Skip to content
Technology

Nvidia does not have a No. 2 position, and it is difficult to retain a No. 2 position in the primary market

Text | Gao Jian Pro

One of the oldest institutional problems in the primary market has been re-illuminated by NVIDIA.

On November 12, Nvidia announced that Huang Renxun would resign as president, CEO and board member, effective immediately. Chief Financial Officer Colette Kress serves as interim CEO, and the board of directors simultaneously initiates a formal CEO selection process.

Going forward, around November 6, it was revealed that Huang Renxun had an intimate relationship with Huang Le, a former executive of NVIDIA customer Megaspeed; on November 8, Huang Renxun sent a letter to employees on NVIDIA’s official website, admitting the extramarital affair and apologizing; on the same day, the board of directors announced the launch of an independent investigation. Since then, it has been reported that he also had a close relationship with Nvidia employees.

It took less than a week from the disclosure of the incident to Huang Renxun’s resignation. The incident brought to the fore a more common void: Nvidia never had a No. 2 position.

Nvidia’s organization is so flat that there are almost no middle layers. About 60 senior executives report directly to the CEO, and there is no public succession plan. Now that the temporary team is in place, no one dares to say whether this course that has been in arrears for 33 years can be made up overnight.

“Takami Pro” is more concerned about the fact that the same test question was asked earlier in the primary market and is more difficult to answer.

When GPs are raising funds, some LPs will ask a question that sounds a bit “offensive”: Does the “number two position” exist and who will occupy it? If you left tomorrow, what would be left of this organization? Most answers will stop at titles.

In the history of venture capital development, different institutions have used different organizational forms to respond to the “second position” problem. Dual core, large internal manager, strong echelon, investment committee, each form is the answer given by a generation of founders that is most suitable for the evolution of their own organization under certain constraints.

But the real question is:Do Chinese investment institutions need the second position?

What does the “No. 2 position” in the primary market look like?

In the primary market, as the organization evolves and the team is inherited, management is also changing. Based on the history of the institution and the core pattern formed over the long term, there are roughly four forms of the second position of investment institutions.

第一種,Partner co-governance type.

Jingwei Venture Capital has Zhang Ying and Xu Chuansheng, and Wuyuan Capital has Liu Qin and Shi Jianming. In a co-governance structure, the organization has two “top decision makers.”

Usually, the division of labor between two people is relatively clear. For example, one is responsible for externally responsible for branding, fundraising and LP relations; the other internally is responsible for investment rhythm, investment committee and project management. Different organizations may have different divisions of labor.

At the same time, the two people naturally back up each other, and the handover cost is the lowest. This is the simplest risk hedging. But on the other hand, most organizations will use “industry division of labor” to circumvent the common issue of the boundary of authority and responsibility, but circumventing does not mean solving it.

When two people have different opinions, how to resolve them is an art. In addition, if one of them has much higher public influence than the other, it may actually converge to “one star + one shadow number two”, and the parallel backup relationship can easily degenerate into a master-slave relationship.

第二種,““Ouchi General Manager” style二號位。

The global model is Blackstone. Jon Gray is the President and Chief Operating Officer of Blackstone Group. He joined Blackstone as an analyst after graduating from college in 1992, and became President and Chief Operating Officer of Blackstone Group in 2018. Within Blackstone, he is regarded as Schwarzman’s “quasi-successor.”

The one closest to this form in China is Hillhouse. The founder and top person in charge of Hillhouse is Zhang Lei, and the founding partner and co-chief investment officer Yi Nuoqing is the apparent number two. His relationship with Zhang Lei can be traced back to the time when the two worked together at China Minmetals Group in 1995. When Zhang Lei returned to China to found Hillhouse in 2005, Yi Nuoqing was called to join. In certain core fields, especially biomedicine, he has a strong voice and influence.

From some perspectives, this type of second position is more like an institutionalized management rights node. Succession can be promoted as planned, provided that the founder is willing to truly delegate power. Although there are number two positions in many domestic investment institutions, power is still concentrated in the hands of number one positions.

第三種,Personal IP single core,強梯隊。

Sequoia China is a sample of this type. Shen Nanpeng’s IP strength and public influence as the founder make it difficult for any single partner to be regarded as an equal “number two” by the outside world.

Shen Nanpeng ranks second in the world on the Forbes Midas List in 2026. He has been on the list for 15 consecutive years since 2012 and has reached the top four times. His official website shows that his position is founding and managing partner of Sequoia China. There is a group of strong partners behind him, but the strong echelon can produce Midas-level investors, but it may not produce the “second position” confirmed by the system.

From the management team page of Sequoia China’s official website, Zhou Kui, Liu Xing, Sun Qian and others are displayed side by side as “partners”, with no hierarchical distinction or hint of succession. From a certain perspective, Sequoia China’s model essentially uses a strong echelon to hedge against the risk of missing the No. 2 position, rather than using a clear No. 2 position to solve the inheritance problem.

