On July 8, 2026, Momenta debuted on the Hong Kong Stock Exchange at an IPO price of 295.6 Hong Kong dollars, giving it a market capitalization of approximately 69.6 billion Hong Kong dollars. The stock initially surged to 314.8 Hong Kong dollars on the first day of trading but ultimately closed at the offer price, supported by the greenshoe mechanism. Throughout the 30-day stabilization period, the greenshoe mechanism was gradually exhausted, yet it still failed to sustain the stock price, which at one point fell to a low of 243.6 Hong Kong dollars per share. As of August 14, Momenta's stock was trading at 268 Hong Kong dollars, representing a decline of approximately 9.34% from the IPO price, with a market value of about 63.1 billion Hong Kong dollars.

Momenta's financial performance does not appear poor. From 2023 to 2025, revenue increased from 743 million yuan to 2.413 billion yuan, demonstrating a compound growth rate exceeding 80%. Gross margin improved from 17.5% to 71.6%, while the adjusted net loss narrowed from approximately 1.09 billion yuan to 303 million yuan. Despite this positive trajectory, why did the stock still fall below its IPO price? The market widely used Horizon Robotics as a comparable company for Momenta's pricing. During the book-building phase, Horizon Robotics had just hit a stage low of 52.9 billion Hong Kong dollars, with the market reflecting considerable pessimism. By the time Momenta officially listed on July 8, Horizon Robotics' market capitalization had rebounded to 64.6 billion Hong Kong dollars. Momenta, with an IPO market cap of about 69.6 billion Hong Kong dollars, had a revenue scale in 2025 that was only 60% of Horizon Robotics, yet its issue market value was still approximately 5 billion Hong Kong dollars higher than that of Horizon Robotics on the listing day. This is more clearly seen in the static 2025 price-to-sales ratio. Momenta's IPO market capitalization, including the greenshoe, corresponded to a 2025 price-to-sales ratio of approximately 28.8 times, while Horizon Robotics on the day of Momenta's listing had a 2025 price-to-sales ratio of about 14.9 times. Momenta carried a nearly doubled price-to-sales premium, and it needed to adequately demonstrate to the market why it deserved this valuation premium.

Why the "Physical AI" narrative failed to materialize

Just before its listing, Momenta shifted its positioning from a "passenger car intelligent driving solution provider" to the "world's first physical AI stock." However, the main problem with this narrative is that 100% of the company's revenue still comes from autonomous driving solutions for passenger cars, with no commercialized revenue from robotics or industrial AI. Notably, the IPO prospectus only mentions the R7 World Model once and does not include the term "physical AI" at all, suggesting the concept was more of a new packaging for external promotion before the listing. While Momenta's R7 World Model has technological foresight, the distance from concept to commercialization remains vast. Robotaxi operations and overseas business are still in their early stages, while extended applications like humanoid robots and industrial AI remain in the research and development phase. Each new concept proposed is an attempt to break through the existing valuation ceiling, but the market ultimately only recognizes one thing: commercial implementation. The market's reaction was quickly reflected in the market cap trends of the two companies. After listing, there was a complete reversal in the market capitalizations of Momenta and Horizon Robotics. Before the listing, although Horizon Robotics served as an important pricing anchor, its market cap was at a low point, while Momenta had a clear premium in both market value and price-to-sales ratio. After listing, Momenta's stock price remained below the issue price, while Horizon Robotics' stock price gradually rose from its lows. From late July onward, Horizon Robotics' market cap surpassed that of Momenta. As of August 14, Horizon Robotics had a total market capitalization of approximately 79.9 billion Hong Kong dollars, while Momenta was at about 63.1 billion Hong Kong dollars, with the gap widening to approximately 16.7 billion Hong Kong dollars. It is worth noting that Momenta's current 2025 price-to-sales ratio has fallen to 22.7 times, which is still higher than Horizon Robotics' 18.4 times. Despite the significant price drop, Momenta still retains a price-to-sales valuation premium. However, a high price-to-sales ratio does not necessarily mean a high market cap. Horizon Robotics' revenue scale is about 1.5 times that of Momenta, and a larger revenue base multiplied by a lower price-to-sales ratio can still generate a higher market capitalization. This is precisely why Momenta has a price-to-sales premium but its market value lags behind.

