Speed runs require foresight, not just reaction.
On August 19, 2025, Unitree Robotics, the Chinese quadrupedal and humanoid robot maker, made its debut on the STAR Market (Shanghai’s Nasdaq-like tech board) with a jaw-dropping 629% opening pop. Priced at 150.8 RMB per share, it opened at 1,100 RMB, pushing its market cap to 444.9 billion RMB (approximately $62 billion). The headline-grabbing number: Astrend IV, a Shunwei Capital (Lei Jun’s VC arm) vehicle, saw a paper gain of 15.2 billion RMB ($2.1 billion) on its early-stage stake.
Context: Why now?
Unitree is not just another robot company—it’s the first breakout IPO from the so-called "Hangzhou Six Dragons," a cohort of AI-hardware darlings backed by local government subsidies and a narrative shift from software AI to embodied intelligence. The market is pricing a future where humanoid robots become the next iPhone, and Unitree, with its mass-produced quadrupedals (Go2, B2) and cheap humanoid prototypes (G1 starting at $13,000), is the closest China has to a Tesla Bot challenger. But this IPO is also a liquidity event for the early crypto-friendly capital that has been quietly rotating into hard tech. Shunwei’s 152 billion RMB paper gain is a signal: the same capital that fueled the 2017 ICO frenzy and 2020 DeFi summer is now betting on physical AI.
Core: The numbers tell a story of discount and euphoria
Let’s unpack the math. Astrend IV held 16.106 million shares before the IPO. At the opening price of 1,100 RMB, that stake is worth 17.7 billion RMB. The 15.2 billion RMB paper gain implies an average entry cost of roughly 56.4 RMB per share—a 63% discount to the IPO price. That’s typical for early-stage VCs, but the magnitude of the pop (629%) means the IPO was deliberately underpriced to create a "first-day winner" effect. This is a classic Chinese IPO tactic: leave money on the table for retail investors, generating positive sentiment for subsequent offerings.
But the real kicker is the valuation. At 444.9 billion RMB, Unitree is trading at a price-to-sales ratio that would require annual revenue of 55–90 billion RMB (assuming 5–8x PS). Yet Unitree’s 2024 revenue is estimated at under 2 billion RMB. Even with 100% annual growth, it would take 5–8 years to justify the current market cap. The market is effectively paying a massive call option on the humanoid robot thesis—one that may or may not materialize by 2028.
From the noise of 2017 to the signal of today. The 2017 ICO mania was about pure speculation; today’s premium is anchored by a real product, real revenue, and a clear path to mass production. But the risk is the same: expectations can overshoot reality.
Contrarian: The blind spot few are talking about
Most coverage focuses on the wealth effect and the "star capital" narrative. But the untold story is the structural risk in the balance sheet. Unitree’s competitive advantage is in hardware integration and cost control—not AI core competence. Its humanoid robots use relatively basic motion control algorithms; they lack the large-model-driven end-to-end reasoning that players like Figure AI (backed by OpenAI) or Tesla’s Optimus are building. If humanoid intelligence becomes a software game—where the brain matters more than the body—Unitree could be relegated to a low-margin hardware assembler, competing with dozens of Chinese copycats on price.
Furthermore, the 152 billion RMB paper gain is locked up for 1–3 years under China’s IPO lock-up rules. That’s not real cash; it’s a phantom fortune that could vanish if the stock corrects. History shows that STAR Market IPOs often drop 30–50% within six months after the initial pop. The risk is asymmetric: the upside is capped by the lock-up, but the downside is open.
The ledger does not lie, but it rewards patience. The real alpha here is not in chasing the stock—it’s in understanding the capital rotation. Shunwei’s early exit from crypto (via its 2018–2020 fund) and re-entry into hard tech mirrors a broader trend: crypto-native VCs are moving into AI robotics, DePIN, and compute. Unitree’s IPO is a validation of that thesis, but it also signals that the easy money in crypto has already been made.
Takeaway: What to watch next
Unitree’s stock will be a sentiment proxy for the entire humanoid robot sector. If it holds above 800 RMB in the first month, expect a wave of follow-on IPOs from Chinese robotics startups (Zhiyuan, Fourier, Galaxy General). If it crashes, it will dampen the entire sector’s valuation. But the bigger signal is for crypto: the same capital that once chased NFTs and DeFi is now pouring into physical AI. The next bull run in crypto will likely be driven by AI–DePIN convergence, not memes. Watch for tokens like Render (RNDR), Akash (AKT), and iExec (RLC) as proxies for this trend.
Speed runs require foresight, not just reaction. The question is: are you positioned for the next cycle, or are you still chasing the last one?
