Liang Wenfeng’s High-Flyer Asset Management crushed the market in 2025, creating a massive financial war chest for GPUs and talent.
- A Financial Windfall: Liang Wenfeng’s hedge fund, Zhejiang High-Flyer, averaged a stunning 56.6% return in 2025, ranking second among China’s largest quant funds.
- Funding the AI Future: With estimated revenues exceeding $700 million, the fund’s success provides ample capital to cover DeepSeek’s staffing and hardware costs, which dwarf the reported sub-$6 million budget for their headline model.
- A Golden Era for Quants: High-Flyer’s success reflects a broader trend in the Chinese market, where quantitative funds significantly outperformed global peers, capitalizing on favorable conditions for systematic trading.
While the tech world buzzes about the disruption caused by DeepSeek’s accessible artificial intelligence, a quieter but equally impressive financial engine has been roaring in the background. Liang Wenfeng, the founder of DeepSeek, is also the driving force behind Zhejiang High-Flyer Asset Management, a hedge fund that just concluded a spectacular 2025. With returns averaging 56.6% across its funds, High-Flyer has not only solidified its standing in the financial world but effectively secured the runway for Liang’s ambitious AI projects.
The Billion-Dollar Engine Behind the AI
Headquartered in Hangzhou, High-Flyer oversees more than $10 billion (approx. 10 billion yuan) in assets. In 2025, this massive capital base was put to work with devastating efficiency. According to rankings released by data provider Shenzhen PaiPaiWang Investment & Management, High-Flyer ranked second among all Chinese quant funds managing over $1.4 billion.
The financial implications of this performance are staggering. Because High-Flyer stopped accepting outside capital years ago, Liang captures the lion’s share of the upside. A conservative fee-based estimate—assuming a standard 1% management fee and a 20% performance fee—suggests the firm generated over $700 million in revenue last year. To put that in perspective, reports indicate that the budget tied to DeepSeek’s headline model work was less than $6 million. This financial asymmetry suggests that Liang possesses a self-sustaining war chest capable of funding top-tier staff, thousands of high-end GPUs, and other critical hardware without solely relying on external venture capital.
A Strategic Pivot to “Long-Only”
High-Flyer’s 2025 victory lap was not accidental; it was the result of a calculated strategic shift. In 2024, the firm moved away from market-neutral strategies—which seek to profit regardless of market direction—and went “all-in” on long-only stock strategies. This bet paid off handsomely as the funds capitalized on market movements using advanced quantitative models.
These “quant” funds utilize complex mathematical and statistical techniques, paired with automated algorithms, to detect investment opportunities that human traders might miss. The 2025 performance was based largely on the 10 index funds operated under the Ningbo High-Flyer Quant Investment Management Partnership, validating the firm’s reliance on data-driven decision-making.
The Year of the Quant
High-Flyer was not the only firm to thrive; 2025 proved to be a golden year for quantitative trading in China. The top funds posting returns between 50% and 70% indicates that market conditions were unusually favorable for systematic stock strategies, rather than a single firm simply getting lucky. In fact, among the heavyweights, High-Flyer was bested only by Ningbo Lingjun, which posted an incredible 73.5% return.
The broader data underscores this trend: Chinese quant funds averaged a 30.5% return in 2025—more than double that of their global peers—while long-only quants averaged 35%. All of the top performers in the large-fund category were quantitative firms, highlighting the dominance of algorithmic trading in the region. While some smaller, nimbler players achieved even higher yields—such as Shanghai-based Hengshui Asset Management with 239%—High-Flyer’s ability to generate nearly 60% returns while managing billions in assets remains a standout feat of scalability.
The Symbiosis of Finance and Tech
Ultimately, the success of High-Flyer serves as a stark reminder that cutting-edge AI requires massive computational resources. For Liang Wenfeng, the lines between financial algorithms and artificial intelligence are blurring. The massive profits generated by his trading algorithms are now being reinvested into the learning algorithms of DeepSeek, creating a self-reinforcing cycle of innovation that few competitors can match.