Behind the human-like robot bubble: How policies, local governments and capital markets work together to create high valuations

China’s humanoid robotics industry is experiencing a typical acceleration process of “policy pre-determination, local racing, capital re-pricing.”China's regulators begin to slow down the pace of some human-formed robotics companies listingThe focus is on reviewing valuations, revenue quality, and whether local government support projects are packed into sustainable business needs. It is really worth asking not only “is the capital market not too crazy,” but an unrivaled industry, why can it be pushed to such a high valuation position in a short time.

High valuation is not grown out of emptiness, it is often spread along the chain of policy.

Human-formed robots and “personal intelligence” have become a popular direction in central and local industrial policy. Local governments build industrial parks, training centers, application demonstration bases, state equity and industrial funds into equity financing, research institutions and local projects form the first batch of orders, and capital markets transform these policy signals into growth expectations.

The biggest problem with this mechanism is that policy support, real demand and corporate profits are easily mixed together. a demonstration project purchased by a local government can become corporate income; a round of state-owned investment can raise valuation; an industrial park signature can be packaged as an order reserve; once these numbers enter the bidding book and market narrative, what investors may see is not a "policy trial", but a "business model has been established".

The rise and withdrawal of cosmetic technology, exposed is not the emotion of a company, but a whole set of pricing logic.

August 19th, Shandong Science and Technology landed in Shanghai.The first day, the share price rose six times. Previously, its IPO retail section received more than 8,000-fold approvals, valued at 219 times the profit in 2025 and 36 times the sales.

Such sharp price changes have demonstrated that the market has not found a stable way to measure how much a human-formed robot business is worth.The question is not whether the tree has technology – it is really competitive in movement control, hardware integration and cost control – but whether these technologies can be converted into a replicable, sustainable, policy-free cash flow in a short enough time.

If a company generates revenue primarily from research institutions, local demonstration projects, training data collection and policy procurement, the quality of these revenue is completely different from the quality of stable factory orders.Both appear in financial statements but cannot be given the same long-term value.

Why are local governments willing to survive? because industrial performance and fiscal pressure exist at the same time

Chinese local governments have long undergone the pressure of hiring investment, industrial upgrading and job growth.A new industry clearly encouraged by central policies tends to quickly become a local competitor: who built the park first, who introduced the leader first, who first set up the fund, who first got the label "National Leader".

This competition can accelerate the formation of industries, and can also cause rebuilding and over-investment. New energy vehicles, photovoltaics, chips and other industries have emerged in many places at the same time investing in large amounts of capital, subsidizing enterprises and expanding production capacity.

A industry that has not yet verified a business model, if first massively pushed by local fiscal and industrial funds and then priced by the capital market, the cost of failure will not only belong to the shareholders of enterprises, but may also spread to the public finances through state capital losses, park vacancies, local debt and industrial fund losses.

The so-called "national strategy", can not automatically become a proof of investment value

This is where this round of regulatory tightening is most notable. Beijing has not given up human-like robots, but instead, the industry is still within the scope of policy support.But regulators have begun to distinguish between two things: an industry worth the state’s investment, not that every enterprise is worth high valuation; a technology direction is strategic, not that today’s business model has matured.

This border has often been blurred by capital markets in the past.As soon as a certain industry enters policy documents, investors can easily understand "state support" as "future guaranteed", and local governments turn policy direction into indicators, and enterprises turn these policy resources into financial stories.

If the securities market cannot separate policy orders, commercial orders, research orders and one-off projects, valuation will be increased by administrative resources.

There is a huge gap between the market size of 7,000 units and hundreds of times the market profitability

International Federation of Robotics statisticsGlobally, around 7,000 humanoid robots will be sold in the industry and professional service scenario in 2025.In comparison, conventional industrial robots will be installed worldwide at 54,2 million units in 2024 and professional service robots will be sold around 19,9 million units.

This set of data shows that human-like robots are still a small market. It can have a lot of technical imagination, but the current business size is not enough to naturally support all high valuations. If the capital markets convert the industry scale in 2030 and beyond into today’s prices in advance, any technical delays, cost issues or insufficient demand will result in a sharp retreat.

More dangerously, the cost of failure can be socialized and the profits privatized.

When an enterprise’s valuation rises, founders, early investors, and some institutions can make huge gains; when a project fails, losses can fall to local industrial funds, state-owned platforms, secondary market investors, and financial support projects.

If local governments use public funds to support a high-risk technology industry, they should publicly disclose the size of investments, withdrawal mechanisms, performance criteria, and failure conditions.

Similarly, securities regulation should not only question the quality of revenue after the stock price has risen.More importantly, before the listing, make a clear distinction: which revenue comes from market customers, which comes from government procurement, which comes from affiliated industry funds, and which is just a one-time demonstration project.

What really matters is not whether the robot will play the fight, but who is paying for it.

Games, conferences, and short videos can demonstrate the ability of robots to act, but can’t demonstrate business models. Really meaningful data for the next year should include: repeated customer percentage, the length of effective work per unit, revenue after government subsidies and demonstration projects, changes in gross interest rates, and the proportion of orders from non-government customers.

Only when these indicators continue to improve can the human-formed robotics industry truly turn from a policy industry to a market industry.

So this round of IPO warming isn’t just a “rational return” to capital markets, but it’s exposing a deeper problem: when national strategies, local governments, industrial funds and securities markets simultaneously force in one direction, prices are easy to run ahead of technology and demand.