription of about 44 times.

With half of the supply frozen upon listing, a greenshoe option in place, and an all-star lineup supporting it—this configuration only managed to maintain a non-breach of the issue price at closing.

The stock price trend of Momenta (06880.HK) on July 9 (the day after listing) can be summarized in four words: calm and steady.

**What kind of company is this?**

Setting aside the glamorous exterior of "physical AI," Momenta's financial fundamentals are actually the most solid among this batch of autonomous driving IPOs From 2023 to 2025, the company's revenue is projected to be 743 million yuan, 1.325 billion yuan, and 2.413 billion yuan, respectively, tripling over three years, with an average annual compound growth rate exceeding 80%. The gross profit margin will increase from 17.5% to 71.6%. Among these, software licensing revenue will soar from 23 million yuan in 2023 to 968 million yuan in 2025, growing over 41 times in three years, accounting for 40.1% of total revenue.

In terms of losses, the net loss attributable to shareholders is indeed expanding year by year. From 2023 to 2025, it will be 2.57 billion yuan, 3.206 billion yuan, and 3.458 billion yuan, with a cumulative total of 9.234 billion yuan over three years. However, the core source of the loss is non-cash items such as changes in the fair value of preferred shares. The adjusted net loss will significantly narrow from 1.093 billion yuan in 2023 to 303 million yuan in 2025, with the adjusted net loss rate dropping from 147.2% to 12.6%.

In terms of R&D investment, the total R&D expenditure for 2025 will reach 1.869 billion yuan, accounting for 77.5% of annual revenue, with a cumulative R&D investment of 4.66 billion yuan over the past three years. By the end of 2025, the company will have 1,157 R&D personnel, accounting for nearly 82%, with more than two-thirds holding a master's degree or higher.

Regarding market share, according to data from the China Association of Automobile Manufacturers, from January to November 2025, the number of Momenta's urban NOA installations reached 414,400 units, accounting for approximately 61.06% of third-party suppliers. According to CIC Consulting, from March 2025 to February 2026, this figure will further rise to 65%, ranking first in the industry. Among the top ten global automakers, nine have already partnered with Momenta.

As of the listing date, the number of mass-produced vehicles equipped with Momenta's solutions has exceeded 1 million. The company has secured over 210 designated models.

The financials are solid, but the issue lies in the market's valuation logic, which has long exceeded these financials.

**"Physical AI": Truth or Label?**

On the eve of its IPO, Momenta quietly changed its label: from "autonomous driving company" to "a global leader in physical AI."

This switch reflects a collective identity transformation occurring in the autonomous driving industry. As urban NOA transitions from a technical highlight to a standard feature for automakers, and as advanced driving solution providers become embroiled in price competition, customer acquisition, and mass production delivery, merely discussing "intelligent driving" has become insufficient to sustain higher expectations. Physical AI has become the new label and a new language for valuation.

However, this switch has also sparked sharp criticism. Critics bluntly state that Momenta is "named physical AI, but its financials are from autonomous driving"—currently, 100% of its revenue still comes from intelligent driving solutions. New businesses related to "physical AI," such as robotics, have yet to contribute any revenue However, if we broaden our perspective to the entire industry chain, we find that Momenta's "physical AI" narrative is not an isolated case, but rather an inevitable choice under industry division of labor. Zhaoguan Electronics, sharing the same "hat," has chosen a different path—focusing on "eyes" rather than "brain."

Founded by a former chip director at AMD, this company develops its own VPU visual chips specifically to solve the problem of "seeing" in the physical world. Its products have landed in eight major fields and over 200 downstream scenarios, with revenue growth of 500% year-on-year in the first five months of 2026. In tests with some leading robotics clients, its chips achieved an environmental recognition accuracy rate 12% higher than industry giant Intel, while reducing power consumption by 30%.

The rise of Zhaoguan Electronics precisely confirms that the physical AI track is not merely a conceptual hype, but a genuinely differentiating industry. Momenta and Zhaoguan Electronics represent two approaches: one entering from the high-dimensional perspective of software and data, attempting to define a universal "brain"; the other rooted in the low-level of hardware and perception, striving to create the most extreme "eyes." The comparison between the two reflects a core proposition of the physical AI industry chain division of labor: in this emerging ecosystem, is the platform value of the "brain" higher, or are the hardware barriers of the "eyes" more solid?

Supporting Momenta's "brain" story is the R7 reinforcement learning world model, which will be mass-produced in April 2026. This is a foundational model capable of understanding physical laws and simulating world evolution. Its technical architecture is divided into three layers: the first layer is the pre-training of the world model, which compresses physical laws, common sense, and causal relationships using vast amounts of real driving data; the second layer is the simulation of the world model, used for closed-loop simulation in autonomous driving, simulating long-tail scenarios; the third layer is reinforcement learning, allowing the system to repeatedly trial and error and autonomously optimize decisions in a virtual training ground.

Momenta CEO Cao Xudong stated that this marks the transition of intelligent driving from "seeing the world" to "understanding the world," with physical AI officially moving from concept to large-scale production.

