Interview with an early investor of Unitree: In the future, there may be only three players that thrive in the embodied intelligence track, ten that get by, and the rest cannot survive | Hard Krypton Exclusive Interview

Author | HUANG Nan

Editor | YUAN Silai

HUANG Jinping, Founding Partner of Rongyi Investment, has been investing in hard technology for 15 years.

HUANG Jinping is one of China's first-generation wireless communication practitioners. He joined ZTE in the 1990s and worked there for 15 years, witnessing the entire process of China's communication industry growing from scratch to a global leader.

HUANG Jinping speaks with the rationality precipitated by his past resume, showing little emotional fluctuation. The tracks and companies he invests in are just like him, with a strong technical background, covering communication, intelligent vehicles, AI robots, eVTOL and XR glasses industries. The industry is more familiar with the cases Rongyi has invested in, such as Unitree Robotics, Galaxy Universal, Thunderbird Innovation, Future Mobility, Lante Optics, Jiangnan New Materials and so on.

Unitree is probably the most outstanding one among them. Rongyi Investment participated in Unitree's Series A financing as early as 2021, continued to increase its investment in the later stage, accompanied Unitree all the way to its listing, and witnessed the growth and transformation of a hard technology company from being unknown to winning cheers from the general public watching the Spring Festival Gala, investors and industrial parties at the same time.

The capital circle has long reached a state where "everyone is paying attention to robots". Investors focusing on consumer, semiconductor and pure consumer tracks have shifted collectively, and a large number of embodied start-ups have seen their valuations quickly exceed 10 billion yuan, heading for the Sci-Tech Innovation Board and the Hong Kong Stock Exchange.

However, Rongyi Investment, which has invested in Unitree and Galaxy Universal, has chosen to wait and see. HUANG Jinping has been immersed in the industry for decades and has seen many cycles of ups and downs. He has formed an almost instinctive way of thinking: don't chase the trend, focus on the essence of the industry. "A technical route cannot avoid three problems: whether the technology can be realized, whether the engineering can be implemented, and whether the business can form a closed loop," HUANG Jinping said.

HUANG Jinping told Hard Krypton that under the current embodied bubble, market anxiety drives capital to chase short-term valuation dividends, ignoring the long technical iteration and supply chain polishing cycle of hardware robots. The unavoidable industrial growth stage cannot be accelerated by financing, nor can it be bypassed by concepts.

"We can make a deduction: assuming that 20 robot companies will emerge in the future and divide a market with a scale of 500 billion yuan, the ideal state is to expect all of them to live well, but the reality is probably not. Maybe only three are particularly excellent, the 4th to 10th are passable, and the rest may not survive," HUANG Jinping said.

He even poured cold water on the embodied companies that are constantly raising financing now, because HUANG Jinping believes that founders who have had too smooth financing experience are likely to overestimate their own capabilities, leading to deviations in business decisions that easily deviate from the long-term route.

"When a lot of enthusiasm and capital pour in in a short period of time, the market will imagine the development pace too fast, and the excessive rise of enterprise valuation is actually a projection of this over-optimistic expectation," he told Hard Krypton.

When money is surging in the embodied track, we talked with this veteran hard technology investor, hoping that he could provide a more calm perspective for the industry.

HUANG Jinping (first from right), Founding Partner of Rongyi Investment, at the listing bell-ringing ceremony of Unitree Robotics

The following is the transcript of the conversation between Hard Krypton and HUANG Jinping, Founding Partner and Chairman of Rongyi Investment, the content has been edited:

Investing in Unitree

Hard Krypton: Rongyi invested in Unitree in the second half of 2021, which was a very early entry. What made you attracted to it?

HUANG Jinping: When we invested in Unitree, the concept of large models had not yet emerged. We were simply optimistic about the robot track, and spent time looking at almost all related projects on the market at the same period.

First of all, we believe in the long-term value of this track. Although the commercialized results were not clear at that time, we recognized the underlying technology and the future market space. After the logic was established, we compared the differences between different enterprises from various dimensions.

