Authors: Bohan, Chemistry

Compiled by: BlockBeats

*Slightly edited without changing the original meaning:

Last month, I went to China, visited most of the top-tier investment institutions, and met with the management teams of several leading robotics and biotech companies.

A narrative is currently popular in Silicon Valley: China is winning several key fields of the future—open-source AI, biotech, and robotics. The reasons supporting this concern are quite substantial.

Chinese open-source models have become some of the most frequently used models by Silicon Valley startups. After the U.S. government restricted Fable, ironically, China has instead taken on the role of supporting the global open AI tech stack.

In terms of biotech, most projects in Chinese clinical trials are innovative therapies, while about half of the drugs in U.S. FDA clinical trials are licensed from China.

In the robotics field, China possesses both the structural advantage of scaled production of training data, and more importantly, its hardware development and feedback iteration speeds are astonishingly fast.

But even with these advantages, the Chinese do not seem complacent. On the contrary, there is a widespread strong desire there to understand what Silicon Valley is thinking. Silicon Valley is still regarded as the global innovation center.

Even a person from a top VC firm told me that whenever Benchmark or Sequoia releases a new podcast episode, he lists it as must-watch content for the entire company.

The Chinese know everything that is happening in the West. Content I post on X and LinkedIn is usually translated within a few hours by mainstream media in China's AI circle (AI Era, Synced, or Qbitai). Even comments in the X comment section are translated in screenshot form. You might not even know it yourself, but you are already somewhat famous in China.

This information asymmetry boosts their learning speed and may ultimately help China narrow the gap with Silicon Valley. But at least for now, they still look up to Silicon Valley.

Overall, the maturity of China's capital market is relatively low, and it is much harsher on founders. This environment may breed companies with stronger execution and more aggressiveness, enabling them to defeat competitors in the global market; but huge pressure and personal responsibility may also stimulate more bubbles and fraud.

From Seoul to Tel Aviv, most global tech centers take Silicon Valley as a template. China, however, constitutes another parallel universe in many ways. Understanding how China funds innovation is an interesting path to observe how China got to where it is today and where it is heading.

IPO or Exit

When meeting with founders of many robotics and AI companies, the most surprising thing to me was: almost all of them are planning to IPO next year, and have already started sprinting at full speed.

None of these companies have reached the scale of Unitree Robotics or Moonshot AI—even for the latter two, listing on Nasdaq would probably not be easy—but everyone told me they are preparing to list.

Why? Because they have no choice. In China, many startups list not because they are already qualified for listing, nor because the market timing is mature enough, but because they are forced to list.

One of the facts that shocks American founders the most is that many investment agreements signed by Chinese founders stipulate: they must return investors' capital within a limited period according to a standard higher than a certain minimum rate of return. The period is sometimes six to eight years. If they fail to do so, the company or even the founders personally may bear repurchase and repayment responsibilities.

Chinese LPs and GPs have less patience and have more direct demands for results. There is even a specific term in Chinese to describe this phenomenon: "Equity in name, debt in reality," meaning it is equity on the surface but actually debt.

It is hard to imagine how innovation actually happens in an ecosystem where founders have to bear huge personal responsibilities to create high-risk enterprises. With stakes so high, who still has the courage to start a business?

But Chinese founders are indeed willing to bet everything.

Such incentive mechanisms have shaped a batch of the leanest and toughest companies globally, and also shaped founders who truly devote themselves entirely to the company. When they cannot make money in China's cruel competitive environment, they often choose to expand overseas and quickly overwhelm local competitors.

They are not sophomore students at Stanford University who just participate in a Y Combinator session during summer vacation for the experience. For them, this is a win-all or lose-all game.

This leads to two questions.

Why No M&A Exit?

Why must the exit method be an IPO? Cannot the company be acquired, allowing investors to recover funds through M&A?

The answer is basically no.

There is almost no truly mature M&A market in China, so startups usually can only exit through IPO and must go all the way.

Chinese company valuations are cheap, and labor costs are also low. Rather than acquiring a startup, large companies prefer to directly copy its ideas, and likely copy them faster.

Chinese companies are usually also extremely ambitious and accustomed to horizontal expansion. A smartphone company might simultaneously produce sports cars and develop enterprise software. These factors collectively reduce their willingness to acquire other companies. There, there is also almost no opportunity to provide a soft landing for startup teams through "acqui-hiring."

However, for Chinese founders, a relatively favorable factor is that the IPO threshold is generally lower than Nasdaq or the New York Stock Exchange.

The lower threshold mentioned here does not necessarily mean looser regulatory requirements, but rather that the public market has a higher acceptance of these companies, meaning investors are more willing to buy these companies' stocks.

In recent years, many Chinese tech companies have had almost no revenue or customers. According to current U.S. tech market valuation standards, their scale is far from enough, yet they still successfully completed IPOs.

Of course, the Hong Kong Stock Exchange is currently in a bull market. Zhipu, as one of the few pure large language model listed companies, has also seen its stock price rise significantly. But if these companies were in the U.S., they probably could not complete listing.

One explanation is that individual investors account for a higher proportion in Asian stock markets. Nevertheless, the Hong Kong Stock Exchange, as the preferred listing place for tech companies, still has a higher degree of institutionalization than the A-share market.

We do not know how long this bull market in Asia can last. Many local institutional investors have already started preparing for potential declines in some of the hottest industries, hoping to buy stocks at low prices after a market crash.

Three Types of Capital Pools

Another question is: Why are founders willing to accept such harsh terms?

Shouldn't the free market competition among VCs, like in the U.S., driven by institutions like Founders Fund and a16z, gradually become more founder-friendly?

