Humanoid robots: hype or golden opportunity for investors?

Keytrade Bank

keytradebank.be

May 07, 2025

3 minutes to read

Robots that fetch a drink from the fridge for you, entertain your cat or help assemble a car? They already exist. But does that make them an interesting investment?

1. What are humanoid robots?

Humanoid robots are robots with a human-like appearance or movement pattern. They often have two legs, two arms and a head. They are designed to operate in the same environment as humans. They can walk, pick up objects, talk, water the plants and even run a half marathon.

Until recently, they remained mostly prototypes that barely ever left the laboratory. Earlier examples were Honda’s ASIMO (which was able to walk and climb stairs in the 2000s) and SoftBank’s friendly Pepper robot. But for a few years now, we have seen the first humanoids popping up in real work environments. For example, Amazon has been using human-like robots to move objects and organise bins since the end of 2023. BMW is also working with humanoid robots in one of its factories to assemble cars. A humanoid robot does not necessarily mean that it is as intelligent as a human. Until recently, they knew only pre-programmed tricks; they were 'dumb' mechanical puppets. New breakthroughs in AI are now rapidly changing their capabilities. This allows them to better perceive their environment, make decisions and learn new skills. This makes the latest generation of humanoids different from what we saw before.

2. Is there a future for humanoid robots?

Yes, although it depends on further technological breakthroughs, the price tag and social acceptance. The need is already there: in many developed countries, the population is ageing and companies are looking for ways to meet the shortage of qualified profiles. In factories, distribution centres and even healthcare facilities, it is becoming increasingly difficult to find enough people for supportive, repetitive, physical or unattractive work. Humanoid robots can play a role in meeting this need. The fact alone that they can move and orientate themselves in an environment designed for humans gives them an edge over classic robots that are attached to a rail or that follow a pre-defined path.

In addition, there are other challenges that humanoid robots can help tackle:

Reducing occupational accidents: in industries such as construction, mining or logistics, they can take over risky tasks, which not only protects people but also reduces insurance costs for companies.
Increasing productivity: robots don't get tired, make mistakes due to fatigue, and don’t need breaks.
Improving service delivery: in high-contact industries, such as retail or hospitality, humanoid robots can help with peak times or repetitive customer interactions. For example, check-ins at hotels or information points at airports.
Increasing accessibility: in households of people with disabilities, robots can make everyday tasks easier - opening doors, bringing objects, keeping company or providing assistance with mobility. Sooner or later, they may even become a familiar sight in every living room, tasked with tidying up, vacuuming or doing laundry.

3. What does the playing field look like right now?

Hundreds of startups are now competing to build multi-use and affordable humanoids. Asia in particular appears to be a hot spot. The bank Morgan Stanley estimated that 73% of humanoid companies are based in that region. In 2024, more risk capital flowed into humanoid-robot startups than ever before.

Some well-known players in the field:

Boston Dynamics: Known for the Atlas robot, one of the most advanced and dynamic humanoid robots in the world, which can perform a backflip.

Figure AI: This start-up is working on a universally usable humanoid for warehouses and retailers, among others.

Agility Robotics: Developer of Digit, a commercially available humanoid that is mainly used in production, warehousing and logistics.

Hanson Robotics: Developer of Sophia, known for her realistic looks and facial expressions. Hanson Robotics focuses on socially intelligent robots.

Apptronik: Developer of Apollo, a versatile humanoid robot. Apptronik works with NASA, Mercedes-Benz and Amazon, among others.

Agibot: one of the fastest-growing Chinese producers of humanoid robots, whose flagships are the Raise A1 and Yuanzheng A2. These robots are used in factories and industry, but the company also aspires to have applications in households.

1X Technologies: Specialised in light, human-like robots for home environments. The company is backed by investors like OpenAI.

LEJU Robotics: Focuses on intelligent humanoid robots such as the Kuafu, which excels in mobility and versatility. The robots are used in industry, including at car manufacturer NIO.

Tesla: In 2022, a prototype of the Optimus robot was shown at the Tesla AI Day. Elon Musk called the robot "the most important product Tesla is working on" and suggested that the robot business could eventually surpass Tesla’s car sales.

The value chain for humanoid robots naturally branches out far beyond the actual developers. Take, for example, brain suppliers such as Nvidia, Qualcomm, Palantir, Siemens and Horizon Robotics for semiconductors, analytics and data science. and body suppliers such as Valeo, Schaeffler, Infineon and RBC Bearings for motors, batteries, sensors, screws, etc.

