This year's return is 56.45%. Zhang Yinxian: Humanoid robots are the most promising golden track for growth in the next decade, and short-term corrections present opportunities for positioning

Recently, Ping An Fund Manager Zhang Yinxian shared his views on humanoid robots at the 2025 Mid-term Investment Strategy Conference.

The investment representative summarized the key points as follows:

1. Driven by three factors: technology, demand, and policy, the investment era of "robot + " has arrived.

2. Humanoid robots will be the most growth-oriented golden track in the coming years and even the next decade, harboring enormous development potential and investment opportunities.

3. We have now entered the fourth industrial revolution dominated by artificial intelligence and humanoid robots. Do we have the opportunity to become leaders? I believe we have the chance, and we should be confident about this.

4. Referring to the development trends of smartphones and new energy vehicles over the past decade, humanoid robots are currently in the budding stage, that is, the 0 to 1 phase. They will gradually move into the growth phase, entering the 1 to N stage.

5. When will large embodied models emerge? It is actually difficult to judge. Just like the sudden emergence of DeepSeep during the Spring Festival, it was an unexpected explosion. What we can do is to closely follow the progress of the three elements: computing power, data, and algorithms.

6. Humanoid robots may still be in a consolidation phase in the short term, but we must not underestimate their long-term potential. We should be more optimistic; a pullback is a better (allocation) position.

My focus is still on finding new processes, new materials, and new technologies.

Zhang Yinxian has 9 years of experience in the securities industry. His managed Ping An Advanced Manufacturing Theme Stock A has performed excellently, with a return of 56.45% year-to-date as of July 16, ranking 18th among nearly a thousand similar funds, and a performance of 106.94% over the past year.

In the ranking of ordinary stock funds in the first half of the year compiled by Capital Deep Dive, Ping An Advanced Manufacturing Theme A returned 53.31%, ranking sixth.

Zhang Yinxian shares the same view as Zhang Lu from Yongying Fund, with both representatives mainly allocating towards humanoid robots, believing that humanoid robots will become a market worth billions Similar to the previous February viewpoint, Zhang Yinxian firmly believes that humanoid robots are a highly growth-oriented golden track in the coming years and even the next decade. The rapid iteration of artificial intelligence, the continuous rise in labor costs, humanity's pursuit of a better life, and the elevation of strategic positioning by policies have all formed an irreversible trend.

Below are the highlights of Zhang Yinxian's sharing, organized by the investment workbook representative (WeChat ID: touzizuoyeben), shared with everyone:

The theme of this report is "Do not fear the floating clouds that obscure your vision; ultimately, we will carry the sparks that illuminate the world." My report is roughly divided into three chapters.

The first chapter discusses how technology, demand, and policy have driven the arrival of the robot investment era. In the second chapter, we can see that both domestically and overseas, various robot manufacturers are iterating and upgrading, so we are currently in a flourishing state. The third chapter will discuss my investment thoughts on humanoid robots.

The Investment Era of "Robots +" Has Arrived

In the first chapter, the investment era of "Robots +" has arrived.

This is mainly due to three driving factors: technology, demand, and policy.

In terms of technology, it includes the rapid iterative development of artificial intelligence, as well as the continuous prototyping and small-batch trial production of core components.

In terms of demand, the intensifying aging population and decreasing labor force will continuously push up the rising labor costs, coupled with humanity's ongoing pursuit of a better life, so this demand will also form a rigid need.

The third point is policy. Both domestically and overseas, the positioning of humanoid robots has been elevated to a very high strategic level. For example, during our country's two sessions, the government work report mentioned intelligent robots for the first time.

In such an irreversible context, I believe humanoid robots will be the most growth-oriented golden track in the coming years and even the next decade, harboring tremendous development potential and investment opportunities.

By 2025, we can see that from around the end of October 2024 to around March 2025, the entire robot sector will perform very strongly, with many targets and indices seeing increases of over one hundred percent, maintaining a time span of about five months.

