Morgan Stanley: Humanoid robot commercialization enters the real-world validation stage, and component leaders benefit more
AI summary card
Morgan Stanley: Humanoid robot commercialization enters the real-world validation stage, and component leaders benefit more
Feedback from the China summit shows that the humanoid robot industry is shifting from concept to commercial deployment, data flywheels, and full-stack hardware-software integration, with the report favoring leading companies in reducers, hydraulics, and automation.
- Dobot and AI^2 Robotics both target sales of 1,000 units in 2026, UBTECH targets delivery of 5,000 industrial humanoid robots, and Linkerbot targets sales of 50,000 to 100,000 dexterous hands, implying 5x to 10x year-over-year growth.
- Real-world deployment is viewed as more important than simple demos, as it can simultaneously generate revenue, scarce data, and a feedback loop for model improvement.
- The robot 'brain' path has not yet converged; neither world models nor Visual-Language-Action models are considered mature enough, and industrial scenarios may require rule-based safeguards and hybrid systems.
- Data is becoming the key bottleneck and a long-term moat; internet, video, and simulation are used for pretraining, while real robot data is the scarcest and most valuable layer.
- The report favors OW-rated component companies such as Leaderdrive, Hengli Hydraulics, Shuanghuan, and Inovance, and also highlights the April 19 humanoid half marathon and IPO catalysts later this year.
Report interpretation
Overview
This report is Morgan Stanley's conference feedback on China's humanoid robotics industry after China Summit 2026 and the firm's China trip. The report covers OneRobotics, Orbbec, Dobot, UBTECH, LimX, Paxini, X Square, AI^2 Robotics, Zeroth, Linkerbot and other complete-system, integration, and related companies, and focuses the investment narrative on commercial deployment, data flywheels, model paths, hardware-software integration, and upstream core component opportunities.
Core views
The report's core view is that the humanoid robotics industry is moving from thematic discussion into real-world testing and delivery ramp-up, with deployment itself becoming both a revenue engine and a data engine; the robot 'brain' paths of world models and VLA have not yet formed a mature consensus, and early-stage adoption may still rely more on rule-based protection and hybrid systems; the data flywheel is becoming the most important moat, especially real-world robot data; Chinese companies in the current immature stage are emphasizing full-stack integration of AI, data infrastructure, and hardware; from an investment perspective, the preference is for core component leaders that can confirm revenue as complete-system shipments scale.
Analysis framework
The report makes its judgment based on feedback from China Summit and company research, combined with company delivery targets, technical path discussions, commercialization progress along the industrial chain, data infrastructure construction, and component-company valuation methods. It uses P/E, P/S, and DCF methods to assess the potential value of core businesses and humanoid robot businesses for component companies.
Methodology notes
Core business valuation
For some companies, the core business is valued using a 2026e P/E multiple; for example, Shuanghuan's core business at 25x 2026e P/E and Inovance's core business at 35x 2026e P/E.
Potential humanoid-robot business valuation
Under the assumption that global humanoid-robot sales exceed 1 million units, humanoid-robot-related sales are valued at 5x to 6x P/S to reflect a longer adoption cycle and stronger growth potential.
Long-term cash flow valuation for robot component businesses
Hengli Hydraulics' humanoid-robot component business and Leaderdrive are valued using DCF, discounting 2025 to 2050 cash flows with an 11% WACC and a 4% terminal growth rate.
Feedback loop among deployment, data, models, and application expansion
Real-world deployment generates revenue and data, data improves models, models enhance capability and expand application scenarios, and this in turn further drives scaled deployment.
The two main directions for the robot 'brain'
The report believes neither route has yet been proven sufficient on its own to support reliable robot capabilities, and industrial scenarios, because they require high repeatability and precision, may need rule-based safety mechanisms and hybrid systems.
Asset mapping & comparison
Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).
- Leaderdrive (688017.SS)A beneficiary of humanoid-robot core components; the report lists it as an OW-rated preferred name
- Strengths
- A foundation in harmonic reducers and industrial robot demand; if it enters the supply chain of humanoid-robot integrators, it can benefit from high-growth scenarios; the DCF valuation reflects long-term cash flow potential.
- Weaknesses
- Still highly dependent on industrial robot demand, manufacturing capex, and the harmonic reducer technology path.
- Comparison
- Compared with complete-system makers, component companies are more likely to convert into revenue through supply-chain penetration in the early phase of industry scaling.
- Risks
- China manufacturing capex comes in below expectations, humanoid robot development is slower than expected, and harmonic reducers cease to be the mainstream solution.
- Hengli Hydraulics (601100.SS)A potential beneficiary of hydraulics and robot components
- Strengths
- Its core business and pump-valve demand provide a base, while the humanoid-robot component business adds upside through DCF valuation.
- Weaknesses
- It is still affected by traditional demand cycles such as excavators and pump-valves.
- Comparison
- Compared with pure-play robot theme names, it has cash flow from traditional businesses plus an option on robot upside.
- Risks
- A sharp decline in China excavator and pump-valve demand, failure to expand non-excavator component share, and humanoid-robot penetration slower than expected.
- Shuanghuan (002472.SZ)A gearbox, reducer, and smart actuator-related component name
- Strengths
- It can benefit from higher penetration of dual-motor, triple-motor, and coaxial gearbox solutions, overseas capacity expansion, and progress in humanoid-robot reducers.
- Weaknesses
- Humanoid-robot revenue still depends on the adoption timeline and customer demand realization.
- Comparison
- The valuation reflects both a 25x 2026e P/E on the core business and P/S potential for the humanoid-robot business.
- Risks
- Market share gains are slower than expected, overseas demand is weaker than expected, and competition in gears and actuators in China intensifies.
