Urgent.News

600+ sources. One page. See who else covered it.

Editions

AI

How China’s young managers grappled with billion-yuan mandates as AI shocks hit portfolios

As Leopold Aschenbrenner’s US hedge fund saw assets wiped off by more than two-thirds in a single month, some of China’s new portfolio managers also felt the shock across the Pacific, learning bitter lessons early in their careers. The 50-day market turmoil, sparked by a global correction in artificial intelligence stocks in June, turned some of China’s rookie managers into an unwitting focal…

How China’s young managers grappled with billion-yuan mandates as AI shocks hit portfolios

China's emerging portfolio managers faced significant challenges as AI stocks plummeted in June, learning critical lessons early in their careers. The rapid market decline exposed rookie managers to the harsh realities of their roles. Even seasoned investors had to address concerns from clients as their portfolios suffered substantial losses.

Yuan Zeqiang, a relatively inexperienced manager with three and a half years of experience, inherited two portfolios at Caitong Fund Management, each with combined assets of 7.28 billion yuan ($1.08 billion) at the end of the second quarter. Within 46 days, his portfolios plummeted by 36 percent and 33 percent respectively, driven by heavy exposure to tech stocks that had retreated after earlier gains in the year.

Wu Dongdong, who had been with Fullgoal Fund Management for four years, faced a similar scenario when his newly launched portfolio dropped nearly 34 percent within 46 days of its June 16 launch. Meanwhile, Hengyue Fund Management's Wu Haining switched focus to AI hardware upon taking charge on June 10, resulting in her fund's decline by nearly 41 percent in 55 days as top holding GigaDevice slumped more than 40 percent over 20 trading days.

These situations highlight the challenges faced by China's young portfolio managers, who not only need to navigate volatile markets but also contend with the lack of funding support that could cushion their drawdowns. Their concentrated bets left little room for error.

Across the Pacific, US hedge fund manager Leopold Aschenbrenner also faced the consequences of the AI stock plunge. His Situational Awareness LP saw a staggering 67 percent drop in July, prompting him to offload $16 billion in leveraged public equity positions to Citadel to prevent total liquidation.

The volatility has led to a surge in portfolio adjustments across China's market. Over 1,000 portfolio managers have resigned from funds this year, while 768 have registered, according to Securities Daily. China's securities investment has experienced explosive growth in recent years, with net assets managed by mutual funds reaching 39.7 trillion yuan by the end of June, up from 34.4 trillion yuan a year earlier.

The broader private fund industry has also seen a record 23.6 trillion yuan in assets under management, with an additional 1.5 trillion yuan added in the first half of 2026.

Written by urgent.news from SCMP Tech's reporting — not their text. Machine-written — it may contain errors, so check the original before relying on it.

This story

This is one outlet's version. Read the fullest account.

Read the original at scmp.com →

More in AI

Toast 1: A New Embedding Model That Rivals OpenAI at a Fraction of the Cost

Toast 1: A New Embedding Model That Rivals OpenAI at a Fraction of the Cost Mixedbread AI announced Toast 1, a new embedding model that claims to match or exceed OpenAI's text-embedding-3-large on…

  • Toast 1 is a new embedding model from Mixedbread AI
  • It rivals OpenAI's text-embedding-3-large at lower cost
  • Toast 1 supports 50+ languages with variable dimensionality

I Ran 4,200 Trials Testing LLM Agent Reliability. Here’s What Broke.

We know when an AI agent gets a response from a tool, getting a response back doesn’t necessarily mean that response should be trusted.

  • I conducted 4,200 trials to test LLM agent reliability.
  • ReliAgent identified signals for cautious response treatment.
  • Distinction between model behavior and detector failure crucial.

Maximizing Your Claude Code Sessions: 7 Tips from Anthropic's Own Engineers

Maximizing Your Claude Code Sessions: 7 Tips from Anthropic's Own Engineers Anthropic published a guide on maximizing the value of Claude Code sessions, and it hit the Hacker News front page with 130…

  • Provide clear project context before starting Claude Code sessions
  • Use CLAUDE.md file for persistent context and instructions
  • Break work into small, verifiable tasks for testing and error detection

Introducing Murya AI

Dear DEV Community, It feels great to be here, and thank you for the opportunity to be a part of this platform! This is my first post, and I am thrilled to share a recent project that kept me working…