When a Limited Partner Teaches Me How to Invest Using Doubao: A Private Equity GP's Career Change Story

Original by Odaily (@OdailyChina)

Author: Golem (@web3_golem)

As LPs begin to use AI, the days for small private fund managers are getting tougher.

Er Gou (@ryansoon777) was a general partner (GP) at a small offshore USD-denominated private equity fund focused on U.S. stocks before the Chinese New Year. However, after the holiday, he resigned to join an AI startup.

“Raising capital for small private funds is already difficult. With the rise of AI, many LPs would rather use tools like Doubao to assist with stock trading than entrust their money to us.”

Er Gou said his decision to switch careers was largely due to observing the subtle impact AI has had on the relationship between LPs and GPs. On the surface, information and analytical capabilities have been leveled by AI. LPs are finding it easier to question the professional judgment of GPs, which can increase friction between the two parties, potentially leading to capital withdrawals or redemptions.

The Struggling Small USD Private Funds

The USD private fund Er Gou previously worked for was actually in decent shape, managing tens of millions of dollars in assets. It primarily invested in highly liquid U.S. stocks, with a small allocation to crypto asset management. Its annualized returns over the past three years had significantly outperformed the Nasdaq index.

Logically, with strong performance and increased investor demand for overseas wealth management in recent years, capital raising shouldn't be too difficult. However, Er Gou revealed that, in reality, it's nearly impossible for small USD funds like theirs to attract institutional LPs.

Currently, top-tier, multi-billion-dollar USD private funds in China (such as Jinglin, Hillhouse, and Boyu) typically use an "offshore + onshore" structure. The fund entity is based in the Cayman Islands, often registered as an exempted company or SPC, while the management entity is based in Hong Kong or Singapore.

In recent years, due to changes in regulatory and fundraising environments, more and more USD private funds are adopting purely onshore structures like Hong Kong LPF or Singapore VCC.

However, the small USD funds Er Gou joined still rely on the most "primitive" USD fund structure: the Cayman SPC + BVI (British Virgin Islands) fund manager structure.

A common saying in the fund industry is that "LPs determine the structure." One reason top Chinese USD funds stick with the "Cayman" structure is that their overseas LPs include U.S. university endowments, Middle Eastern sovereign wealth funds, and large European family offices. These top-tier international "old money" investors have been familiar with the Cayman structure for decades. Continuing to use this framework reduces communication and trust costs between them.

However, small domestic private USD funds also based in the Cayman Islands cannot attract these top international capital flows. Their LP base is primarily in Asia, putting them in an awkward position.

From an Asian perspective, the capital behind USD private funds mainly comes from private banks, mainland China (overseas investment capital), Hong Kong local family offices, and wealthy individuals in Southeast Asia.

Even for similarly sized small USD funds, these circles have a natural affinity and sense of security towards Hong Kong or Singapore, making them more willing to invest in Hong Kong LPF or Singapore VCC structures over Cayman SPCs.

Besides the constraints of fund structure and scale, differences in investment strategy also make fundraising difficult for small funds like Er Gou's.

Among the investment strategies used by private funds, the main categories are discretionary strategies and quantitative strategies. Discretionary strategies involve the GP deciding what to buy and sell based on their own research, experience, and judgment, where profitability hinges on the fund manager's market insights. Quantitative strategies involve writing investment logic into mathematical models and programs, executing trades automatically or semi-automatically at high frequency, where profitability relies on statistical patterns identified by the model.

“Currently, funds using quantitative strategies find it easier to raise capital than those using discretionary strategies. Especially with the empowerment of AI, LPs are more convinced by quant strategies,” Er Gou stated, noting that the enthusiasm for quant strategies increased significantly after DeepSeek (Odaily note: incubated by the quant fund High-Flyer team) became popular last year.

Furthermore, the key difference between quant and discretionary funds is that quant strategies can demonstrate data and algorithms to gain LP trust. Whether the fund is profitable or facing drawdowns, it remains within controllable parameters. Top-tier quant funds can even be presented as fixed-income products. Discretionary strategies are more abstract. GPs need to expend significantly more communication effort to gain full LP trust, especially during large drawdowns when LPs can easily question the GP's investment ability.

Therefore, in summary, the survival space for small USD funds of the type Er Gou worked for in China has been compressed by the macro environment, making fundraising increasingly difficult. The remaining large LPs in these funds are also questioning whether AI's "investment ability" far surpasses that of the GP.

The "Diverse" LP Base

“In the past, LPs generally listened to us because we were professionally trained. Now, they take our reports, feed them to AI to translate into plain language, and then come back to 'teach' us how to do things,” Er Gou said. Since AI became widespread, LPs who previously only cared about the final results have shown significantly increased “interest” in his investment operations.

Er Gou once even had to redeem an LP because of this. This LP was a 50-year-old real estate business owner, very much an "alpha male" type. He had invested about $1 million USD in the fund where Er Gou worked. However, he didn't take a hands-off approach. He frequently argued with Er Gou using fragmented market information and conclusions drawn from AI. “His attitude was terrible. He thought I was just a young guy who didn't understand anything. We couldn't build trust, so eventually, after some coordination, we had him redeemed.”