This change is indeed happening. But the Chinese venture capital ecosystem is still younger than the U.S. More importantly, the different capital sources Chinese founders can choose correspond to completely different incentive mechanisms behind them.

Chinese founders can usually obtain three types of institutional venture capital.

Local RMB Funds

These funds are often supported by provincial or municipal government funds, and the attached conditions are usually the harshest. They often require enterprises to set up offices or factories locally to create jobs and attract talent.

The goal of Chinese RMB funds is usually not just to obtain capital returns, but also to bear the task of driving economic development in the location of the LPs, and their incentive mechanisms are not the same as Western funds that only focus on investment returns.

These requirements often focus on creating jobs and attracting talent, and employment and population inflow help stabilize the local real estate market.

Since this is the case, why do founders still accept this type of funding?

The reason is, if you want to enter the hottest fields such as AI, semiconductors, and robotics, these industries are also highly concerned by the state. Sometimes, only RMB funds can invest, for example, DeepSeek.

Local USD Funds

These institutions include traditional top-tier Chinese VCs, such as Sequoia China, Hillhouse, ZhenFund, Qiming Venture Partners, and IDG. Qiming Venture Partners actually has little relationship with U.S. Matrix Partners anymore. Many of these institutions manage both USD funds and RMB funds simultaneously.

Compared to the first type of capital, these funds are usually more friendly to founders. Over the past twenty years, many household-name companies in China have had their support behind them.

This is the funding source Chinese founders most hope to obtain, especially those companies planning to go global. By the way, from the establishment of Sequoia China until before it later split from Sequoia, it was always the best-performing part in the Sequoia system.

Foreign Funds

The last category is pure Western funds like ours.

Historically, many Western funds once made a fortune in China, such as Coatue and Tiger Global. But now, foreign capital's direct investment in Chinese companies has decreased significantly.

Benchmark's Series B investment in Manus is an abnormal case and is likely the last similar transaction. Obviously, the consequences triggered by this transaction later further dampened the enthusiasm of foreign investors.

Of course, investors always hope they can think contrarily. Perhaps, investing in China is already the last truly contrarian investment thesis in the market.

I once asked a member of Founders Fund what investment directions still count as contrarian now. He also admitted that crypto and defense tech are already very crowded, and China might be the only remaining contrarian thesis.

FA Intermediary Layer

The existence of FAs is also a unique feature of the Chinese venture capital industry.

FA is the abbreviation for Financial Advisor, but everyone directly calls it FA.

They are not wealth management institutions that the name might suggest, but investment bankers serving early-stage financing, responsible for packaging and marketing projects, and matching transactions between startups and VC firms.

The existence of such a huge layer of intermediaries in the entire financing ecosystem confuses me very much.

VC firms actually outsource project sourcing and the first round of due diligence to FAs. FAs are often the first stop for founders to access capital. Many founders are also more willing to cooperate with FAs, letting them help negotiate with shrewd and aggressive VC firms.

But there are obviously conflicts of interest therein.

FAs cannot continuously push reverse-screened, poor-quality companies to a VC firm, otherwise they will lose the trust and access qualifications of that firm. FAs usually charge 2% to 5% of the financing amount as commission. In that system, this has almost become a fixed tax.

I asked a top investor why VC firms would allow this situation to happen. Relying on FAs, wouldn't they lose the excess returns brought by exclusive project sources, and also fail to see good projects earlier than others? His answer was: This is how the industry operates.

Of course, they will also invest in projects without FA participation, but many of the best projects will be led and coordinated by FAs in the first few rounds of financing. FAs will even design the entire financing relay plan in advance: Sequoia China handles the seed round, Hillhouse handles the Series A, and both participate in the Series B. This way, the company can build financing momentum and truly operate quickly.

Invisible Relationship Network

China's social relationship network is neither public nor easy for outsiders to understand. This is both the result of China's relationship-oriented culture and in turn continuously reinforces this culture, profoundly affecting the way daily business activities operate.

China is a society that operates on "guanxi."

LinkedIn never truly entered the Chinese market, and local imitators did not succeed either. You usually can only know others through acquaintance introductions; stepping back, you can only join larger group chats. The WeChat group member limit is 500 people; in comparison, the iMessage group chat limit of merely 32 people is simply not worth mentioning.

Imagine no cold emails, no LinkedIn DMs, and basically no proactive outbound calls. This might explain to some extent why China has never truly developed a mature B2B SaaS industry. This culture naturally also shapes the interaction method between VC firms and founders. Generally, investors will not directly DM a founder.

This is also another reason why FAs exist: they provide "relationship liquidity" for the closed relationship network.

Most Chinese people maintain a certain state of anonymity online and on WeChat. If you add someone's WeChat, the other party is likely to use anime, cartoon, or landscape images as avatars, and use nicknames or pseudonyms as usernames. I have even encountered some Chinese people who refuse to reveal their real names, only willing to use nicknames, or use relatively impersonal English names.

The Hand of the State

The last factor is the development direction and goals set by the state, that is, industrial policies driven by national planning. For this model, the Western attitude depends on whether you ask Capitol Hill or Silicon Valley, or which faction within them.

The government plays a far more important role in China's innovation ecosystem than in the West. The government is both the main LP for many funds, and will also attract startups to settle locally by formulating attractive regulations, providing tax incentives and land incentives. The government also influences the investment direction of VC firms by clearly hoping a certain industry develops. Over the past decade, the most typical example is China's domestic semiconductor industry.

China's brain-computer interface industry provides a more vivid and also more personalized industrial policy case.

Because a certain local government is an identified supporter of brain-computer interface technology, I have communicated with members of its affiliated investment institutions, and they have already invested in many startups in this field. They explained that their primary goal is to establish this strategic industry, rather than pursuing venture capital returns. This is a bit like America's In-Q-Tel.