On the same day of the listing, Cadillac announced the industry's first mass-produced model equipped with the Momenta R7 world model. The SAIC Volkswagen ID. ERA 9X also launched as the first high-level intelligent driving mass-produced model equipped with the R7 world model.

However, a key fact is that the current application scenarios of the R7 world model are still concentrated in the intelligent driving field. It still has a long way to go to reach the physical AI blueprint described by Cao Xudong, which covers passenger cars, Robotaxis, Robovans, and even Robotrucks and embodied intelligence.

Some analysts point out that Momenta's valuation anchor is not Horizon Robotics or Pony.ai, but whether it can become a "platform rent collector" in the intelligent driving track. The narrative of physical AI essentially seeks a more imaginative valuation language for this "platform rent" business model.

**Capital Discrepancy: International Heavy Investment, Domestic Caution** Momenta's IPO is like a mirror, clearly reflecting two different capital logics.

International capital is paying for dreams. GIC, Fidelity International, BlackRock, Oak Tree Capital, Franklin Templeton... these top international long-term funds and sovereign funds are heavily invested. They are buying into the grand narrative of "physical AI," betting that Momenta can become the "universal brain" driving cars, robots, and all physical entities.

The international offering part covers sovereign and long-term funds from 15 countries and regions. The cornerstone subscription totaled approximately HKD 3 billion, accounting for about 49.95% of the total shares offered globally.

Domestic industrial capital, on the other hand, is focused on realistic valuations. Companies like SAIC, General Motors, Mercedes-Benz, Toyota, and BYD are also shareholders of Momenta, but their attitudes are more complex. On one hand, Momenta is an important supplier of intelligent driving technology for them; on the other hand, they are also developing their own intelligent driving solutions and remain vigilant about external suppliers, always ready to "replace them at any time."

This skepticism about valuation can also be corroborated by the situation of another company in the same field. Zhaoguan Electronics, which focuses on creating "eyes" for physical AI, has its core product as a self-developed VPU visual chip, following a "hardcore chip" route. Although its D round valuation is only about 1.5 billion yuan, far less than Momenta, its growth path aligns more with domestic industrial capital's preference for "hard technology"—its products have been implemented in over 200 scenarios, with a revenue growth of 500% in the first five months of 2026. In contrast, Momenta's cumulative net loss of over 9.2 billion yuan in the past three years, along with the vague expectation that it needs to reach a capacity of 4 million vehicles to become profitable, presents a higher risk in the eyes of local capital that seeks certainty.

Some analysts point out that the market often compares Momenta with Horizon Robotics, the latter's valuation being three times that of Momenta. Critics argue that Momenta's "pure software" business model has less technical barrier and profitability visibility compared to models that include hardware like chips.

As of the listing date, Horizon's total market value was USD 7.58 billion, and Pony.ai's was USD 3.017 billion, both lower than Momenta's pricing. How much of this pricing is supported by "physical AI" remains unknown, but it is certain that the majority should come from the fundamental data in the prospectus.

**A daring leap into the future?**

Momenta's future path is essentially a daring leap "from a leader in intelligent driving to a physical AI platform."

The upward momentum is clear. A 65% market share in the third-party city NOA, 1 million mass-produced vehicles equipped, a gross margin of 71.6%, and a tripling of revenue over three years—these constitute the most solid fundamentals. Cao Xudong has provided a clear timeline: losses will continue to narrow in 2026, break-even will be achieved in 2027, and profitability will be realized in 2028. Before the IPO, the company had approximately 10 billion yuan in cash, and after the IPO, it supplemented with several billion yuan, totaling nearly 20 billion yuan The downward resistance is equally real. On the competitive front, Momenta's "pure software" model is facing dual pressure from giants like Huawei and Horizon, which adopt a "soft and hard integration" approach, as well as from automotive companies that are pursuing "full-stack self-research." Horizon's founder, Richard Yu, published a statement as Momenta sprinted towards its IPO, claiming that "Horizon is unlikely to mix in society and has never boasted about being the 'first stock in XX'"—a statement filled with tension.

On the client side, major clients such as BYD and Mercedes-Benz are also players in self-research smart driving. This "same bed, different dreams" relationship puts Momenta at constant risk of client loss. The revenue from its top five clients accounts for over 60%, and the operational uncertainty cannot be ignored.

In terms of narrative, the label of "the first stock in physical AI" is experiencing rapid inflation. In December 2025, Wuyishijie will land on the Hong Kong Stock Exchange as the "global first stock in Physical AI"; in June 2026, Haiqing Zhiyuan will substantiate the title of "first stock in physical AI" with a first-day surge of 270%. When everyone claims to be the "first stock in physical AI," how much value does this label retain?

On the profitability front, Cao Xudong admits that the company needs to reach a deployment volume of 4 million vehicles to potentially become profitable. Currently, it has just surpassed 1 million vehicles. The gap between 1 million and 4 million is a full 3 million vehicles, countless point battles, and innumerable price wars.