“Honestly, our LPs are very successful people in their own fields. They are authorities in their domains. But now that they have AI as an assistant, they believe they have become authorities in investing as well,” Er Gou lamented.

Small USD funds often have a "diverse" LP base because their fundraising channels are narrow, primarily relying on friends of the boss or referrals. According to Er Gou, the LPs in his fund included high-net-worth individuals from mainland China, business owners, and FoFs (Fund of Funds). “Our LPs include a Shanxi coal mine boss, a billionaire ranked in the top 300-400 on the Forbes list, and even some second-generation rich kids who were friends with us and introduced their fathers.”

Their relationship with LPs was also nuanced. For some LPs, the fund wouldn't even charge the standard 2% management fee, only taking a 20% performance fee. The biggest characteristic of this LP structure was enthusiasm for participating in financial markets and "moving capital overseas," but they lacked the time and energy to quickly learn and research market trends.

Therefore, in a sense, the core value of a GP lies in undertaking tasks like information gathering, market research, opportunity screening, and investment judgment for the LP. They use their professional expertise to compensate for the LPs' lack of time, energy, and knowledge, thus completing the conversion process from information to decision-making.

However, with the popularization of AI tools, the past information processing and research capabilities that heavily relied on professional institutions are being rapidly democratized. Except for the final capital allocation and trade execution, a significant portion of the traditional GP functions is now being replaced by AI at a lower cost and higher efficiency.

“It's not difficult for our LPs to open a brokerage account with someone like IBKR. With AI assistance, they can buy whatever stocks or sectors they like on their own.” Er Gou believes AI has a particularly significant impact on funds employing discretionary strategies. Since investing is ultimately result-oriented, if an LP catches the right trend and their personal investment returns exceed the fund's performance, they will naturally begin to question the fund's capabilities.

In contrast, the "information equalization" brought by AI has a smaller impact on quantitative funds and might even widen the gap between funds.

The parameters and algorithms within quantitative fund strategies are constantly evolving. The addition of AI accelerates this iteration process. This is a domain of pure competition in efficiency and intelligence. An ordinary person using a standard AI to build a quantitative strategy can absolutely not compete with large quant funds without specialized knowledge in mathematics, finance, and other related fields.

“Quantitative strategies, by nature, require continuous innovation ahead of market peers to generate excess returns. If you think your ordinary AI has built a good strategy, chances are it's already been discovered and iterated upon by most smart people,” Er Gou explained the advantage of top-tier quant funds.

Will AI Replace GPs?

However, Er Gou isn't worried that AI will completely replace professions like GPs or analysts. AI will always be neutral and accessible to everyone. It's a lever. GPs can use AI to enhance their knowledge systems and investment strategies to generate more returns for LPs. What truly frustrates Er Gou is that AI has increased friction between GPs and LPs.

“Some LPs will even question why you didn't invest in a hot stock, and they'll explain their reasoning quite convincingly. They don't understand that GPs aren't just buying whatever's trendy at the moment,” Er Gou finds this phenomenon slightly exasperating, especially since U.S. AI and semiconductor stocks became a major theme, allowing retail investors to achieve excess returns just by betting on the sector leaders.

During a bull market, retail investors can easily outperform funds. First, personal investing is more flexible, has more room for error, and capital is more focused. Second, with AI-assisted research, a retail investor's research efficiency is greatly enhanced, akin to having a versatile expert on call 24/7.

Especially in this year's U.S. stock market, if a retail investor correctly bet on popular memory chip stocks like SanDisk, Micron, or SK Hynix, their returns could surpass most funds. “At this point, LPs might either consider allocating more capital to their own accounts and less to the fund, or simply redeem their capital from the discretionary fund entirely,” Er Gou says. In a bull market, everyone tends to think they are a 'stock god'.

But all of this presupposes that retail investors know how to use AI correctly. Using inferior AI can be counterproductive, and Er Gou says this is the biggest source of friction with his LPs. “High-net-worth individuals in China primarily use companion-style conversational AIs like Doubao. More analytical tools like ChatGPT or Claude are not as widespread. To provide emotional value to users, these companion AIs are highly prone to hallucinations in professional domains.”

Essentially, the problem isn't the quality of the AI, but that most people don't truly understand how to use it. AI can integrate vast amounts of information in seconds to construct a logically consistent analytical framework. However, logical consistency doesn't equate to factual accuracy. For LPs without professional backgrounds, it's often very difficult to distinguish which conclusions are based on real data and which are merely probabilistic inferences generated by the model.

Therefore, rather than seeking analysis from AI, most investors are actually seeking validation. The ultimate goal of these conversational AIs isn't to help investors "sift truth from falsehood," but to complete the dialogue.

So, will AI replace GPs? AI can cheaply generate ten thousand logically sound investment research reports. But asset management, at its core, is an "ancient service industry" built on trust and the entrustment of one's capital. The relationship between a GP and an LP is also a process of mutual selection.

However, in a future where all "tasks" might ultimately be executed by AI to maximize "results," "human private funds" should also learn from AI – perhaps by practicing how to provide more emotional value.