The IPO of Yushu Technology on the Shanghai STAR Market closed with a lottery rate of 0.0181%. That is the lowest in the board’s history. The hype machine is running at full capacity. Media outlets estimate a single lot profit of 200,000 to 300,000 RMB. Strategic investors include the National Social Security Fund, DeepSeek, PetroChina Kunlun Capital, and Tencent affiliates. The company is being crowned the "first humanoid robot stock" in A-shares. But the ledger does not lie, and the narratives are piling up faster than the data.
Source code is the only truth that compiles. For a public company, the source code is its financial statements, its product breakdown, and its unit economics. Yushu Technology has provided none of those. The prospectus—if one can call it that—offers shipment numbers, market share estimates, and a list of strategic backers. It does not disclose revenue, gross margin, net income, or operating cash flow. The valuation of 60.993 billion yuan (approximately $8.5 billion at current rates) is built on a foundation of promises, not audited ledgers.
Context: Yushu Technology is a robotics company based in Hangzhou. It started with quadruped robots (similar to Boston Dynamics’ Spot) and later expanded into bipedal humanoid robots. In the first half of 2026, it shipped 5,900 units combined, claiming a 31% global market share in the quadruped segment. The company asserts that 90% of its core components are developed and manufactured in-house. Its strategic investors include DeepSeek, the AI lab that is positioning itself as China’s answer to OpenAI. The IPO raised 6.099 billion yuan, with the stock priced at 150.80 yuan per share. The post-IPO market cap was 60.993 billion yuan.
Core: The data gap is a confession.
The first problem is the unit economics. If we assume an average selling price of 100,000 to 300,000 yuan per robot (a reasonable range for industrial and educational robots), the first-half revenue would be between 590 million and 1.77 billion yuan. Even at the high end, the market cap of 60.993 billion yuan implies a price-to-sales ratio of 34x to 100x. That is not a hardware company multiple. That is a software-as-a-service multiple, or a biotech pre-revenue multiple. The market is pricing in exponential growth that has not yet appeared in any shipment data.
Second, the 90% self-sourced component claim is ambiguous. Based on my experience auditing hardware supply chains, that percentage is almost certainly measured by the number of component types, not by bill-of-materials cost. The high-value items—chips, LiDAR, high-precision sensors—are likely still purchased externally. The company’s own annual report, if it exists, would clarify this. But the silence in the data is a confession.
Third, the humanoid robot narrative is a tail that wags the dog. The 5,900 units shipped are overwhelmingly quadruped robots. The humanoid model, likely the H1, has not been disclosed in terms of volume. The market is labeling Yushu as a "humanoid robot stock" based on a few prototypes and a strategic partnership with DeepSeek. That partnership is a press release, not a product roadmap. There is no evidence that DeepSeek’s large language model has been integrated into Yushu’s robot control stack. The collaboration may be real, but it is not yet reflected in any shipped unit.
Fourth, the strategic investors have lock-up periods. The Societal Security Fund, DeepSeek, and the energy giants are locked for 12 to 36 months. The free float in the early days will be tiny. That means the price can be manipulated by a small amount of capital. The hype around the 0.0181% lottery rate is a self-fulfilling prophecy: everyone wants to get in for the first-day pop, but few plan to hold for the long term. The early investors who bought in at a 2016 valuation of 16.85 billion yuan (a 840x return) are not holding for the next decade. They are selling into the public market.
Contrarian: What the bulls got right.
To be fair, the bulls have some valid points. The policy environment is extremely favorable. The STAR Market approved Yushu in 73 days, a record. That is a clear signal that the Chinese government wants to support "new quality productive forces" and hard-tech companies. The strategic investor lineup is not just for show. The involvement of energy giants like PetroChina and Southern Power Grid suggests that Yushu’s robots are being deployed for dangerous environment inspection and power grid maintenance. Those are real, paying use cases. The quadruped robot business is profitable enough to generate cash flow, even if the margins are thin.
Moreover, the company’s vertical integration is a genuine moat. By controlling 90% of component types, Yushu can drive down costs faster than competitors who rely on off-the-shelf motors and reducers. In a price war, that advantage compounds. The partnership with DeepSeek, if it matures, could give Yushu an exclusive high-intelligence embodied model, similar to the OpenAI-Figure pairing. The internet giants—Meituan, Tencent, Alibaba, Ant Group—are not just investors; they are potential customers for last-mile delivery and service robots. The ecosystem effect is real.
But these strengths do not justify the current valuation. The market is pricing in a future where Yushu captures 50% of the humanoid robot market by 2030, with margins comparable to software platforms. That is a bold assumption. The gap between the promise and the proof is still wide.
Takeaway: The audit is just beginning.
The IPO is a landmark for the embodied AI industry. It provides the first public valuation anchor for humanoid robot companies. It will force other players to accelerate their own listings. But for the investor who buys at the opening bell, the risk is asymmetric. The upside is limited by the lock-up expiration and the eventual dilution from follow-on offerings. The downside is a 50% correction if the first quarterly report shows that humanoid robot revenue is negligible.
History is written by the auditors, not the poets. The numbers will come. The quarterly filings will reveal the truth. Until then, treat the 0.0181% lottery rate as a warning, not a signal. The ledger does not lie, but the narrative does. And the narrative around Yushu Technology is still waiting for its first real data point.