How China’s Venture Capital Misfires Are Fueling a Parallel World of Startup Success
In China, Two Startup Worlds Collide: The VC Mirage and the Builders' Frontier
In the chaotic terrain of China's tech ecosystem, two parallel realities are unfolding. On one side: well-funded venture capital firms with track records marred by failure, pouring capital into startups that seem destined to falter. On the other: visionary founders, often overlooked by investors, quietly building world-class companies that defy expectations and reshape industries.
This duality has become increasingly visible in recent years as once-celebrated venture capital brands fall under scrutiny. Among them is Innovation Works, the high-profile fund launched by Kai-Fu Lee, whose poor returns and eventual rebranding into AI startup 01.AI have stirred controversy across Chinese social media.
Critics claim the fund's average DPI (Distributions to Paid-In capital) was below 1, signaling that investors failed to recoup even their initial stakes.
DPI, or Distributions to Paid-In Capital, is a crucial performance metric in Venture Capital. It measures the actual cash returned to investors (Limited Partners) relative to the total capital they have contributed to the fund, showing realized gains.
"Please remove the word 'possibly' when calling it a failure," wrote one user on a viral forum thread, echoing a growing sentiment among limited partners (LPs) and founders alike: that China’s early-stage VC landscape has become a playground for hype and incompetence.
A Market Overrun by Mediocrity
Despite abundant capital, many Chinese VC firms have adopted a risk-averse yet paradoxically reckless approach. They flood the market with cash, backing trendy sectors and flashy pitch decks, but often neglect the crucial due diligence required to evaluate founder capability or product feasibility. Incompetence in founder selection, not regulation or macroeconomics, has emerged as the quiet killer of venture returns.
Venture Capital funding volume in China over the past years
| Year | Funding Volume (USD) |
|---|---|
| 2018 | ~$150.2 Billion |
| 2020 | ~$56.4 Billion |
| 2021 | ~$88.5 - $132.7 Billion |
| 2022 | ~$44.2 - $67 Billion |
| 2023 | ~$45 - $68.8 Billion |
| 2024 | ~$33.2 - $40.2 Billion |
In an ironic twist, this environment has rewarded survival over vision. VCs are incentivized to spread money across dozens of similar, low-conviction bets. The result: a bloated pipeline of mediocre startups that burn through funds without clear paths to product-market fit.
"It's not the policies or Xi Jinping that's throttling innovation," said one early-stage founder who declined to be named. "It's the people writing the checks. They're chasing the wrong things."
The Builders in the Shadows
While capital chases noise, innovation has found refuge in the overlooked corners of China’s tech scene. Take DeepSeek, the Hangzhou-based AI company whose breakthrough R1 model has been embraced by institutions across the country. Or DJI, the drone titan that continues to dominate the global market through relentless product execution.
These are not anomalies. They are proof that, beneath the surface noise, China remains fertile ground for extraordinary innovation. But many of these founders operate without the backing of elite VC networks. Instead, they grow through strategic partnerships, international attention, or in some cases, state-supported R&D.
Kai-Fu Lee, now leading 01.AI, appears to be navigating this complex landscape. While critics suggest his shift to AI development represents a strategic pivot from his venture capital challenges, 01.AI has produced open-source large language models like Yi-34B and the Yi-Coder series, which gained some initial attention.
However, skepticism remains high, as the company recently ceased pre-training and pivoted to platform solutions built on third-party models. Analysts pointed out Yi-* models failed to build a substantial user base and were quickly outperformed by newer models.
The Looming Correction
Industry observers warn that this split ecosystem may be unsustainable. A correction seems inevitable. Poorly performing VC portfolios could soon face capital withdrawals, shrinking LP appetite, and loss of access to high-quality deal flow. Meanwhile, builder-led companies may continue to scale, driven by performance rather than perception.
If China is to reclaim its reputation as a global innovation hub, the tide must turn. Investors must recalibrate their criteria—moving beyond hype cycles and reengaging with the hard, often unglamorous work of identifying real talent and technical rigor.
"This is the epiphany," said one angel investor. "China doesn't have a funding shortage. It has a filtering problem. The VCs are playing a different game. But the real builders? They're still out there. They're just playing to win."