Why is there such strong performance? The main reason is that 2025 will be the year of mass production for robots, with many participants setting goals to deliver over a thousand units this year, cumulatively reaching the opportunity for 10,000 to 20,000 humanoid robots globally. This is also the core factor driving the strength of the sector.

Referring to the development trends of smartphones and new energy vehicles over the past decade, I believe humanoid robots are currently in the budding stage, that is, the 0 to 1 phase. Subsequently, they will gradually move into the growth stage, entering the 1 to N phase With a strong manufacturing foundation accumulated in the fields of smartphones and electric vehicles, along with numerous visionary entrepreneurs and skilled engineers, China's supply chain is bound to play a pivotal role in the humanoid robot sector.

Another point that cannot be ignored is that China's manufacturing industry accounts for over 30% of the global market, providing excellent training, iteration, and upgrade opportunities for robot manufacturers, which will also accelerate the promotion of humanoid robots. Recently, the application of some niche robots has garnered significant market attention, including unmanned logistics vehicles.

We believe that humanity has now entered the Fourth Industrial Revolution.

The first three industrial revolutions each had iconic products. For example, the first industrial revolution in the 1860s was the steam engine era, followed by the electrical era of the second industrial revolution, and then the information technology revolution of the 1960s, with about 70 years passing since then.

From a timing perspective, this aligns well. The iconic product of the Fourth Industrial Revolution is likely to be artificial intelligence combined with humanoid robots.

Embodied intelligence is the intersection of robotics and artificial intelligence development, where artificial intelligence empowers robots with brains, and robots provide bodies for artificial intelligence. Looking back at the performance of the past two to three years, the rapid rise of artificial intelligence has also been accompanied by a swift increase in the robotics sector.

I also want to emphasize why artificial intelligence is so important for the development of robots. In fact, the species of robots has been researched and developed early on in both Japan and the United States over the past 50 years, but has yet to enter the homes of ordinary people, mainly due to the complexity of robot motion control.

Current mechanical motion control includes model-based control methods and learning-based control methods, commonly referred to as model-based and model-free.

In the past, robot development was mainly based on model-based methods. Now, due to the rapid development of artificial intelligence, we can leverage AI effectively. Model-free control methods such as reinforcement learning and imitation learning have emerged, helping robots achieve better motion control.

For example, domestic robot manufacturers that have gained popularity this year, including those involved in handkerchief flipping, combat competitions, and marathon races, have benefited from reinforcement learning and imitation learning, leading to rapid advancements in robot performance and motion control.

Having discussed the complexity of mechanical motion control, returning to the embodied large models of artificial intelligence, this is very beneficial for the performance transformation of humanoid robots.

Large models have been frequently mentioned in recent years, and they can be divided into non-embodied large models and embodied large models. Non-embodied large models can be understood as text, image-text, and video generation; embodied large models combine physical robots with AI algorithms, allowing them to interact with the physical world, including autonomous driving and humanoid robots.

When will embodied large models emerge? It is actually difficult to judge. Just like the sudden emergence of DeepSeep during the Spring Festival, it was an unexpected explosion. What we can do is to follow the continuous explosion of computing power, the accumulation of data, and the optimization of algorithms, so we must keep up with the progress of the three elements: computing power, data, and algorithms In the first three industrial revolutions, unfortunately, China was either a follower or a bystander. Now, with the fourth industrial revolution dominated by artificial intelligence and humanoid robots, do we have the opportunity to become leaders?

Source: Ping An Fund

The performance and capabilities of our large models are gradually narrowing the gap with the United States. Combining our experience accumulated in the fields of smartphones and electric vehicles, I believe we have the opportunity to become leaders, and we should have confidence in this.

From the demand perspective, many people have made predictions, of course, with significant differences. Here, we present the forecasts from brokerages for 2030, estimating that global demand could reach a cumulative total of 2 million units. According to the targets set by Tesla's Musk, it is expected to reach around 1 million units by 2029 or 2030. The long-term potential is even more vast.

Musk's guidance suggests that the number of humanoid robots in the future could be one to two times the human population on Earth, potentially reaching 10 billion to 20 billion units, far exceeding the number of electric vehicles. Therefore, this is a market worth tens of trillions or even hundreds of trillions, much larger than the current electric vehicle and smartphone markets.