In terms of strategy, Momenta has chosen a "horizontal platform" path—using a single world model to conquer all physical terminals. In contrast, its peer, Zhaoguan Electronics, has opted for a "vertical deep cultivation" path—focusing solely on being the "eyes" of the physical world and perfecting visual perception.

Both strategies have their pros and cons. The appeal of Momenta's model lies in "winner takes all"; once it becomes the universal brain of physical AI, its valuation could be limitless. However, the risk is that by trying to cater to too many scenarios, it may encounter entrenched competitors in each niche. The allure of Zhaoguan Electronics' model is its "irreplaceability"; when your chip becomes the industry standard, you gain pricing power. However, the limitation is that the ceiling is relatively clear.

The comparison between Momenta and Zhaoguan Electronics is, in fact, a microcosm of a deeper issue in the physical AI industry: in this emerging supply chain, will the platform that masters the "brain" ultimately prevail, or will the hardcore tech companies that control the "senses" be better positioned to navigate cycles? The answer may emerge in five years, but Momenta must prove itself before then.

Momenta's IPO debut is a successful "face project"—under the strong support of 14 cornerstone and greenshoe investors, it maintained its issue price.

But the real test has just begun. Is "physical AI" a natural extension of Momenta's technical route, or is it a new story that autonomous driving companies must tell as the capital window narrows? The rise of Zhaoguan Electronics reminds us that physical AI is differentiating into a complex system driven by both "brains" and "eyes," as well as "software" and "hardware." Understanding Momenta requires looking not only at itself but also at its position within the entire industry chain.

Cao Xudong said at the listing ceremony: "We have set three ten-year visions: saving one million lives in ten years, freeing up 100% of time in ten years, and doubling logistics and travel efficiency in ten years."

The vision is grand, but the patience of the capital market usually does not last ten years. Whether Momenta can successfully transform the grand narrative of "physical AI" into one scalable and sustainably profitable business after another before the capital's patience runs out will be one of the most noteworthy business stories in the next two years.

As an observer put it: "The real test is in January 2027."

### Related Stocks

- [06880.HK](https://longbridge.com/en/quote/06880.HK.md)
- [GIC.US](https://longbridge.com/en/quote/GIC.US.md)
- [00388.HK](https://longbridge.com/en/quote/00388.HK.md)
- [BLK.US](https://longbridge.com/en/quote/BLK.US.md)
- [BEN.US](https://longbridge.com/en/quote/BEN.US.md)
- [01211.HK](https://longbridge.com/en/quote/01211.HK.md)
- [002594.CN](https://longbridge.com/en/quote/002594.CN.md)
- [02601.HK](https://longbridge.com/en/quote/02601.HK.md)
- [601601.CN](https://longbridge.com/en/quote/601601.CN.md)
- [AMD.US](https://longbridge.com/en/quote/AMD.US.md)
- [INTC.US](https://longbridge.com/en/quote/INTC.US.md)
- [600104.CN](https://longbridge.com/en/quote/600104.CN.md)
- [GM.US](https://longbridge.com/en/quote/GM.US.md)
- [7203.JP](https://longbridge.com/en/quote/7203.JP.md)
- [TM.US](https://longbridge.com/en/quote/TM.US.md)
- [09660.HK](https://longbridge.com/en/quote/09660.HK.md)
- [06651.HK](https://longbridge.com/en/quote/06651.HK.md)
- [01392.HK](https://longbridge.com/en/quote/01392.HK.md)
- [80388.HK](https://longbridge.com/en/quote/80388.HK.md)
- [HKXCY.US](https://longbridge.com/en/quote/HKXCY.US.md)
- [BTX.US](https://longbridge.com/en/quote/BTX.US.md)
- [BIT.RT*.US](https://longbridge.com/en/quote/BIT.RT*.US.md)
- [BDJ.US](https://longbridge.com/en/quote/BDJ.US.md)
- [BSTZ.US](https://longbridge.com/en/quote/BSTZ.US.md)
- [BIT.RT.US](https://longbridge.com/en/quote/BIT.RT.US.md)
- [HYDD.SG](https://longbridge.com/en/quote/HYDD.SG.md)
- [BYDDY.US](https://longbridge.com/en/quote/BYDDY.US.md)
- [BYDDF.US](https://longbridge.com/en/quote/BYDDF.US.md)
- [81211.HK](https://longbridge.com/en/quote/81211.HK.md)
- [TOYOF.US](https://longbridge.com/en/quote/TOYOF.US.md)

## Related News & Research

- [Momenta makes public debut in Hong Kong, puts physical AI in spotlight](https://longbridge.com/en/news/292088574.md)
- [NIO Capital-backed Momenta lists in Hong Kong under](https://longbridge.com/en/news/292463659.md)
- [AI glasses have a new hater: Lorde](https://longbridge.com/en/news/292517914.md)
- [The whirlwind 72 hours of rival AI announcements](https://longbridge.com/en/news/292338769.md)
- [Our AI investments in China focus on physical AI and applications, says Temasek](https://longbridge.com/en/news/292165452.md)