What impressed me most about Wang Xingxing at that time were two points. The first is his almost obsession with robot technology. Whether it is the early quadruped robot or the subsequent humanoid product, the R&D of the body itself has an extremely high technical threshold. Wang Xingxing was in a fanatical research state, willing to calm down to polish the product and achieve extreme performance. At that time, the supporting facilities of the industrial chain were not perfect, and the leading supply chains were unwilling to supply goods for niche robot manufacturers, so a large number of parts and debugging work had to be done by themselves. This kind of obsession and extreme characteristics were particularly important.

The second is a strong sense of cost control. He not only paid close attention to the production cost of hardware products, but also strictly controlled the overall operating expenses of the company, which is a very rare quality for young entrepreneurs.

So to sum up, we first locked the track, then checked the feasibility of the technical route, and finally invested in Wang Xingxing through horizontal comparison. Of course, there is also a part of luck in making this accurate bet. The subsequent outbreak of AI large models has also greatly promoted Unitree.

Hard Krypton: At that time, Unitree was still a quadruped robot company and had not yet entered the humanoid body field. Was there big internal disagreement when Rongyi promoted the investment decision?

HUANG Jinping: There was almost no disagreement inside Rongyi. On the contrary, we not only invested ourselves, but also took the initiative to introduce more than 20 familiar institutions to talk to Unitree. As a result, none of them followed up the investment. When these institutions wanted to enter later, the window period had passed.

The bell-ringing ceremony of Unitree Robotics (Photo by Rongyi Investment)

When we evaluated Unitree, we not only looked at "whether its technology is good", but also "where the market is". Unitree's robot dog adopts the electric drive route. Compared with Boston Dynamics' expensive hydraulic solution, electric drive can greatly reduce the hardware cost, and the technology is universal, which can not only be used for quadrupeds, but also extended to humanoids. The engineering implementation and commercial space are very clear.

But it must be admitted that the short-term market space corresponding to the quadruped robot dog at that time was very limited, mainly in local markets such as inspection tasks, with small scale, and scenarios that can undertake heavy load and industrial operations were still being explored.

Most financial-oriented institutions mainly focus on the current order size to calculate accounts, and lack the research and judgment of the long-term value of the technical route. This is the most fundamental difference in the judgment logic of all parties.

Hard Krypton: Has Unitree's competitiveness structure changed in recent years? Which of the traits you valued back then have been verified?

HUANG Jinping: In the early stage, an enterprise mainly relies on technical route judgment, product engineering and market capabilities to gradually build its brand. Unitree is also in this state.

But what we talk a lot about with Unitree is that in the next stage, we need to turn the best suppliers, the most cutting-edge innovative technologies and the highest-quality industrial resources into our partners, form systematic advantages, and gradually transfer the real barriers to the ecological level. The ecological construction at this stage requires the power of capital, especially CVC, to leverage. Your moat depends on whether you can gather the best partners.

Hard Krypton: Has Wang Xingxing himself changed?

HUANG Jinping: Over the years, he has formed his own set of methodologies and business concepts for running an enterprise. After the external aura came, all kinds of affairs and external disturbances increased significantly, which was a big challenge to the operation of the enterprise. But the benefits are also real: Unitree hardly spent any advertising cost, but enjoyed huge brand dividends, and all kinds of resources came to the door actively.

Overall, the Unitree team has done a very good job in this process, with strict management, always focusing on the main business, and not being distracted by external voices. When the technology and products are gradually mature and the market demand rises, they are ready to take over, and the capital chain is also very abundant.

The core still returns to the founder himself, whether he can control the changes of the industry and the growth rhythm of the company. From this point of view, Wang Xingxing's determination and judgment are absolutely very prominent among his peers.

Wang Xingxing (Source: Unitree official website)

Hard Krypton: From the perspective of investment return, is Unitree's growth rate within your expectation?

HUANG Jinping: It definitely exceeded expectations. But this over-expectation comes more from the changes of the industry itself. We did not expect that embodied intelligence would develop so fast at the beginning, and the large model wave has pushed the rhythm of the whole track forward by a big step.

However, external perspectives often overestimate the short-term implementation speed of technology. Comparing with autonomous driving, which has been developed for many years, the driving itself is a relatively standardized scenario - with limited control dimensions and relatively clear rules, it still experienced a long climbing period. Robots are far more complex. To work in the real environment, the data accumulation is far from enough, and the model capability still has a long way to go. These can all be solved, but enough patience and time are needed.

