Hook
1100 yuan. That was the opening price. 629% above the IPO price of 150.8 yuan. A market cap of 444.9 billion yuan โ roughly $62 billion โ appeared on the tape in the first second of trading. For a robotics company that hasn't even filed its first post-IPO quarterly report. The bid-ask spread was a chasm. The order book screamed one thing: retail momentum, not institutional conviction. The price discovery mechanism on the STAR Market broke down, and the result was a valuation that assumes Unitree will generate more revenue in the next five years than the entire Chinese robotics industry did in the last decade. That's not a thesis. It's a prayer. And I don't trade on prayers.
Context: The Event and the Players
Unitree Robotics, a Hangzhou-based maker of quadruped and humanoid robots, listed on Shanghai's STAR Market on August 19, 2025. The IPO price was set at 150.8 yuan per share, raising approximately 24.3 billion yuan (based on 161.1 million shares outstanding). The underwriters โ a syndicate led by CITIC Securities โ priced the deal at a conservative multiple, likely to ensure a strong first-day performance. And they got it. The stock opened at 1100 yuan, giving early investors a windfall. Among them: Astrend IV, a vehicle controlled by Shunwei Capital (Lei Jun's venture arm), which held 16.106 million shares. At the opening price, that stake was worth 17.7 billion yuan. The cost basis? Around 56.4 yuan per share, based on the IPO price and the disclosed unrealized profit of 15.2 billion yuan. That's a 19x return on paper.
But the article you read โ the one that gushed about the 152 billion yuan paper profit โ buried the critical detail: lock-up periods. Early investors cannot sell immediately. Those 16.1 million shares are locked for at least 12 months, and often longer for controlling shareholders. The '152 billion yuan' is a number on a spreadsheet, not cash in hand. The market doesn't care about your unrealized gains until they become realized. And the path to realization is a narrow door: sustained price above 300 yuan for at least 30 consecutive trading days, or a negotiated block trade. Neither is guaranteed.
This is the essence of the story. A robotics company with legitimate hardware chops โ Unitree has shipped thousands of quadruped robots (Go2, B2) and launched humanoid prototypes (H1, G1) โ is now valued at a premium that would make even Nvidia blush. The question is not whether Unitree is a good company. The question is whether the market is pricing it as a good company or as a narrative.
Core: Order Flow, Valuation, and the Hidden Leverage
Let's strip the narrative and look at the order flow. The opening trade was a massive imbalance. Retail investors โ the 'scattered households' of Chinese equity markets โ piled in, driven by the 'Hangzhou Six Little Dragons' hype and the Lei Jun aura. Institutions, who had access to the IPO at 150.8 yuan, were net sellers on the first day. The volume was extraordinary: over 80 million shares traded in the first hour, representing nearly 50% of the float. That's not accumulation. That's distribution.
Now, the valuation. 444.9 billion yuan. For context, the entire global robotics market (including industrial, service, and humanoid) was estimated at $45 billion in 2024, with a CAGR of 15%. At that rate, the market reaches $90 billion by 2030. Unitree's current market cap implies it will capture 70% of that market โ an absurdly optimistic assumption. Even Tesla, with its massive manufacturing scale and AI ecosystem, is valued at a fraction of that for its robotics division. The reality is that Unitree's 2024 revenue was likely under 2 billion yuan (based on its quadruped sales and early humanoid pre-orders). A 444.9 billion yuan market cap translates to a price-to-sales ratio of over 200x. That's not a growth stock. That's a lottery ticket.
And the lottery ticket is backed by leverage. Not financial leverage โ I'm talking about the implicit leverage of future expectations. The market is betting that Unitree will achieve a 50% CAGR in revenue for the next decade, with margins expanding from low single digits to 30%+. That's a bet on a technology inflection that hasn't happened yet. The humanoid robot market is still in the lab stage. Figure AI has shipped fewer than 100 units. Tesla's Optimus is still in development. The industry is years away from mass production. Unitree's G1, priced at 99,000 yuan, is a marvel of cost engineering, but it's not a revenue driver yet. The order book for humanoid robots? Mostly pre-orders from research labs and corporate pilot programs. Not recurring revenue.
