Unitree's Robot Stock Lost Half Its Value Because the Robots Aren't Selling

Unitree Robotics has lost more than half its market value in three weeks, and the company's own numbers explain why: almost none of its humanoid robot revenue comes from actual factory work.

Unitree Robotics listed on Shanghai's STAR Market on August 19, 2026, and for a few hours it looked unstoppable. The Hangzhou robot maker priced its IPO at 150.80 yuan a share and opened at 1,100 yuan, a 629% jump, according to Bloomberg and the South China Morning Post. The first trade valued the company at about 445 billion yuan, or $66 billion. Retail demand was just as wild: The Asset reported that the online tranche was oversubscribed more than 8,200 times.

That was August 19. By September 9, the party was over.

Unitree shares closed at 513.93 yuan on September 9, according to FT market data and Investing.com, more than 53% below the debut high. Its market value had fallen to roughly 208 billion to 212 billion yuan, depending on the data provider, wiping out more than 230 billion yuan from that first-day peak. This wasn't a slow bleed. It was a reversal almost as violent as the surge that created it.

The factory story is thinner

The answer sits inside Unitree's own IPO materials. In the first three quarters of 2025, 73.6% of the company's humanoid robot revenue came from research and education customers: universities, labs and developers buying robots to study, not to put to work. Commercial and consumer uses made up another 17.39%. Industrial applications, the use case every humanoid robot pitch deck wants you to picture, accounted for just 9.01%.

Break that last slice down and the story gets smaller. Unitree's filing says corporate guided tours made up roughly half to 70% of industrial application revenue. That means robots walking visitors around showrooms may have been a bigger industrial use than robots doing actual industrial work. What's left for smart manufacturing and inspection is closer to 3% or 4% of total humanoid revenue.

That is the gap.

Then there's the profit line. Unitree's first-quarter 2026 revenue rose 69% to 422.8 million yuan, but net profit after deducting non-recurring gains and losses fell 53% to 40.3 million yuan, according to Yicai Global and a South China Morning Post report on the company's updated filing. That's a real company selling real hardware. It is not a company whose earnings clearly justified a $66 billion opening valuation.

Frankly, the market did the arithmetic and didn't like the answer.

Unitree isn't alone in getting caught out here. The entire humanoid robotics trade in China has run on a bet that these machines are close to real industrial deployment at scale, replacing line workers the way robotic arms did a generation ago. Unitree shipped more than 5,500 humanoids in 2025, genuinely strong volume for the category. But volume and industrial revenue are different things, and Unitree's own disclosures show the distance between them is still large.

Beijing noticed

Regulators are now moving around the same problem investors just repriced. Reuters reported on September 9 that China's securities regulator had given informal window guidance to some investment banks and companies, citing The Information, after Unitree's volatile debut. The China Securities Regulatory Commission is lifting the bar for humanoid robotics IPO candidates by asking for recurring revenue, a path to narrower losses or real innovation before approvals move forward, according to that report. Reuters also said it couldn't independently verify The Information's account and that Chinese financial regulators didn't immediately respond to a request for comment.

That's a real shift. A month ago, Unitree's IPO was being treated as proof that China had produced the first genuinely investable humanoid robot maker on a mainland exchange. Now it's a warning sign, and it's regulators, not just bruised retail investors, drawing the lesson.

None of this makes humanoid robots a dead end. Unitree still builds machines that customers pay for, and a 53% profit decline is a bad quarter, not a bankruptcy filing. The issue is price. Investors briefly valued a company with about 40 million yuan of first-quarter adjusted profit as if factory adoption had already arrived. Unitree's filing says it hasn't.

The business is real. The valuation got ahead of it.

Also read: SDK.finance Adds a Standalone Ledger Layer for Banks and FinTechs • LinkedIn wants your government ID, but its own leak record argues against it • OpenAI Puts AI Doomer Paul Christiano on Its Safety Board

This article is posted in Business News, check it out for more related stories.