When a lot of enthusiasm and capital pour in in a short period of time, the market will imagine the development pace too fast, and the excessive rise of enterprise valuation is actually a projection of this over-optimistic expectation.

Teams that have not experienced cycles are prone to failure

Hard Krypton: Since the beginning of this year, the embodied intelligence track has been very hot. As a front-line investor, what obvious changes have you observed in market sentiment?

HUANG Jinping: The feeling is quite obvious. There has been a clear bubble in some popular directions in the track, which is not a good thing from the investment perspective. Capital is essentially asset allocation. If a large amount of capital crowds into the track, enterprises will use investors' money to engage in disordered involution, and excessive competition will eventually hurt the development of the whole industry.

At the same time, once an enterprise obtains excessive financing, the return expectation of capital will quickly turn into the pressure of business growth, forcing the enterprise to pursue a path that can produce results in the short term, rather than polishing technology and products steadily.

Hard Krypton: How to judge whether a company can survive and have the ability to continuously accumulate strength?

HUANG Jinping: To verify the technical feasibility and business competitiveness of a company, we still need to judge in a long cycle.

We all hope to invest in leading enterprises. Of course, there is another common strategy: invest in several more companies outside the leading ones. As long as the track is hot enough, the valuation will rise in the later stage, and you can make money by selling and exiting at that time. This is essentially an arbitrage logic between the primary and secondary markets, based on the assumption that the capital heat of the whole track will continue to spill over, and non-leading companies can also be driven up.

But back to the industry itself, this logic does not hold. The final state of the industry must be that the leading players are getting stronger and stronger, the middle and lower players will gradually differentiate within three to five years. Many companies, even if they get high valuations, their business may not keep up, and the valuation will inevitably fall back in the end.

Before 2023, the robot track was not clear, and there were very few teams that could get through the technical path and have engineering capabilities. We judged that Unitree's leading position would become stronger and stronger. In 2023, when embodied intelligence broke out, more teams entered and more capital poured in, but our view on the final state of the industry has not changed. Therefore, we have laid out some vertical companies around the upstream and downstream links, but our investment logic is still only investing in leaders.

So this is not a simple trade-off between two options, but a clear industrial judgment: the short-term market verifies the survival ability of enterprises, and the long-term track determines the growth space of enterprises. Only by forming a progression between the two can enterprises go far.

HUANG Jinping, Founding Partner of Rongyi Investment (Source: the enterprise)

Hard Krypton: There are many embodied intelligence companies that were established in a short time but have high valuations. From the investment perspective, compared with companies like Unitree that have experienced a relatively complete industrial cycle, what are the specific gaps of these companies in terms of technology accumulation, product engineering, and team running-in? Does the short-term high valuation mask these gaps to some extent?

HUANG Jinping: Pure short-term valuation is indeed not equivalent to the real operating status and technical strength of an enterprise. We always have a metaphor: the growth of an enterprise is the same as the growth of a child. No stage of growth can be skipped. Even if there are genius exceptions, skipping the necessary growth stage will inevitably lead to various shortcomings.

Founders who have had too smooth financing experience are likely to overestimate their own capabilities, leading to deviations in business decisions that easily deviate from the long-term route. The core root is that after getting a large amount of capital, they are burdened with the performance expectation of capital. In order to deliver satisfactory results in the phased stage, the enterprise's business actions will inevitably be distorted, which is a common phenomenon.

Take Unitree as an example. The company entered the market in 2016. Its technology and products have been polished for so many years, and the industry accumulation is built up through years of trial and error and iteration, which cannot be achieved in a short time. In the early years, the upstream supply chain supporting facilities were not perfect, there were no mature suppliers for core components, and manufacturers were very unwilling to cooperate. Unitree had to do independent R&D, which in turn helped it control the cost. But as a new team that has been established for less than one or two years, even the complete machine has not been fully developed, where does it have the energy and ability to independently develop key components?

Furthermore, many new teams are put together temporarily. They look very strong on the surface, but in fact, the running-in time is too short, and they have not experienced tough battles and setbacks. Once the market environment changes or the supply chain has problems, all kinds of hidden problems will be exposed at once.

The growth of an enterprise cannot be rushed. It needs to be polished slowly in actual combat.