The capital market always favors the next big story. On September 1, 2026, Mercamander (09615) listed on the Hong Kong Stock Exchange as “the first stock with smart eyes and intelligence”, breaking the Hong Kong stock robot IPO record by oversubscribing by 3,835 times.
However, the contrast between 3,835 times the popularity of oversubscription before listing and the pressure on stock prices after listing ripped a layer of capital narrative. As of the close of trading on September 28, its stock price was HK$79.5, which was more than 21% off the issue price (HK$101.7).
On the one hand, there is a capital frenzy of 3,835 times oversubscribed, and on the other hand, the cold treatment of the market, which continued to decline after listing. Behind this valuation gap is a question that must be answered: Is Mercamander's PS, which was about 30 times larger when it was released, a reasonable pricing for future intelligence, or is it an overly optimistic overestimate?
Regarding this issue, we may be able to get an idea of its latest financial statements, technical performance, and market prospects.
Financial report on the AB side: impressive growth, but profit and scale effects have yet to be realized
On September 24, Mecamand handed over its first interim report after listing: revenue, orders, and overseas revenue were all high, and gross margin continued to rise — the growth curve was impressive enough. However, aside from the hustle and bustle of data, the other side is also clear: profits have been slow to arrive, and the scale effect is still an unpaid check.
In the first half of the year, the company's revenue was 237 million yuan, up 54.7% year on year; new orders were signed at 335 million yuan, up 75.3% year on year; overseas revenue was about 100 million yuan, up 68.3% year on year. The overseas market has become an important growth engine.
During the period, gross margin increased from 61.4% to 65.0%, and the profitability of the product itself continued to be optimized. Customer quality is also outstanding, with a repurchase rate of 93.9% in the first half of the year. Customers include more than 100 Fortune 500 companies, covering nearly 50 countries and regions.
Under the impressive growth curve, profit pressure is still heavy. In the first half of the year, the company's operating loss increased to 774.54 million yuan, and R&D investment of 80.1 million yuan, an increase of 72.9% over the previous year. R&D expenses accounted for 33.8% of total revenue. Looking at it over a long period of time, the company's net losses from 2023 to 2025 were 401 million yuan, 283 million yuan, and 360 million yuan respectively, with a cumulative loss of 657 million yuan over three years; the first quarter of 2026 continued to lose 57 million yuan. Although adjusted losses narrowed slightly by 5.9%, the revenue growth rate was 54.7%, and operating losses increased by 18.8% at the same time.
It is worth mentioning that this set of data created a great deal of tension with what management said “operating leverage is clearly evident.” The core logic of the scale effect is that revenue expansion drives unit fixed costs to be diluted, and losses narrow at an accelerated pace with revenue growth. Mercamander's current performance shows that the scale dividend is far from being realized.
Of course, changes in Mecamand's inventory are also worth watching: from 63.2 million yuan at the end of 2025 to 89.3 million yuan at the end of the first half of 2026, an increase of 41.4%. Although it is lower than the revenue growth rate, considering the company's high degree of product standardization, the backlog of inventory may still indicate that downstream demand is not as optimistic as order data.
Selling expenses are another sign. Zhitong Finance observed that in 2023-2025, the company's sales expenses were 186 million, 162 million yuan, and 168 million yuan respectively, which continued to exceed R&D expenses for the same period (119 million, 109 million, and 113 million). The sales expense ratio once reached 102.7% in 2023, which means that the total revenue for that year was not enough to cover sales costs. This raises a key question: is the current high growth due to natural demand driven by product value, or is it an order in exchange for continued high sales investment? Once sales investment shrinks, it is still unknown whether growth can be maintained.
From the above, it is easy to see that when Mercamander's order growth rate does not beat the rate of loss expansion, and when the announcement of “operating leverage is showing” and the reality that operating losses have increased by 18.8%, the gaps in the company's valuation and fundamentals will not automatically close.
Technology fog: A “smart first company” earns 99% of its revenue from 3D cameras
The core of Mercamander's external narrative is an “eye and hand” integrated solution to create intelligent robots, but the robot guidance business that currently contributes all revenue is essentially a mature set of 3D cameras and supporting software to serve traditional industrial robots and complete object recognition, grasping, and loading and unloading in structured scenarios.
This set of products has been iterated for more than ten years. It is a mature solution for an industrial vision circuit. It is not a embodied intelligent system for humanoid robots in an open environment.
Deconstructing Mercamand's revenue structure, the answer is more intuitive. In 2025, the company's total revenue was 389 million yuan, of which the revenue from intelligent robot guidance products was about 361 million yuan, accounting for 93%. The revenue from intelligent inspection and measurement products was about 2.37 million yuan, accounting for 5.9%. Together, the two types of business contribute approximately 99% of revenue. In the first half of 2026, revenue from intelligent robot guidance products was 216 million yuan (accounting for 91.14% of revenue), and revenue from intelligent inspection and measurement products was 21.2 million yuan (accounting for 8.9% of revenue). The total contribution also exceeded 99%.
The two products actually known as “embodying intelligence” — the Mech-GPT multi-modal large model and the mech-hand dexterous hand, account for less than 1.1% of total revenue. According to the prospectus plan, these two products are expected to be launched on a large scale at the end of 2026, and “revenue with actual commercial scale” will only be generated in the second half of 2027. In other words, a company listed as “the first stock of Embody Intelligence” comes from an industrial 3D camera that has been sold for ten years. I'm afraid this is not in the early stages of commercialization of physical intelligence, but commercialization of physical intelligence has not yet begun.
This can be seen from the technological paradox of “breakthrough progress.” Management claimed in the financial report that “physical brain technology has made a major breakthrough” and “verified the correctness and effectiveness of the investment direction.” However, R&D expenditure in the first half of 2026 was 80.1 million yuan, an increase of 72.9% over the previous year, accounting for about 33.8% of revenue.