The price-to-sales premium: Market recognizes Momenta's short-term efficiency advantage

The current price-to-sales premium for Momenta reflects the market's recognition of its operational efficiency. First, it has a higher gross margin and smaller losses. Momenta's gross margin is 71.6%, with an adjusted net loss of only 303 million yuan. In contrast, Horizon Robotics has a gross margin of 64.5% and a full-year 2025 adjusted net loss of 2.812 billion yuan. Momenta primarily sells pure software products, where marginal costs approach zero, while Horizon Robotics' chip business must continuously bear rigid hardware costs such as tape-outs, materials, and inventory. Second, Momenta is closer to breakeven. In the first half of 2026, Horizon Robotics is expected to report an adjusted net loss of 1.4 billion to 1.7 billion yuan, expanding by 5.1% to 27.6% year-over-year. Momenta's losses have stabilized at around the 300 million yuan level, bringing the breakeven point significantly closer. The market is willing to pay a valuation premium for a company that is "achieving profitability sooner." Third, the software business has global replicability. Momenta's client base includes major global automakers such as Mercedes-Benz, BMW, Audi, Volkswagen, Toyota, and General Motors. As of February 2026, it had partnered with 24 automakers, including nine of the world's top ten, and held a 65% market share in third-party city NOA, ranking first in the industry. Once a software solution is developed, the marginal cost of replicating it for global automakers is extremely low. In contrast, Horizon Robotics' chip expansion overseas requires repeatedly completing automotive-grade certifications and supply chain adaptations in different countries, incurring higher expansion costs.

Long-term risks behind the valuation premium from operational efficiency

Once Horizon Robotics chips are mass-produced in vehicles, the cost for automakers to switch the entire computing platform platform is very high, requiring new hardware adaptation, safety certifications, and extensive road testing, which naturally creates strong customer lock-in and more sustainable revenue. Although Momenta's pure software solutions do not have chip-level hardware lock-in, switching a full suite of intelligent driving algorithms still requires more than half a year of adaptation and validation, so automakers are unlikely to change suppliers arbitrarily in the short term. However, a prominent long-term risk is that leading automakers like BYD, Li Auto, and NIO continue to develop full-stack self-driving capabilities in-house, and a mature in-house system will continuously compress the overall market space for third-party software providers. While Momenta has initiated plans to develop its own chips to fill the hardware gap and build long-term barriers, this new business cannot contribute substantial revenue in the short term. Meanwhile, Horizon Robotics is also expanding into areas like vehicle operating systems and general physical AI, intensifying competition. The external industry environment further suppresses valuations. The ongoing price war among automakers is causing them to cut budgets for high-level autonomous driving, and the secondary market's tolerance for unprofitable hard-tech growth stocks is decreasing. However, Horizon Robotics also faces operational challenges. Its reported net profit of 3.5 billion to 4.0 billion yuan in the first half of 2026, which appears to show a turnaround, is actually derived from the fair value gains on CARIAD convertible bonds and disposal gains from the terminated merger with D-Robotics, both non-cash items unrelated to core operations. After stripping these out, the adjusted operating loss has widened year-over-year. In 2025, its R&D investment was 5.154 billion yuan, a 63.3% increase, accounting for 137.1% of revenue, indicating a long path to profitability. This explains why, although Horizon Robotics' market cap is higher than Momenta's, the market has not granted it a high valuation premium either.

Conclusion

Momenta still has opportunities for a valuation recovery. If its self-developed chips enter mass production, its physical AI business achieves commercialization, or its global overseas orders ramp up significantly, any one of these scenarios could reshape its valuation system. However, until then, the market will continue to question a core issue: As automakers' full-stack in-house development becomes an industry trend, how can a pure software intelligent driving company lacking a hardware moat maintain its long-term irreplaceable industrial value?