Then, at the policy level. Humanoid robots are a powerful weapon to address the ongoing rise in labor costs. Whether in the United States or China, how to maintain the competitiveness of the real economy and industries?

Controlling labor costs is a necessary step. Therefore, humanoid robots have been elevated to a high position. This year, the government work report during the Two Sessions first included embodied intelligence and intelligent robots in the report. Some regions are actively promoting investment attraction, providing various related preferential policies, and encouraging or subsidizing relevant enterprises for research and development.

The same goes for overseas, where countries like the United States, Japan, and South Korea place great importance on the humanoid robot industry, providing subsidies and tax reduction policies.

Domestic and International Resonance Among Manufacturers, Flourishing Diversity

In the second part, I want to show you the state of resonance among domestic and international manufacturers, with a flourishing diversity. What players are there among humanoid robot manufacturers both domestically and internationally?

It's not just Tesla and startups; there are also many well-known large enterprises. These large enterprises are looking for new major tracks, as only such major tracks are meaningful and flexible for the company.

Therefore, whether it is vehicle manufacturers, home appliance companies, or internet platform technology companies, they are all entering the humanoid robot track.

Overseas humanoid robot manufacturers include Tesla, Figure, Agility, Boston Dynamics, etc. These are some of their forms. The list of domestic humanoid robot companies includes the already listed UBTECH ROBOTICS, and Yuejiang, as well as Zhiyuan and Yushu, which are likely to soon enter the capital market, along with some unlisted but strong-performing companies. This shows some forms of domestic robots.

Additionally, besides many companies entering the robot track, companies in the supply chain must also be taken seriously. Supply chain companies refer to those that assist in the development and iteration of robots, such as NVIDIA and Huawei The robotics sector will definitely see a surge of players emerging like mushrooms after rain. Similar to the Qualcomm chips and Android systems during the smartphone era over a decade ago, which provided a great development platform for many players, the catalytic factors will continue to emerge.

Investment Ideas for Humanoid Robots: New Processes, New Materials, New Technologies

The last part mainly discusses my investment ideas for humanoid robots.

First, let's briefly review humanoid robots.

Source: Ping An Fund

From this chart, we can see that the humanoid robot index and the artificial intelligence index have shown similar trends since the launch of ChatGPT at the end of 2022. The robotics sector may be stronger than artificial intelligence after this year's Spring Festival, primarily because this year marks the beginning of mass production, along with many new companies entering the sector. However, ultimately, humanoid robots are the final carriers of artificial intelligence applications, so their trends are highly correlated.

Looking at the recent timeline, from January to June 2024, the market trends show that until the end of October, the humanoid robot index and the artificial intelligence index were very much in sync.

Source: Ping An Fund

Starting from the end of last year, the humanoid robot sector has performed stronger, with some catalytic events. For example, at the beginning of November 2024, Cybers announced its entry into robotics, and by the end of February, the index had continued to rise for four to five months, peaking before starting to correct. Then, after a sharp drop due to tariff disputes in early April, it rebounded and led the gains.

There were also some catalysts at that time, such as the Beijing Half Marathon on April 19, which, despite some criticisms, gained significant attention. Tesla released its Q1 report on April 23, with the target of one million units being pushed back. These two events were somewhat pessimistic, but stock prices rose instead, which is related to the position of the humanoid robot sector.

Around May 20, there were many videos released from competitions and Tesla's Figure, but stock prices began to correct after mid-May, still related to their position. Investors in growth sectors often overestimate short-term impacts and underestimate long-term impacts, being overly optimistic in the short term and driving the index very high.

At high levels, even with catalysts, the sector's performance can be relatively weak. Therefore, I believe that the humanoid robot sector may still be in a consolidation phase in the short term, but one should not underestimate its long-term potential. Being more optimistic, corrections present better (investment) positions.

