China's quant funds just crossed $474 billion in assets, and investors are throwing money at AI-driven strategies faster than firms can open the doors.
In May, Ubiquant opened a new fund and closed it in under two hours after raising 2.6 billion yuan, about $384 million, according to Bloomberg. Days later, investors bought more than 100 million yuan of a product from Shenzhen ChengQi Asset Management within seconds. That is not normal demand. It is what happens when a market that regulators nearly cooled off decides AI is the next edge.
China's quantitative fund industry ended 2025 managing 3.22 trillion yuan, roughly $474 billion, Bloomberg reported, putting it ahead of any other emerging market and much closer to the US and European quant industries than looked likely only two years ago. If you're watching this from outside China, the number is the point. A local strategy business that once looked like a rough copy of Wall Street's quant shops is now large enough to move money, talent and computing capacity on its own terms.
That scale looked unthinkable in 2024. Assets across China's quant industry had fallen to about 1.13 trillion yuan by the end of that year, down from 1.92 trillion yuan in 2023 and 1.5 trillion yuan in 2022, after regulators cracked down on high-frequency trading practices they blamed for worsening market swings. The rebound since then hasn't come from regulators suddenly becoming relaxed. It has come from performance, and from the belief that machine learning can find trades human managers miss.
More than 60% of China's quant funds now use some form of AI or machine learning in their strategies, according to the Bloomberg report. You should treat that figure carefully. AI can mean a serious research stack, or it can mean a fund manager has attached new language to an old model. But money is not waiting for the distinction. Investors are behaving as if the better firms have already proved enough.
No firm captures that shift better than High-Flyer. The Zhejiang-based quant fund, controlled by DeepSeek founder Liang Wenfeng, returned about 57% in 2025 while managing more than 70 billion yuan, roughly $9.7 billion, according to data compiled by PaiPaiWang Investment & Management and cited in public profiles of the firm. Liang built High-Flyer's AI infrastructure years before DeepSeek became a global name. In 2019, the firm began building its first computing cluster, Fire-Flyer, at a reported cost of about 200 million yuan. By 2021, it had bought roughly 10,000 Nvidia A100 chips for a larger system.
Those chips now look like one of the best early purchases in Chinese finance. DeepSeek, founded in Hangzhou in 2023, used the same culture of model-building and compute discipline to rattle the US AI market with low-cost models. The Wall Street Journal reported in June that DeepSeek raised more than $7.4 billion in its first funding round at a valuation above $50 billion, with Liang himself investing around $3 billion. That's a hard fact worth sitting with: one of China's most watched AI companies came out of a quant shop, not a university lab or a cloud giant.
Frankly, that makes the China quant boom harder to dismiss. This isn't only a story about funds putting AI in a pitch deck. It is also about people who learned to turn data, chips and models into trading returns, then carried the same habits into frontier AI. Liang is both the founder behind DeepSeek and the figure tied to one of the funds proving that AI can still produce a trading edge when the market gives it enough data to chew through.
The rest of the field is scaling quickly too. By April, 71 China-based quant managers had crossed 10 billion yuan in assets, up from 61 in March, according to industry data reported by Caixin Global. China's private fund industry reached a record 23.5 trillion yuan, about $3.5 trillion, by the end of April. The useful question is not whether quant has become important in China. It has. The question is how many managers can keep producing returns once everyone is looking for the same signals.
Minghong Investment shows both sides of that problem. The firm, founded by Qiu Huiming, a physics PhD from the University of Pennsylvania and former Millennium Management employee, manages roughly 80 billion to 90 billion yuan across strategies including index enhancement, market neutral, CTA and macro products. It sits near the top of the domestic quant league table. It also suffered a roughly 15% loss in one macro product over two weeks in March, a useful reminder that faster models don't make drawdowns disappear.
A market can't add 10 billion-yuan managers at this pace without crowding the trade. When dozens of AI-enhanced funds are hunting short-term inefficiencies in the same A-share market, each fund is fighting not only human rivals but machines trained to notice the same patterns. The best firms may still win. The weaker ones will discover that calling a strategy AI-driven doesn't protect you when returns compress.
China has built a huge quant industry in a very short time, and it has done so while AI talent, cheap domestic ambition and scarce high-end chips are all pulling in the same direction. Whether the next 71 managers reach the 10 billion yuan mark as easily, or find that the easy money has already been competed away, is the question investors have not had to answer yet.
Also read: Crypto's coders are quitting for AI, and the industry can't ignore it • Shanghai Bets on a New Quantum Computing Zone to Beat Rival Chinese Cities • Singapore Will Fine Cloud Giants a Million Dollars for Going Dark
