Leverage the “Law of Large Numbers” to secure long-term competitive advantages in an industry fraught with uncertainties.
On the last day of August 2026, Momenta, which had just listed on the Hong Kong Stock Exchange for over a month, released its first earnings report.
The figures are indeed impressive: revenue in the first half of the year reached 1.602 billion yuan, representing a 75.9% year-on-year increase. What draws more attention is the loss issue that has been widely discussed across the industry: the adjusted net loss narrowed by 96.6% to 14.097 million yuan.
Key financial indicators of Momenta in the first half of 2026
For a company that wears the halo of “the first stock of physical AI” and has long been questioned for “burning money”, this result almost conveys a message: just give me a little more time, the break-even point is right around the corner.
In the subsequent earnings call, the management focused on a longer-term layout: cover more vehicle models through the “blanket coverage” mass production strategy, keep the million-unit-scale driven flywheel accelerating continuously, and pave the way for the global implementation of Robotaxi and Robovan in the next three years.
“Competition in the automotive industry is becoming increasingly fierce. Simply betting on a single vehicle model and pinning hopes on creating a one-off hit is not enough to support the long-term survival of an enterprise.” Therefore, Momenta takes the “Law of Large Numbers” as the underlying logic of its strategy. Through extensive deployment of a large number of OEMs and vehicle model matrices, the market risks brought by a single automaker or vehicle model can be effectively hedged. Our strategy strives to secure long-term competitive advantages in the fast-changing and highly uncertain automotive industry.
Financial Performance: From “Money-burning Machine” to “Approaching Profitability”
In the past few years, the two words that the autonomous driving industry feared most were “loss”. R&D investments amounting to billions of yuan have caused many star companies to fall on the eve of commercialization.
The signal released by Momenta’s this earnings report is very clear: Large-scale operation is reshaping the financial model.
Earnings data shows that in the first half of 2026, Momenta achieved an operating revenue of 1.602 billion yuan, a 75.9% year-on-year increase.
Revenue from Momenta’s two core businesses
Both of the two major businesses are high-margin software and service businesses with extremely low marginal costs, which also drives the profit data to improve.
In the first half of this year, Momenta’s gross profit reached 1.173 billion yuan, up 79.4% year-on-year, with a gross margin of 73.2%, higher than the 71.8% in the same period of last year. This is largely attributed to its business model: most of Momenta’s revenue comes from pure software algorithm licensing and services, without the cost pressure of hardware.
Based on the adjusted net profit caliber that can better reflect the actual operating conditions (excluding non-cash items such as equity incentives, listing expenses, and changes in fair value of preferred shares), Momenta’s adjusted net loss in the first half of the year was 14.097 million yuan, which narrowed significantly by 96.6% compared with 416 million yuan in the same period of last year. It is less than 15 million yuan away from the break-even point.
Momenta’s loss status as of the end of June this year
In accordance with IFRS standards, the loss for the period was 16.537 billion yuan, a year-on-year increase of 871.8%. Thehe financial statements: 16.31 billion yuan comes from “changes in the fair value of preferred shares and other financial liabilities”
Simply put, the company issued a large number of preferred shares before its listing. In accordance with the requirements of International Financial Reporting Standards, these preferred shares need to be re-measured at fair value at the end of each reporting period. The higher the company’s valuation, the higher the fair value of preferred shares, and the greater the resulting “book loss”. This is an accounting treatment result required by accounting standards, which is essentially an advance record of the equity conversion cost that may occur in the future, and will not lead to future cash outflow. Moreover, all preferred shares have been converted into common shares after the listing in July, and the financial statements in the second half of the year will no longer be affected by this.
However, the figure of 367% for the asset-liability ratio still deserves vigilance. This is not “amplified by the accounting treatment of preferred shares”, but the result of pushing up the total liability caliber after preferred shares are included in financial liabilities before listing, reflecting the capital structure of the company that relied on financing for expansion in the past few years. With the completion of the IPO and the conversion of preferred shares into shares, this liability will drop significantly and net assets will increase substantially, so this indicator will improve obviously in the future.
In terms of R&D, as of June 30, 2026, the company has 1102 R&D personnel, accounting for 79.1% of the total number of employees. In the first half of this year, R&D expenses were 1.16 billion yuan, up 18.6% year-on-year, and the R&D expense ratio was 72.6%.
In terms of cash flow, the earnings report shows that as of June 30, 2026, the net outflow of cash flow from Momenta’s operating activities was 381 million yuan, narrowing by 47% compared with 716 million yuan in the same period of last year. With the IPO fundraising of about 6.8 billion Hong Kong dollars in July, the company’s cash reserves have exceeded 15 billion yuan. In the short term, the capital reserve is sufficient to support subsequent R&D investment and global expansion.
Business Performance: Clarify the “Blanket Coverage” Strategy
Why can it get close to profitability so quickly? The answer lies in the “mass production scale”.