My approach is still to look for new processes, new materials, and new technologies, using my engineering background to filter and select opportunities closely aligned with industrial development. For example, the Beijing Half Marathon exposed many issues, such as motor heat dissipation, lightweighting, insufficient power density, and speed limits, indicating the potential for robot downtime. Investment opportunities should be sought based on these events Currently, I am still focused on the perception field, including depth cameras, LiDAR, electronic skin, and six-dimensional force sensors.

Source: Ping An Asset Management

In the control field, I am looking for general-purpose motor drive control chips, preferably those that are self-controllable and have significant potential for import substitution.

There are also new manufacturing processes, such as additive manufacturing for more complex components, especially hollow components, which have inherent advantages and can quickly produce samples. Additive manufacturing is also a key area of my research in new manufacturing processes.

Additionally, there are some new design solutions, such as how to improve motor torque density and power density. Traditionally, radial motors have been predominant; is there a possibility for axial motors to emerge through the humanoid robot industry chain? This also needs to be tracked.

There are also applications of new materials. Therefore, I am willing to seek such investment opportunities.

For example, in the past two years, many outstanding stocks have emerged in areas like lead screws and reducers. However, both investors and practitioners in the industry have rushed in upon seeing their significant value, leading to changes in the supply chain or competitive landscape. Of course, it is still too early to talk about the landscape, but the supply chain is indeed changing rapidly, with new players entering every day.

I believe it is necessary to rely on my industrial background, study mechanics, and look at automobiles and machinery for many years to identify relevant capabilities. Coupled with a valuation perspective, considering the risk-reward ratio, I will conduct screening and identification.

My investment philosophy is focused. Many other investors say that mass production of robots has not yet appeared, making it difficult to discern the truth. However, I believe that as long as one is willing to analyze performance parameters, even using physics learned in school, many things can be identified and screened. Combined with cost reduction trends, company capabilities, and resources, it is still possible to minimize the risk of stepping on landmines.

To give a few examples, lightweight design is definitely a continuous pursuit. Other industries, such as automotive, have also been emphasizing lightweight design, and the same applies to robots.

But what materials are used for lightweight design? There are magnesium alloys, aluminum alloys, and engineering plastics, as well as higher-end materials like PEEK. Many investors and market participants only look at density, without considering strength or specific strength.

Regarding costs, I believe some need to be identified. There are also those who claim to make large investments, announcing plans to invest tens of billions, but do they have the capability and foundation to undertake such projects?

This can actually be seen from financial statements, including six-dimensional force sensors, whether you have the corresponding calibration equipment, rather than just having one-dimensional calibration equipment while making six-dimensional force sensors.

Some directly claim that using direct-drive motors eliminates the need for reducers, but application scenarios must also be considered. For example, can lightweight materials and direct-drive motors be suitable for robots? What power levels and power densities are generally required? All of these need to be analyzed. I will try to utilize my knowledge to reduce the risk of stepping on landmines.

Of course, the market has different styles of profit-making methods, and there is no superiority or inferiority. I hope to find a method that suits me and can yield sustainable benefits.

My operational idea for the Ping An Advanced Manufacturing theme product is: First, to select companies with a high probability of entering the humanoid robot industry. Second, to actively explore companies with improving fundamentals or supportive main businesses, while actively expanding into the robotics field based on technological synergy. When such companies can promote and apply new processes, new materials, and new design solutions in the humanoid robot field, the potential is very large After all, achieving a breakthrough from 0 to 1 often presents the opportunity for a Davis double. This is the type of target I prefer to invest my energy in and research.

Related articles:

This year's return is 39.7%, Zhang Yinxian: The humanoid robot market has a space of hundreds of trillions, and the next few years, even 10 years, may be the largest growth track

"Billion Champion" Zhang Lu: The second half of this year is a key node for the mass production of humanoid robots, and investment should focus on three points

Source: Investment Workbook Pro Author: Wang Li

For more insights from industry leaders, please follow↓↓↓

Risk warning and disclaimer

The market has risks, and investment requires caution. This article does not constitute personal investment advice and does not take into account the specific investment goals, financial situation, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article are suitable for their specific circumstances. Investment based on this is at your own risk