The stock opened at 1,100 yuan on August 19. By Wednesday's close, it was 513.93 yuan — a 53% drop from the debut high, wiping out roughly 237 billion yuan in market capitalization from the peak.
Last week, the signal was Unitree down 44% from its opening price. This week, the halving is confirmed at more than 50%, with roughly $30 billion in paper value erased in 11 days. The Shanghai STAR Market has delivered its verdict on the first publicly traded humanoid robot company.
The numbers are brutal in their specificity. Unitree priced its IPO at 150.80 yuan, opened at 1,100, and closed day one at 845 — a 460% gain. The allocation rate for retail investors was 0.018%, the lowest in STAR Market history. Only 7.44% of shares were tradable in early days. Investors joked that winning shares was like hitting the lottery.
The allocation rate for retail investors was 0.018%, the lowest in STAR Market history. Only 7.44% of shares were tradable in early days.
Then the lottery ticket lost half its value in three weeks.
The sell-off has a logic to it, and the logic is in Unitree's own prospectus. In the first nine months of 2025, 73.6% of humanoid robot revenue came from universities and research institutions. Only 9.01% came from industrial applications. More than half of that industrial slice was for showroom tours and demonstrations.
The robots dance. They backflip. They fight in kickboxing quarterfinals at the World Humanoid Robot Games. They do not yet work in factories at scale.
Founder Wang Xingxing said as much himself, the day after the listing. Speaking at the World Robot Conference in Beijing, he acknowledged that robots remain less efficient than humans at simple assembly work and lack the ability to adapt to new tasks without retraining. The industry's "ChatGPT moment," he said, would arrive "in two to three years at the earliest, or five to 10 years at the latest." That was a more cautious timeline than he offered a year ago. Unitree's shares fell 16% that day.
When the founder tells you the product isn't ready for large-scale factory deployment, the market should listen.
The valuation math was always going to be the problem. Unitree reported 1.699 billion yuan in 2025 revenue, with humanoid shipments topping 5,500 units. Even after the September drop, the company was still worth around $30 billion — roughly 125 times 2025 revenue. At its peak of $66 billion, it traded at more than 250 times revenue.
Compare that to Agility Robotics, which recently opened its books as part of a proposed SPAC. Agility generated $1.78 million in net sales in 2025, recorded a $140.2 million operating loss, and is valued at approximately $2.5 billion — roughly 1,400 times its 2025 net sales. Unitree is profitable. Agility is not. The market is now trying to figure out what that distinction is worth.
Figure AI, the best-funded American humanoid rival, announced in September 2025 that it had raised more than $1 billion at a $39 billion post-money valuation. Figure is private. Its valuation does not get marked down on a screen every trading day. Unitree no longer has that luxury.
The regulatory backdrop makes the sell-off more than a simple repricing. Beijing had already warned the sector before Unitree listed. In November 2025, National Development and Reform Commission spokesperson Li Chao said more than 150 Chinese companies were working on humanoid robots and cautioned against "highly repetitive" products crowding the market.
Since the sell-off, the response has sharpened. Reuters carried The Information's report that the China Securities Regulatory Commission has given informal guidance to banks and companies, asking future humanoid listing candidates to show recurring revenue, a path to narrower losses, or real technical innovation before approvals are considered. Reuters could not independently verify the report, and Chinese regulators have not commented. Keep that caveat. It matters.
The CSRC's reported guidance frames what comes next. The first Unitree submission was the IPO itself. The second submission — the one that matters now — is whether the company can demonstrate that humanoid robots generate repeat revenue from customers who buy again after the first sale.
Dong Chen, chief investment officer for Asia at Bank J. Safra Sarasin, put the valuation problem plainly to the South China Morning Post: "Given that level of uncertainty, Unitree's high valuation is probably not justified."
The robots are impressive. The valuation was more impressive than the robots.
None of this means China is backing away from robotics. The opposite is true. The sector sits inside Beijing's broader push around embodied intelligence and advanced manufacturing. Unitree raised about 6.1 billion yuan in the IPO, above its earlier target. That cash did not disappear when the share price fell.
The company is still shipping. On the same day the stock hit its lows, Unitree announced the Dex5-S dexterous hand: 22 degrees of freedom, 1:1 real-hand size, starting at $6.5K. All 22 joints support smooth backdrivability, with limit impact torque protection on each. It's a component play, not a factory deployment, but it's real product.
The tension is whether component sales and research orders can sustain a company valued at 125 times revenue while the industrial market develops. Wang Xingxing holds about 121.4 million shares, worth roughly 103 billion yuan at the first-day close. Meituan, an early backer, had an 8.7% post-IPO stake worth nearly 30 billion yuan at the same close — a return of about 70 times its original investment. The paper wealth was created quickly. It has been marked down quickly.
The question for public markets is simpler than the question for technologists. Technologists ask whether humanoid robots will eventually work in factories. Public markets ask when, and how much of that future is already priced in today.
Unitree's first-half 2026 revenue came in at roughly 1.15 billion yuan, up 48.5% year-over-year — ahead of guidance, but a sharp deceleration from 2025's 332% growth. Adjusted profit fell about 19%. The growth is real. The margin expansion has stalled.
That combination — real revenue, decelerating growth, thin industrial adoption — is what a $30 billion valuation now has to justify. The alternative is that the market continues to mark it down until the industrial numbers arrive.
Agility's S-4 filing said its Digit robots have been deployed at 9 customer sites and accumulated more than 65,000 operating hours, with more than $300 million in multi-year contracted orders for Digit v5. Those are the kind of numbers that would change Unitree's story. Unitree does not have them yet.
What Unitree has is 5,500 humanoid shipments, a profitable core business, and a founder who told the market the ChatGPT moment is years away. The public market heard him.