Summary in Plain Language
The global humanoid robotics industry has long been in a state of hype and speculation: there isn't a single publicly traded company whose main business is the production of humanoid robots that can provide actual orders and profit records. The primary market has invested hundreds of billions, and many companies have seen their valuations soar to tens of billions despite selling only a few prototypes. Opinions about the industry are polarized—some claim the future market size could be ten times that of the new energy vehicle market, while others see it as a flashy, yet useless toy. This all changed when Yushu Technology went public, and its stock price plummeted by nearly half in just five days from a high of 440 billion yuan, shattering the illusion of the industry's potential. Suddenly, everyone realized that the previous valuations were based on unfounded assumptions, and the industry is now forced to shift from focusing on hype to establishing realistic pricing standards.
---
Detailed Analysis
1. Why Did Yushu Technology’s Public Offering Tear Down the Industry’s Valuation Bubble?
The valuation logic for humanoid robots was simple: rarity meant high value. For the past two years, the secondary market only had OneBot as a target for humanoid robots, and despite losing over 5 billion yuan in six years, it still received a high premium because there was no other option for investors in this field.
However, with Yushu Technology going public and at least 20 more similar companies preparing for IPOs, the concept of scarcity became meaningless. JPMorgan Chase immediately cut OneBot’s target price in half, leading to a significant drop in its stock price. More importantly, the valuations of publicly traded companies can no longer be sustained by empty promises; they must be based on practical metrics such as how many robots can be sold in a year and how much profit can be made from each sale. The huge valuations that emerged from speculation will eventually be adjusted to reflect these realities.
The valuations given by three leading securities firms for Yushu Technology varied by a factor of three: CITIC Securities estimated 50 billion yuan, CCB International 109 billion yuan, and Nomura Securities nearly 150 billion yuan. None of them dared to return to the previous high of 440 billion yuan—this was simply a bubble that has now burst, and everyone is trying to find a reasonable price point.
2. Understanding the Commercialization Threshold for Humanoid Robots
All complex research reports can be simplified to a basic calculation that ordinary people can understand: if a boss spends 80,000 yuan per year on a worker and a robot costs 160,000 yuan, including purchase and maintenance over two years, and performs the same amount of work, the boss would definitely choose the robot. This is the recognized “two-year payback threshold” for the industry.
We are still two steps away from reaching this threshold:
- Cost reduction: Current industrial humanoid robots cost 300,000 to 500,000 yuan, far exceeding the 160,000 yuan required for payback. According to industry estimates, with China’s supply chain and mass production, costs could be reduced to around 100,000 yuan, which would be within a reasonable payback range.
- Robot intelligence improvement: Current humanoid robots are quite limited; they need retraining for different tasks and are less efficient than skilled workers. The tipping point for widespread adoption will come when robots can perform 80% of tasks independently in unfamiliar environments without human guidance. This could happen in two to three years at the earliest or five to ten years at the latest.
The first practical applications will likely be in enclosed environments like logistics warehouses and automobile factories, where tasks are highly standardized and do not require advanced intelligence.
3. Don’t Be Misled by the Term “Humanoid Robots”—The Profit Models of Three Types of Companies Differ
The industry is still in its early stages, and many companies are categorizing themselves under the broad term “humanoid intelligence,” but their business models are vastly different, and their valuations cannot be compared:
- Companies that build the robots themselves: Examples include Yushu and Zhiyuan. Their core capabilities are supply chain management and cost reduction to produce stable robots in large quantities. Their valuations depend on annual shipments and profit margins, similar to the manufacturing industry, with clear limits and no excessive premiums.
- **Companies that develop robot “brains”: These are like manufacturers of operating systems like Android or iOS for phones. They license their AI systems to various robot manufacturers and earn fees based on usage. If their system becomes the industry standard, they can earn high profits. However, no current AI system can be adapted to all robots and all scenarios, so their valuations are highly volatile.
- Robot labor platforms: These companies provide robots to customers’ factories on an hourly or task-based basis, similar to hiring a robot employee. This model offers stable long-term revenue, but since robots are not yet widely used, this is still in the conceptual phase.
Currently, companies in all three categories are mixing their roles: those building robots are also developing their own AI systems, and those developing software are integrating with hardware. The boundaries between these categories are blurred, and no company is purely in one category.
4. Investing in Humanoid Robotics: Don’t Believe the Claim of a 10-Time Market Size Compared to New Energy Vehicles
There is significant disagreement in the industry. Some predict a market size ten times that of new energy vehicles, while others see it as a gimmick. The difference lies in the perspective: one group is looking at the long-term potential, while the other is focusing on immediate practicality. Ordinary investors should focus on two reliable indicators:
- Don’t believe claims of being “the best in the world”; first, check the actual shipments—can the company sell thousands or tens of thousands of robots per year? Many so-called “orders” are just for display purposes and not actual sales.
- Don’t believe in grandiose claims of “changing the world”; instead, look at the customer payback period—can the robots help customers recoup their investment within two to three years? If not, the commercialization is not viable.
The industry is still in the process of finding its value. Just as the value of the PC and smartphone industries shifted from hardware to software and ecosystems, the value of humanoid robots is still evolving. The leading companies today may not even make it into the top ten in a few years, which highlights both the uncertainty and opportunities of this stage.