4. Market potential: from niche to billion-dollar business?

Today, the market is still tiny. Worldwide, there are only a few dozen advanced prototypes, and the number of sales transactions in 2025 is low. But this is expected to change. Estimates for the future market vary widely, but are unanimously optimistic:

Morgan Stanley projected that 40,000 humanoid robots could be in use worldwide by 2030. That number could swell to 63 million by 2050. For comparison: 63 million robots are equivalent to roughly one humanoid robot per 150 people on earth by 2050.
Citigroup is even more optimistic and predicts a 7 trillion dollar market by 2050, with 1.19 billion humanoid robots in operation.
Goldman Sachs is slightly more cautious and predicts “significant growth in humanoid robotics” with a global market of 38 billion dollars by 2035.
Macquarie, another bank, expects 6.3 million humanoid robots in use by 2035, representing a market of 139 billion dollars.

Of course, it’s best to take these figures with a large grain of salt. But they do illustrate an important point: should the technology gain a foothold, as is the case with generative AI, the potential return for early investors could be great.

It is also interesting that many of the current projects focus on industrial applications. This market itself is already big. But according to some analysts, most of the future value is not derived from the factory floor, but in robots at home – for example, as personal butlers, household helpers or even nannies. A robot that keeps an eye on the children or does household chores could potentially be adopted in hundreds of millions of households.

Currently (as of April 2025), an advanced humanoid robot easily costs tens of thousands of dollars to produce. Morgan Stanley expects a humanoid robot to cost 150,000 dollars each by 2028, dropping to 50,000 dollars by 2040. Only when the price approaches that level or lower can robots gradually enter the consumer market.

5. Risks and challenges: from hype to feasibility

Technological revolutions often come later than initially predicted. For instance, self-driving cars: in 2015, many thought they would be mainstream in 2020. This means there is a risk that humanoid robots may not mature until much later, or that the technology will run into unexpected barriers. For investors, this means that the time-to-market is unpredictable. Early investors can find themselves sitting on dead money if it takes longer before they see revenue and profits come in. If there are few breakthroughs or people don’t seem ready, the hype can turn into disillusionment. We have seen this in recent years, for example, in VR/AR glasses, the metaverse and 3D printing: a lot of buzz with high valuations, but slow adoption led to share price drops. With humanoid robots, too, the danger is that many promises will not be realised.

There is a proliferation of players, but not everyone will win. As is often the case with new technologies, there are likely to be a few winners and many losers. For investors, the challenge is to determine who will ultimately capture a market share. A practical risk is how society will react. The use of human-like robots raises ethical questions. Are rules needed for workplace safety? What about liability if a robot injures someone or makes mistakes? The potential loss of jobs due to automation can also trigger a backlash. Governments could impose taxes or rules on the import of robots to protect human labour. In addition, there is public sentiment to consider: will people accept humanoid robots in their environment?

And last but not least – even if the technology works, it remains to be seen if it will be profitable. Building robots is expensive, and the first generations are likely to be very expensive to buy and to maintain. In short, the huge potential comes with a lot of uncertainty.

6. Investing in humanoids?

Humanoid robots are an exciting but speculative part of the AI/robotics investment universe. For investors who want and can take their chances, there are a few ways to invest in them:

1. Individual shares

You can buy shares directly from companies active in this niche. Some examples:

(Auto) manufacturers: Such as Tesla, Toyota, Xpeng or Boston Dynamics' owner Hyundai, as well as Ubtech and Xiaomi. Although revenue now comes mainly from cars or other products and services, humanoid robots have the potential to create future value.
Subcontractors: These include Nvidia, AMD, Qualcomm, Ouster, Samsung Electronics, Dassault Systemes and Mobileye.
Industrial robotics players: companies such as ABB, Fanuc, Kuka (China’s Midea) already make robots and can expand their expertise to humanoids or parts for them.
Emerging pure players: unfortunately, many (Figure AI, Sanctuary AI, Agility Robotics, etc.) are not yet publicly listed. However, you could look at conglomerates that have an interest in pure players (e.g. Schaeffler AG acquired an interest in Agility Robotics).

2. ETFs (trackers)

A simpler way to invest in robotics is via a specialised ETF. There are several ETFs that track a basket of robotics and AI-related stocks. Some well-known ones:

iShares Automation & Robotics UCITS ETF
Global X Robotics & Artificial Intelligence ETF
L&G ROBO Global Robotics & Automation UCITS ETF
ARK Artificial Intelligence & Robotics UCITS ETF
Amundi MSCI Robotics & AI UCITS ETF

While there are currently no ETFs that focus exclusively on humanoid robots (as of May 2025), the above do offer exposure to the sector needed to enable humanoids.

3. Active fund management

In addition to ETFs, there are also investment funds that focus on technological disruption, AI or robotics. Consider a possible investment in this segment primarily within a long-term view. You invest because you believe the world will look different, not because the upcoming quarterly figures will be spectacular. So keep a horizon of several years (if not decades) in mind.