This is where the 't measured yet.' signature comes in. The market is pricing Unitree as if it has already solved the scaling problem. But the data โ the actual order flow, the revenue figures, the gross margins โ t measured yet. The only thing that has been measured is the price action. And price action, divorced from fundamentals, is a dangerous signal.

Contrarian: What Retail Sees vs. What Smart Money Knows
Retail sees the 629% pop and thinks, 'I missed the boat.' Smart money sees the distribution and thinks, 'I need to find the exit.' The contrarian angle here is that the IPO's success is actually a bearish signal for the broader robotics sector. When a company with no proven profitability or massive revenue base commands a 200x P/S ratio, it pulls all valuations upward. That means every other robotics startup โ from Fourier Intelligence to Zhiyuan Robotics โ will now demand higher valuations in their next funding rounds. That's fine for venture capitalists, but it creates a bubble that will eventually burst.
Look at the history. In 2021, when the Chinese AI chip company Cambricon listed on the STAR Market, it opened at a 400% premium and later collapsed 80% as it failed to deliver on revenue promises. Unitree faces the same risk. The narrative is compelling: China's hardware advantage + AI = the next Tesla. But the reality is that Unitree's core competency is in motion control and mechanical design, not in AI. Its humanoid robots rely on third-party AI models for perception and planning. The 'true' AI robotics companies โ like Figure with OpenAI, or Tesla with its end-to-end neural network โ have a different moat. Unitree's moat is manufacturing cost, which is replicable. The Chinese supply chain can produce humanoid robots at 99,000 yuan. So can Xiaomi, and DJI, and dozens of others. The competitive advantage is thin.
And the market hasn't priced in the downside risk. What if the US or EU imposes tariffs on Chinese robotics? Unitree's overseas sales, which account for a significant portion of its quadruped revenue, would be hit. What if the government tightens regulations on humanoid robots for privacy concerns? The stock would gap down. The market doesn't care about your thesis; it cares about liquidity. And when liquidity dries up, the 629% gain becomes a 90% loss.
Takeaway: Actionable Price Levels and the Only Thing That Matters
The IPO price of 150.8 yuan is now a distant memory. The opening price of 1100 yuan is the new reference point. For traders, the key levels are: support at 800 yuan (the 50% retracement of the opening day's range) and resistance at 1200 yuan (the psychological round number). If the stock breaks below 800 yuan, the next stop is 600 yuan, which would still be a 300% gain from the IPO but a 45% loss from the open. That's where the retail euphoria will meet the smart money's exit.
For long-term investors, the only thing that matters is the first earnings report. If Unitree reports revenue growth of 100%+ and provides a forward guidance that justifies the valuation, the stock might hold. But if it reports a miss โ and the whisper numbers are already inflated โ the correction will be violent. The 152 billion yuan paper profit will evaporate faster than it appeared.
Here's the bottom line: 't measured yet.' The market has assigned a price, but it hasn't assigned a value. The distinction is everything. The 629% pop is a liquidity event, not a validation of the business model. The real test โ the one that separates smart money from the crowd โ is whether you can hold through the drawdown or exit before it. I know my answer. I'm not holding anything that hasn't proven its revenue stream. And Unitree's revenue steam? T measured yet. The only thing that has been measured is the volume of trades. And that's not a fundamental.

So, to the retail traders chasing the hype: check the gas, not just the gem. The gas is the order flow, the lock-up expiry dates, the short interest. The gem is a fantasy until the P&L proves it. The market doesn't reward narratives. It rewards execution. And Unitree's execution is still in the lab. The IPO is the first chapter of a long book. The next chapter โ the one that matters โ hasn't been written yet.