Hard Tech's Big Squeeze: Primary Market Frenzy Meets Secondary Market Doubt
A group of partners recently flew to Qingdao for a meeting about a world-model project they had been following for a long time.
Last year the work was punishing: extremely high technical barriers, no near-term monetization, and not even a coherent business story to pitch. This year it has become the most desirable theme in capital markets, raising well over a billion yuan in six months, with the next round priced near several tens of billions of yuan — a stake that may not even be available.
That is the surreal reality of today's hard-tech investment market.
Consider the data: in the first half of 2026, China's equity investment market completed 5,944 deals, up 14.7% year on year, worth about 565.4 billion yuan, up 31.9%.
Where did the money go? Industry estimates put 90% of this year's primary-market capital into AI, robotics, world models, quantum technology, controlled nuclear fusion, integrated circuits and commercial space. Nine of every ten yuan chase the same handful of labels — a concentration with little precedent.
The results are visible: a brain-computer interface project tripled its valuation in two months; some startups reprice within a week and may run three funding rounds simultaneously; founders have grown assertive, refusing valuation adjustment or repurchase clauses, and some will not even cooperate with due diligence. Invest or not? Someone else will. FOMO is driving every decision.
Yet while the primary market fights over allocations, the secondary market votes with its feet.
On August 19, Unitree listed on Shanghai's STAR Market at an issue price of 150.80 yuan, opened at 1,100 yuan — a 629% gain — and briefly topped 440 billion yuan in market value, before falling well below its first-day high.
On September 1, Mech-Mind listed in Hong Kong. Its public offer was oversubscribed 3,835 times, yet it closed its first day below the issue price. Two markets, two moods: the primary market's frenzy has not transferred to the secondary market, and the gap between story and earnings is plain to see.
The logic is straightforward. Primary investors bet that someone will take the next round; secondary investors look at real revenue and profit. If a company cannot show recurring revenue, a narrowing loss or genuine technological innovation, why should the secondary market pay for primary-market valuations?
Policy signals deserve closer attention.
Twelve days ago, the National Development and Reform Commission sent the robotics industry two messages: keep building training grounds, real-machine data and embodied models, but meet clear conditions — robots must enter real scenarios and form application loops around real demand, and regions must avoid blindly following trends and rushing in.
This is not isolated. In November 2025 the NDRC noted that China already had more than 150 humanoid robot companies, over half of them startups or cross-industry entrants, and spoke of duplicate products, clustered firms, and industry entry and exit mechanisms. When capital floods in on labels rather than evidence, regulation tends to follow.
With more than 150 companies in the race, building a robot is no longer scarce.