第四kind, collective decision-making by the investment committee.

Investment institutions with state-owned assets background and various fund of funds managers generally adopt collective decision-making by investment committees. Position 2 is diluted into “everyone is position 1/N”: no one can decide alone what to vote for, and no one can decide alone what not to vote for.

This mechanism almost eliminates key man risk, at the cost of almost simultaneously eliminating the personal sharpness required for excess returns. It is more suitable for managing scale, controlling drawdown, and meeting the compliance requirements of state-owned LPs. But it is difficult for it to grow into the No. 1 spot on the Midas list.

Because excess returns often come from a person’s judgment at a certain moment that cannot be replicated by a committee.

If the Investment Committee uses a committee to dilute the No. 1 position, Benchmark uses a non-hierarchical approach to eliminate the No. 1 position. It has maintained multiple partners with equal rights for a long time, does not have a CEO, does not have a platform team, and does not maintain a huge post-investment system. It solves the No. 2 problem in a “no No. 2” way. Because there is no hierarchy, there is no inheritance.

Overall, the above models are actually answering the same trade-off: security and sharpness cannot have both. The institution’s ability to resist risks increases, while the probability of generating a number number decreases. Each model is the answer of a certain generation of founders to the question “Will I leave?” in the way they trust most, under certain constraints.

Some investors believe that if you want to test any model, you don’t have to look at the title, but only three things: Who can sign the day after the founder is suddenly absent? How do you classify carry? Is position 2 a true partner, rather than a senior employee? Who will the LPs meet when raising funds for the next fund?

Only the organization that can answer these three questions truly holds the No. 2 position. Otherwise, all it has is a title.

This is similar to the view of senior investor Wang Ming.

Wang Ming told “Gao Jian Pro”, “It doesn’t matter whether there is a clear No. 2 position. What is important is to see what the No. 1 position gives up? Is it the right to execute, the right to judge, or credit. It only counts when you reach the third step. The hard standard is whether he can raise the next issue by himself.”

爲什麼投資機構It’s even harder to have a “number two”?

A more fundamental question: Why is it that in manufacturing, Internet companies, and investment banks, No. 2 can rely on systems and sequences to stabilize production, but in the primary market, it is always a scarce commodity?

“It depends on the definition of the second position. It is common for the second position of a general manager to be paired with the same position. If you are good at investing and even have a high profile, unless the first position gives up, there can only be one most important person in the center of the stage. This is also the person that the LP values ​​the most.” An LP who invested in multiple funds revealed.

If you focus on the attributes of the industry, you will find that this industry is a “people-to-people partnership” rather than a “funded partnership.”

A chip company can rely on the CUDA ecosystem, patents, production capacity and cash to survive the absence of the founder; a fund cannot. The core assets of the fund are investors’ industry knowledge, judgment of the future, and long-term trust with entrepreneurs.

The value of the former is mostly reflected in the balance sheet, while the value of the latter is stored more in the people themselves and the relationships between people. A company can survive through systems, but a fund will most likely rely on people.

More importantly, investment judgment is typical tacit knowledge. Why a certain project was rejected during due diligence, why a certain founder deserves to raise money for multiple rounds, and why a certain track should be withdrawn when it seems to be the hottest, these can hardly be written into SOPs.

These judgments rely on intuition and pattern recognition abilities accumulated in long-term practice rather than a set of replicable decision-making rules.

Looking at it from another perspective, the true formation of a No. 2 position is not the day of appointment, but after he has invested in multiple projects with the organization and even missed several opportunities together. The trust accumulated through those common experiences is the source of his recognition from the team when he is promoted to the No. 1 position.

This also explains a counter-intuitive phenomenon: almost all the second positions “airborne” in the primary market failed. It may not be a matter of ability, but a matter of time. And time is precisely the most expensive cost in this industry.

What is even more hidden is the endless verification cycle. To prove himself, a candidate must complete the investment period and exit period of a fund; while some LPs must confirm that he is “good” and have to look at the quality of the next fund. The two rounds of verification took a total of nearly ten years. When he was finally confirmed, the next technology cycle had begun and the topic was changed again.

The life cycle of a fund is much shorter than a person’s career cycle.

Generally speaking, the short cycle of a fund is 6 or 7 years, and the long cycle is 10 years. It takes 8 to 10 years for an investor to enter the industry and form a stable judgment; it takes another 5 years to go from being a partner to being able to independently represent an institution. In other words, the complete cycle of a fund is only enough to cultivate half of the second position.

Institutional handovers often hit the fundraising window, and one of the scenarios that LPs least want to see is the simultaneous appearance of “a new fund + a new batch of managers.”

If you stack this scale with the technical cycle, you will see an even crueler pattern.

PC, Internet, mobile Internet, AI, each technology wave lasts about 12 to 15 years, which is exactly equal to the golden period of a generation of GP. Every time technology changes, there will almost always be a rearrangement of organizational positions, and the winner of the rearrangement is often not who seizes the opportunity, but who completes the handover before the opportunity arrives.

