Ningquan Asset's H1 2026 apology letter reveals a 500-billion-yuan fund manager betting against the market's biggest winners — and warning they could lose 80% of their value.
Ningquan Asset's H1 2026 apology letter reveals a 500-billion-yuan fund manager betting against the market's biggest winners — and warning they could lose 80% of their value.

Ningquan Asset's H1 2026 apology letter reveals a 500-billion-yuan fund manager betting against the market's biggest winners — and warning they could lose 80% of their value.
Ningquan Asset, a 500-billion-yuan Chinese private equity firm, apologized to investors after avoiding AI stocks cost it record losses in H1 2026 — even as daily turnover hit 2.74 trillion yuan.
"The AI theme has become a super-sized bubble, and many popular stocks will likely fall 80% to 90% or more," Ningquan said in its semi-annual letter to holders, published this week. The firm manages more than 50 billion yuan in assets.
The warning comes amid one of the most lopsided bull markets in A-share history. Daily average turnover surged to 2.74 trillion yuan in the first half, up from 1.73 trillion yuan in all of 2025, yet the median return across 5,526 stocks was negative 15.4 percent. Semiconductor and optical module stocks nearly doubled, while the Shenwan Securities Index — a traditional bull-market bellwether — fell 7.8 percent.
The divergence reflects an estimated 1 trillion yuan in institutional ETF redemptions during the period, according to market estimates. The Huatai-PineBridge CSI 300 ETF, which once held more than 100 billion units in 2024, has shrunk to below 20 billion units. If Ningquan's bearish call on AI proves correct, a correction in the overheated sector could trigger a sharp rotation into value stocks — but if the rally continues, the fund faces deeper redemption pressure.
The Two Forces That Split the Market
Two super-expected variables drove the divergence, Ningquan said. The first was large-scale institutional selling: an estimated 1 trillion yuan in ETF redemptions by major institutions during the first half, draining liquidity from broad-market indices. The second was a powerful AI-driven tech cycle originating in the United States. The US Smart Semiconductor Index surged 120 percent in H1, the Philadelphia Semiconductor Index gained 101 percent, and the Global Semiconductor Index rose 110 percent. Starting in April, A-share semiconductor stocks began tracking their US peers in near-lockstep, with related shares roughly doubling.
The concentration was so extreme that the CSI Dividend Index — typically a defensive anchor — fell 12 percent in June alone, a drawdown virtually unheard of during a bull market. The dividend index's collapse illustrated how deeply capital was being pulled into the AI trade at the expense of every other sector.
Why Ningquan Sat Out the Rally
Ningquan's portfolio is concentrated in two areas: low-valuation, high-dividend companies across telecom, consumer appliances, power, finance, chemicals, and property services; and deeply depressed cyclical sectors including property development, building materials, and solar manufacturing.
The firm acknowledged it "seriously underestimated the heat of this theme" but defended its decision not to chase AI infrastructure stocks. The manufacturing companies benefiting from AI demand have weak business models and questionable long-term moats, Ningquan argued, while requiring continuous capital expenditure to sustain growth. "Just one round of massive demand-driven prosperity can push valuations and market caps this high — we simply couldn't make the math work," the firm said.
Ningquan did participate in AI through a different route, buying internet giants at low valuations between 2022 and 2024. The logic: these companies combine computing capacity, AI technology, application advantages, and stable cash flows to support long-term capital spending. But the timing misfired. While these stocks performed well during the "DeepSeek moment" in H1 2025, they have since weakened. "Chinese internet giants, with market caps far smaller than their US peers, have moved in the completely opposite direction — this genuinely puzzles us," Ningquan wrote.
The Stuck Capital Problem
The market's structural divide also reflects a deeper issue: trapped capital from the 2020-2021 bull market. Investors who bought consumer and new-energy stocks at peak valuations remain underwater, unable to rotate into semiconductors or provide fresh liquidity to the market. This explains why the 2026 bull market has seen limited retail participation and no significant deposit migration from bank accounts.
Those holding consumer stocks have their own logic. Kweichow Moutai trades at about 18 times earnings, Shanxi Fenjiu at about 12 times, and Yili at about 13 times — valuations that are historically low by bull-market standards. They are waiting for a style rotation that has yet to arrive.
The standoff has changed how many A-share investors operate. Rather than analyzing Chinese fundamentals, a growing number now treat the Philadelphia Semiconductor Index as their primary trading signal for A-share semiconductor and optical module stocks: buy when Philly Semi rises, sell when it falls. The shift reflects a market where fundamentals have become secondary to the US tech narrative.
Ningquan's anti-consensus bet now faces its defining test. If the AI bubble bursts as the firm predicts, its value-heavy portfolio could see a dramatic recovery. If the tech rally continues to broaden, the fund may face accelerating redemptions from investors unwilling to wait.
This article is for informational purposes only and does not constitute investment advice.