A paradox emerged from this: if the technology actually achieved a “major breakthrough,” R&D investment should have entered a stage of marginal decline; if R&D expenses are still growing rapidly, then the technology is far from converging. R&D expenses of only 80 million in the first half of the year should simultaneously support Mech-GPT, Mech-Hand, and iteration of existing 3D vision products — whether this money is enough to achieve so-called “breakthrough progress” in “embodying the brain” is worth putting a question mark on. Meanwhile, Shao Tianlan, the founder of the company, also confessed in an interview with “Late LatePost” that the home scene and open service industry “cannot see a mature path in the short term,” and that some key technologies “are not completely clear.”
It can be seen as big. Although the solid chassis of Mercamand's industrial vision is worthy of recognition, mature industrial products are not directly equivalent to the physical intelligence that is still being explored. 99% of the company's revenue comes from mature industrial 3D cameras, mech-GPT, and dexterous smart products that have yet to contribute to real revenue. This means that Mecamand's grand story of 30 times PS has not yet been endorsed by a corresponding revenue curve.
“First in the world”, born on a narrow 1.8 billion racetrack
For investors, the aura of “number one in the world” is a very attractive label in the secondary market. As a result, Mercamander's “number one in the world” also added a strong impression to its valuation premium.
According to the prospectus, citing Insight Consulting data, the total global market for AI+3D vision-guided general intelligent robot components in 2025 is only 1.8 billion yuan. Among them, Mercamand ranked first in the world with a revenue share of 22.1%.
However, it should be noted that Mercamander's number one crown in the world is worn on an extremely small segmented track. This is the undertone of the entire valuation logic that is most easily overlooked.
Broadening the horizons to the broader market for “vision-guided robotics (VGR) systems” is much larger. According to data from Meticulous Research, the global VGR systems market is US$3.24 billion in 2026 and is expected to reach US$16.92 billion in 2036, with a CAGR of 18.1%. The VGR system includes 2D vision and 3D vision, covering hardware, software and services. Application scenarios include assembly, quality inspection, handling and placement. Mercamander's “AI+3D vision-guided general intelligent robot component” is a very small subset of the VGR market — it focuses on “3D vision-guided components” for “general intelligent robots”.
In the vision-guided robotics market, Mecamand chose the fastest growing yet smallest entry point. The 1.8 billion global market means that even if it had a 50% share, the revenue would be only 900 million. Its 389 million revenue is already close to one-fifth of this market segment.
Moreover, judging from the market growth rate, even if the forecast given by Insight Consulting is very attractive: from 2025 to 2030, the CAGR for this segment will reach 43.2%, and the market size will expand to 10.6 billion yuan in 2030, but compared horizontally with the industry market, this forecast is quite aggressive.
According to data from Interact Analysis, the CAGR of the global machine vision market was only 7.2%, and the CAGR of the Chinese machine vision market was about 10.1% during the same period. This means that the predicted growth rate of the Mecamande circuit is 6 times that of the global market and more than 4 times that of the domestic market.
This high growth forecast is also tied to a strong premise that cannot be controlled by Mecamand: large-scale commercialization of general-purpose humanoid robots in 2025-2030 will drive the penetration rate of 3D vision components to increase from 5.1% to 10.6%. Once the industrialization of humanoid robots slows down, growth expectations for the entire market segment will be drastically lowered, and the underlying logic that supports high valuations will also loosen.
After all, the big fish in a small pond is just a big fish in a small pond. This means that Mercamand is leading the way in a small racetrack, no matter how fast it grows — 389 million revenue corresponds to about one-fifth of its share, leaving room for “number one in the world”, which is visibly tight with the naked eye. The high-growth narrative that supports the imagination of valuations is also based on the suspenseful assumption of commercialization of humanoid robots, and the risks speak for themselves.
epilogue
So, is Mecamander's 30x PS overrated?
Based on the issue price, Mercamand's market value is approximately HK$12.7 billion, corresponding to 2025 revenue of $389 million, and the market sales ratio is more than 30 times. This valuation jump process is worth paying attention to. The company's valuation after the pre-IPO round was 6.367 billion yuan, corresponding to the 2025 revenue of about 16 times PS; the market value of listing and issuance was about 11.6 billion yuan, and the valuation increased by 82% in just two months. There is a huge valuation difference between the primary market and the secondary market, and the prospectus did not give sufficient explanation.
In other words, Mercamander's valuation level is a premium paid by the market for its leading global position, high margin structure, and intelligent imagination. However, Mecamand's 30x PS is already far higher than its peers. On the other hand, for mature industrial vision companies, Cognex PS is about 9.5 times, Opt 9.4 times, and Jienshi is only 6 to 7 times.
According to simple estimates, Mercamand's revenue CAGR for 2023-2025 is 46.6%. If the valuation returns to the normal state of the industrial vision industry, it will correspond to revenue of 389 million yuan in 2025. The reasonable market capitalization range is only 3.5 to 3.9 billion yuan. Compared with the current market value, there is a significant premium. Of course, long-term estimates are not directly equivalent to current estimates, but this comparison is enough to explain: 30x PS implies multiple optimistic assumptions — maintaining a high revenue growth rate over a long period of time, commercializing Mech-GPT and smart hands as scheduled to open up a new growth curve, and ultimately achieve large-scale profits. As long as any assumption falls short of expectations, there is room for revision in the valuation.
The market always has a simple perception, and a high growth rate should enjoy high valuations. However, the Mercamander case reminds us that growth cannot be priced separately from track space. Rapid growth in a small racetrack can easily reach its peak quickly. The break on the first day of listing and the current discount of the stock price is over 21%, which is probably a sign that the market is returning to rationality.