The domestic venture capital circle also has a unique layer of pull.

In the past three years, the fund source structure of the primary market has undergone fundamental changes: US dollar funds have declined, and state-owned assets and industrial capital have become the main investors. The demands of state-owned LPs are depersonalization, accountability, and handover;

This demand has a clear institutional transmission path.

In the relevant policies introduced in 2026, government investment funds and state-owned enterprise investment funds will be included in the penetrating regulatory framework, and the governance responsibilities of state-owned LPs will be significantly increased. In the selection process of managers, state-owned LPs in many places generally require the locking of “key persons” in the LPA. In practice, arrangements with as few as three and as many as five are common. No exit is allowed during the investment period. More than half of the changes will trigger forced exit or major breach of contract clauses.

It should be noted that key person lock-in is a business arrangement between LPs and GPs and is not a regulatory or fund industry association compliance requirement.

And the history of excess returns is almost entirely due to personalization. This constitutes the most profound structural contradiction in China’s primary market in the next ten years.

In fact, when the first generation of domestic RMB and US dollar funds established credit from scratch, the only assets that could be pledged were the founder’s cognition and investment judgment records. This set of investment logic has high efficiency, short decision-making links, and the most concentrated brand premium, but its output is exactly the kind of structure that state-owned LPs are most wary of: once the key person leaves, the value of the fund may shrink significantly, and the corresponding risks will increase.

What state-owned LPs want is an accountable system, and what they want in return are irreplaceable people. Position 2 stands right in the middle of this crack.

Personal LP Zhang Lan believes that this contradiction will eventually fall on people.

“Whether it is the second position or the second echelon, what is tested is the structure of the institutional leader and the ability to absorb talents. Just like when a GP invests in a founder, it depends on whether he can attract and retain key talents, even talents who are stronger than himself in some aspects. Only such entrepreneurs have the opportunity to obtain investment. It is similar when LP chooses GP.” Zhang Lan said.

He selected a GP based on the same criteria for selecting a founder: a No. 1 position without a structure cannot support an echeloned organization. “For funds of a certain size, without a talent pipeline, it is difficult to imagine that they will go through cycles and consistently perform well in different sectors.”

機構化,還是明星化?

When LPs hope or even require GPs to become institutionalized, returns are highly dependent on stardom. So there is a two-way movement.

At one end are star partners setting up their own businesses. For example, Cao Xi left Sequoia China and returned to Midas List with Monolith Capital; Liang Yingyu completed the first phase of raising US$200 million with Huiheng Capital after leaving Kaiming Venture Capital. LP’s buying logic is summarized as “pricing for proven investors rather than paying for track popularity.”

The star partner went solo, taking his judgment record with him to become the No. 1 elsewhere. The “key person” that state-owned LPs desperately target during the selection process is precisely the asset with the most pricing power at the other end of market-based fundraising.

On the other end of the spectrum are established institutions accelerating platformization.

Less than a year after Botha was “forced out”, Sequoia America announced it would raise about $10 billion in new funds, while closing its London office, terminating the Arc accelerator and Scouts programs, merging its early-stage and growth-stage investment teams, and flattening the partnership.

a16z goes further: using its operations team, policy department, and media capabilities to make the institution itself an irreplaceable platform.

The platformization of established institutions is essentially trying to answer the other side of the same question: Can the judgment of “that person” be broken down into assets that the organization can inherit?

Regarding this question, the choices of Cao Xi and Liang Yingyu show that, at least for some people, the answer is no. The organizational dynamics of Sequoia America and a16z also show that even in markets with the most mature institutions, this transformation is far from smooth.

If you look a little longer, position 2 is actually a path that has been traveled repeatedly, but different industries are at different stages.

Family businesses solve it with blood, Wall Street solves it with the MD promotion sequence, and Silicon Valley solves it with the combination of “founders + professional managers”.

The reason why the primary market is particularly difficult is that it has two characteristics at the same time: it is highly dependent on personal judgment, and it must make a commitment to LP in the name of the institution for its survival. The former makes it look like a workshop, while the latter forces it to become a century-old store.

Back to the beginning. Huang Renxun has spent more than 30 years proving one thing: without the No. 2 figure, the company can reach four trillion US dollars. The primary market proves another thing: investment institutions can raise money and invest in unicorns without a second position.

But neither of these two things has proven the third thing: what will happen without the No. 1 position.

The difference is that Nvidia has begun to be forced to answer this question. The answers to Kress’s temporary team and the independent investigation that has not yet been concluded have yet to be revealed.

Sooner or later the primary market will also have to answer.

“The potential No. 2 person is more concerned about this issue than the LP. Because the LP only asks when raising funds and when something goes wrong, whether they are deciding whether to stay or leave every day.” Wang Ming, a senior investor, said.

About Us · 關於我們