Let’s look at a set of comparative data first.
In 2022, Momenta delivered its first 100,000-unit mass production solution, which took 24 months. By the first half of this year, the latest 100,000-unit delivery only took 36 days. From “two years” to “more than one month”, the time unit has been compressed from “year” to “month”.
In terms of specific figures, as of June 30, Momenta has covered 26 OEM customers, with a cumulative total of 105 mass-produced vehicle models, and the cumulative installation volume of mass production solutions has exceeded 1 million units. About 321,000 new units were added in the first half of the year, up 83.7% year-on-year. The increment in just half a year exceeded the total cumulative volume before 2022. Its cooperative brands include Mercedes-Benz, BMW, Audi, GAC Toyota, Hongqi, SAIC Volkswagen, Dongfeng Nissan, IM Motors, etc., and it has reached mass production business cooperation with 26 OEMs.
The greater value of 1 million vehicles running on the road is generating data.
The earnings report discloses that as of the end of June, Momenta has accumulated more than 13 billion kilometers of real driving mileage relying on over 1 million mass-produced vehicles, from which more than 100 million segments of high-value “golden data” have been extracted, with a daily simulation mileage of over 10 million kilometers.
The more data accumulates, the faster the algorithm runs, and the mass production scale in turn accelerates data accumulation, forming a positive feedback loop, behind which Momenta’s flywheel effect is being fully released.
Of course, high data barriers do not mean there are no challenges.
On the one hand, in the supply chain security logic of automakers, core technologies cannot be “supplied exclusively by one party”, and they will insist on independent R&D. While SAIC invested in Momenta, it also introduced Huawei’s solutions; Toyota’s intelligent layout is also diversified, neither giving up independent R&D nor betting all its chips on a single supplier. This means that Momenta is not only facing competition for order acquisition, but also the squeeze on market space brought by the automakers’ independent R&D routes.
On the other hand, Momenta is continuously increasing its efforts to expand overseas. The compliance cost of overseas markets, geopolitical risks, and the differences in data supervision in various regions are all variables that require continuous investment to cope with in the process of large-scale expansion.
In addition, up to now, Momenta has a total of 105 mass-produced vehicle models, but there is a gap of more than 110 models between the number of delivered models and the cumulative number of over 219 designated models. There are still risks of changes in mass production rhythm and customer cooperation as to whether these orders can be smoothly converted into deliveries.
On the whole, the “flywheel” has been spinning, with accelerating scale, accumulating data, and expanding market share. The next thing to see is whether this flywheel can maintain acceleration, and whether its technical leading edge can be continuously converted into commercial orders.
At the level of commercial strategy, the “blanket coverage” strategy was clearly put forward in the conference call. The essential goal is to expand the business coverage as much as possible, access more OEMs, and cover a more diversified vehicle model matrix.
The underlying business logic is that competition in the automotive industry is becoming increasingly fierce, and a single hit vehicle model is difficult to support long-term survival. Through extensive deployment of a large number of OEMs and vehicle models, the market risks of a single automaker or vehicle model can be effectively hedged, so as to strive for long-term competitive advantages in the uncertain industry with the “Law of Large Numbers”.
Next Step: Implementation of Robotaxi and Global Market
If Momenta only sells NOA solutions, its story can at most be regarded as a “profitable good business”, but the capital market is willing to give a high valuation, which must be based on its next growth step.
The earnings report discloses that the self-developed Momenta R7 world model will be deployed in mass-produced vehicles starting from the third quarter of 2026.
Three-layer architecture of Momenta R7 world model
The so-called “world model” means enabling AI to understand the laws of the physical world, rather than just identifying and classifying. In the scenario of autonomous driving, it means that the system has human-like “commonsense reasoning” ability, which can predict the causal relationship between dynamic objects on the road, rather than just responding to static rules.
If this technology is successfully implemented on a large scale, it may open a generation gap at the algorithm level. But it should be noted that it still takes a period of data accumulation and iterative verification from the deployment of the world model to the generation of actual experience differences. The deployment on vehicles in the third quarter is just the starting point, and the user-perceivable experience advantage may not be formed until 2027.
The larger layout is: Robotaxi and the global market.
Momenta teams up with Xiangdao Mobility to build a pre-installed mass production-based Robotaxi fleet
The earnings report shows that Momenta’s Robotaxi has carried out tests or operations in 6 cities across 3 countries around the world, and plans to launch its first Robotaxi in the fourth quarter of 2026.
According to the detailed schedule disclosed in the conference call: the goal by the end of 2026 is to obtain test licenses in more than 10 cities around the world, with a fleet size of hundreds of units; in 2027, it plans to expand to more than 30 cities around the world, and the algorithm safety performance is expected to reach 10 times that of ordinary human drivers; 2028 will be the first year of large-scale operation, and Robotaxi will transform from pilot trial operation to normalized commercial mobility services, completing the key leap from technical leadership to large-scale commercial operation.
