Unitree Robotics' P/E Ratio Reaches Nearly 10x Industry Average as Humanoid Robot Funding Enthusiasm Cools
nashnova research
Unitree Robotics has fallen 53.2% from its first-day peak two months after listing, yet still trades at a trailing P/E of 388× — nearly 10 times the sector average; meanwhile, Q3 embodied-AI funding dropped roughly 29% quarter-on-quarter, signaling the humanoid-robot sector is shifting from hype to commercialization reality-check.
What happened to Unitree's stock?
Unitree (宇树科技) listed on Shanghai's STAR Market on August 19 at ¥150.8 per share and hit ¥1,100 intraday on day one.
By the October 21 close, the stock had fallen 53.2% from that intraday peak.
This means → the market completed a full "euphoria-to-repricing" cycle in two months; the IPO-day premium is being squeezed out.
What does a 388× P/E ratio actually mean?
On October 20 the company disclosed a trailing P/E of 388.25×, versus a one-month sector average of just 39.51× for general-equipment manufacturers.
In plain terms = for every ¥1 of earnings, the average peer commands about ¥40 in market price; Unitree commands nearly ¥390 — roughly 10 times more expensive.
Unitree itself flagged the risk: humanoid robotics remains in early commercialization, with uncertainty across technology maturity, downstream demand, and supply-chain readiness — revenue and profit will take a hit if large-scale deployment lags expectations.
What signal is the regulator sending?
Reuters reported that Chinese regulators have used informal window guidance to slow IPO reviews for some humanoid-robot companies.
The focus: whether valuations are justified and whether revenue from affiliated projects genuinely reflects market demand.
This means → no formal IPO freeze is in place, but the bar for "concept-rich, revenue-light" listings is rising.
How sharp is the funding cool-down?
Per Yicai Global, Q3 embodied-AI deals totaled at least 185 rounds raising ¥33.3 billion — down about 28.7% from the H1 quarterly average of ¥46.75 billion.
Monthly swings were sharper: July ¥8.65 bn → August ¥17.54 bn → September plunged to ¥7.15 bn, a month-on-month drop of roughly 59.2%.
Top-tier VCs pulled back in step — HongShan (Sequoia China) participated in at least 20 deals in H1 but only 8 in Q3; GL Ventures also cut its deal count.
How does physical-AI commercialization differ from software AI?
Generative AI (products like ChatGPT) ships mainly through software and cloud platforms — low marginal cost, fast iteration.
Physical AI (humanoid robots) must also clear hardware cost, motion control, environmental perception, reliability, and real-world deployment hurdles.
In plain terms = software AI can monetize the moment it generates a paragraph; a robot has to work reliably on a real factory floor to count — the road from "tech demo" to "volume shipment" is far longer.
What is the core question behind this repricing?
The market's central question: how many years of expected growth is the current price actually paying for?
And when will that growth show up as real revenue and profit?
This reflects a broader pivot across the humanoid-robot sector — from "paying for the vision" to "paying for results" — and the answer will determine whether valuations can hold at elevated levels.
市场有风险，内容仅供研究参考，不构成投资建议。