Summary of Key Points
Momenta is a company that develops intelligent driving software and went public on the Hong Kong Stock Exchange on July 8th, with an issue price of HK$295.6 and a market value of approximately HK$69.6 billion. The company does not manufacture vehicles or chips; instead, it earns revenue by providing intelligent driving solutions to automakers. Its customer base includes 24 automakers globally (all major domestic automakers in China and 9 out of the top 10 global automakers), with over one million production vehicles already equipped with its systems. Its core strategy is to "walk on two legs": on one hand, it focuses on mass-producing L2 assisted driving solutions to generate current revenue and collect data; on the other hand, it invests in developing L4 autonomous driving technology to prepare for the future, creating a "data flywheel" cycle. Its business model has evolved from charging per project to licensing software by vehicle, significantly improving its revenue structure and gross margin. Momenta positions itself as the "first stock in physical AI," aiming to expand its technology beyond intelligent driving into areas such as robotics. However, it is not yet profitable and faces competition from automakers developing their own solutions. Investors are optimistic about its market share and the potential of physical AI, but whether its vision will become a reality remains to be seen.
Detailed Analysis
1. Business Model: Moving from Project-Based Revenue to Software Licensing
Previously, Momenta primarily provided customized intelligent driving solutions to automakers, charging for each project (a traditional "engineer-hours" model with a gross margin of only 17.5%). Now, it licenses its software, earning revenue with every vehicle sold, resulting in nearly zero marginal costs (for example, developing one solution can generate revenue 100 times if 1 million vehicles use it).
Data shows that from 2023 to 2025, licensing revenue increased from HK$23 million to HK$968 million (a 42-fold increase), accounting for 3.1% to 40.1% of total revenue, with the gross margin soaring to 71.6%. This represents a significant shift from labor-intensive to profit-driven model.
2. "First Stock in Physical AI": A Realistic Vision or Just a Story?
Momenta insists on being called the "first stock in physical AI" for practical reasons: previous autonomous driving companies (such as Pony.ai and Wen远知行) have had mediocre stock performance, and the intelligent driving market is highly competitive (with city-based NOA systems becoming standard and solution providers competing on price). Physical AI refers to the ability of AI to understand the laws of the physical world (for instance, knowing how inertia affects vehicle movement and needing to decelerate in advance). Momenta claims this technology can be applied to robotics and home services (e.g., robots avoiding obstacles while cleaning). The question arises: is this a natural progression of technology or just a bold claim? The key lies in whether its R7 model can truly be extended from vehicles to Robotaxi and other applications. Cao Xudong plans to launch the robotics business in 2027 with 10,000 Robotaxis in operation by 2028; if this succeeds, the company's valuation could soar. Otherwise, investors may lose interest.
3. Investor Interest: More Than Just Intelligent Driving
Momenta has an impressive list of cornerstone investors, including the Singapore Government Investment Corporation and Fidelity International, with each investing $100 million. Mercedes-Benz and BYD also contributed, and 14 institutions subscribed for a total of HK$3 billion (nearly half of the issuance amount), indicating high demand for shares.
Investors are attracted by two main factors: first, Momenta's market share in China's advanced intelligent driving market (65%), making it a leader; second, the potential of physical AI. While digital AI (like ChatGPT) has gained popularity, physical AI is seen as the next wave, with autonomous driving being the only field that combines both large-scale data collection and commercial success.
4. Challenges Ahead
- Profitability Still a Long Way Off: Despite a reduction in net losses from HK$1.093 billion in 2023 to HK$303 million in 2025, R&D expenses account for 77.5% of revenue, indicating a high cost burden. Investors may lose patience with ongoing losses.
- Intensifying Competition: Companies like Horizon Robotics (in chip manufacturing) are entering the intelligent driving market, competing directly with Momenta. Huawei holds a 14.5% market share and is second in this sector, while other players such as Lightship Intelligence and YuanRong Qixing are also pursuing listings on the Hong Kong Stock Exchange.
- Threat from Automakers' In-house Development: Leading automakers (like BYD and Tesla) are increasing their own R&D efforts in intelligent driving, potentially reducing the reliance on third-party suppliers.
5. The Strategy of "Walking on Two Legs": Why Not Go All-in on L4?
Many startups start with L4 autonomous driving, but Momenta focuses on both L2 and L4 simultaneously. Cao Xudong compares this to climbing a mountain by building a rocket (L4 is the ultimate goal, while L2 is the foundation for gathering data and generating cash flow). This "data flywheel" approach ensures a steady supply of data for L4 development, creating a positive cycle: L2 vehicles generate data for L4 improvement, which in turn drives more L2 adoption and more data collection. Without L2-based production data, L4 technology would be unsustainable.
Conclusion
Momenta's listing represents the transition of AI companies from showcasing their technology to achieving commercial success. The HK$700 billion market value reflects confidence in the potential of "physical AI." If the R7 model can indeed be applied beyond intelligent driving, Momenta could become a new giant. However, if its claims remain unproven, its valuation may shrink. Cao Xudong's vision of creating a "GPT moment" for physical AI within the next decade is ambitious but challenging to achieve.