- Inovance (300124.SZ)A name with automation, NEV powertrain, and humanoid-robot related exposure
- Strengths
- The valuation framework combines the core automation business, the value of its NEV powertrain stake, and humanoid-robot business potential.
- Weaknesses
- Macro conditions and the pace of high-end automation product development have a large impact on earnings flexibility.
- Comparison
- The valuation uses a mix of 35x 2026e P/E for the core business, equity value from NEV powertrain, and 5x P/S for humanoid robots.
- Risks
- Failure to develop high-end automation products, lower ASPs for low-end products due to competition, and raw-material inflation causing a larger-than-expected gross margin decline.
- Dobot (2432.HK)A robotics company covered in the summit research, with a 2026 sales target of 1,000 units
- Strengths
- It has a clear commercial sales target and can benefit from the expansion of real-world scenario testing.
- Weaknesses
- The report does not assign a coverage rating, and the commercialization target still needs execution validation.
- Comparison
- Compared with component leaders, complete-system/integration players face application-scenario and model-reliability challenges more directly.
- Risks
- Scenario rollout is slower than expected, the model path is not yet mature, and unit economics fall short of expectations.
- UBTECH (9880.HK)An industrial humanoid-robot company covered in the summit research, targeting delivery of 5,000 units
- Strengths
- The delivery target is sizable, and industrial humanoid scenarios may help drive a data-and-revenue feedback loop.
- Weaknesses
- The report marks it as not covered, with no formal rating or target price disclosed.
- Comparison
- Relative to companies with 1,000-unit targets, UBTECH's 2026 industrial humanoid delivery target is higher.
- Risks
- Industrial scenarios require high precision and reliability; if model quality and rule-based protection are insufficient, deliveries and repeat orders could be affected.
- Orbbec (688322.SS)A company related to robot perception covered in the summit research
- Strengths
- Vision and perception capabilities may benefit from robotics hardware and data-stack development.
- Weaknesses
- The excerpt does not disclose a rating, target price, or specific financial assumptions.
- Comparison
- It sits in the humanoid-robot perception supply-chain segment, and its investment logic differs from reducers and hydraulic components.
- Risks
- Industry scaling is slower than expected, perception hardware competition intensifies, and customer adoption is weaker than expected.
- OneRobotics (6600.HK)A robotics-related company covered in the summit research
- Strengths
- Participation in real use-case development and industry feedback helps validate the commercial direction.
- Weaknesses
- The excerpt does not disclose a formal coverage rating, target price, or specific valuation.
- Comparison
- Compared with core component companies, complete-system or integration segments depend more heavily on application rollout and data flywheels.
- Risks
- Application expansion is slower than expected, hardware-software integration is difficult, and delivery scale falls short of expectations.
Key data
- Dobot and AI^2 Robotics 2026 sales target1,000 units eachIndicates that complete-system companies are starting to set commercial delivery targets.
- UBTECH 2026 industrial humanoid robot delivery target5,000 unitsThe report treats this as a signal of industrial humanoid robot scaling.
- Linkerbot 2026 dexterous hand sales target50,000 to 100,000 setsCorresponds to approximately 5x to 10x year-over-year growth.
- Global humanoid-robot sales valuation assumptionmore than 1 million unitsUsed for the P/S valuation scenario for some companies' humanoid-robot businesses.
- Hengli Hydraulics target priceRmb133Derived from the combination of 35x 2026e P/E on the core business and DCF for the humanoid-robot component business.
- Leaderdrive per-share valueRmb269Based on DCF of 2025 to 2050 cash flows, 11% WACC, and 4% terminal growth.
- Morgan Stanley stock rating horizon12 to 18 monthsRatings such as Overweight, Equal-weight, and Underweight reflect risk-adjusted total return relative to the coverage universe.
Impact & implications
In terms of investment implications, the report is not simply betting on complete-system manufacturers; instead, it emphasizes that as the industry enters the real deployment stage, core component companies with clearer revenue recognition, more verifiable technical paths, and more direct demand elasticity may benefit first. For the industry, commercial scenarios, real-world data, model reliability, and full-stack integration capability will determine the early competitive landscape; for market catalysts, the humanoid robot half marathon, IPOs of related companies, and delivery-target execution will influence risk appetite along the value chain.
Risks
- Humanoid robot commercialization progress and penetration are slower than expected.
- World models or VLA models are not mature enough, and industrial-scenario reliability and repeatability cannot meet requirements.
- Construction of the pipeline for acquiring, labeling, denoising, and evaluating real-world robot data falls short of expectations.
- Competition in gears, actuators, automation, and robot components in China intensifies, compressing ASPs and gross margins.
- Overseas OEM demand or overseas capacity expansion is weaker than expected.
- Macro conditions, manufacturing capex, and weakening excavator and pump-valve demand hurt related companies' traditional businesses.
- Changes in the technology path could weaken certain component solutions, such as harmonic reducers no longer being the mainstream option.
- Investment banking relationships and coverage-company ties may create research conflicts of interest, and investors need to judge independently based on disclosures.
What to watch
- Execution of 2026 delivery targets for Dobot, AI^2 Robotics, UBTECH, and Linkerbot.
- The market attention and technical-validation catalyst from the April 19 humanoid robot half marathon.
- IPO progress of humanoid robot companies later this year and its impact on value-chain valuations.
- The speed at which real deployment scenarios expand from inspection, tourism, retail services, manufacturing, and logistics to home assistance.
- Progress in reliability of world models, VLA models, and hybrid systems with rule-based safeguards in industrial scenarios.
- The pace at which core component leaders enter humanoid-robot integrator supply chains and changes in their market share.
- Whether global humanoid-robot sales can approach or exceed the valuation assumption of 1 million units.
- Gross margins, capacity utilization, product mix, and overseas customer expansion at component companies.