Missing Out on the AI Frenzy and the Retreat of Trillion-Yuan ETFs: The Cost and Logic of Ningquan Asset's Commitment to the "Anti-Consensus" Strategy
```
During the highly polarized market of the first half of 2026, Ningquan Asset Management, a private equity firm with assets exceeding 10 billion, encountered its most severe challenge since its founding. The company’s strategy of sticking to undervalued value stocks and avoiding popular AI sectors caused this institution, renowned for "deep value," to pay a heavy price in a round of unilateral tech bull market.
Recently, Ningquan Asset Management released “A Letter to Holders: 2026 Semi-Annual Investment Report”, publicly apologizing to holders, admitting that this year's performance fell short of expectations, and disclosing that the June product drawdown matched the largest historical bear market drawdown, saying, "Such an event has never occurred in previous bull markets."
Despite performance pressure, Ningquan not only did not increase holdings of hot AI stocks but explicitly warned of a super-sized bubble in related sectors, predicting that a large number of popular stocks "are highly likely to fall by eighty or even ninety percent in the future."
This stance triggered strong reactions in the market. Critics believe that a fund manager unable to accurately judge the beginning of the tech cycle is not qualified to comment on the cycle itself; supporters compared Ningquan's viewpoint to risk warnings in 2007 and 2015, believing its contrarian judgment holds reference value.
At the same time, the controversy also reflects a deep contradiction in the A-share market in the first half of 2026: in a bull market with record trading volume, the majority of investors are still losing money.
Widespread Losses in a Bull Market
Wind data shows that the average daily turnover in A-shares in the first half of 2026 reached 2.74 trillion RMB, far exceeding the 2025 yearly average of 1.73 trillion RMB, with no doubt about the high level of capital activity.
However, as of June 30, out of 5,526 stocks in the entire market, the median price change was -15.4%, and even the Shenwan Securities Index, known as a "bull market indicator," fell by 7.8%. The divergence between abundant liquidity and the absence of wealth effect confused the market.
The roots of this fragmentation point to two major unforeseen variables: massive large-institution selling and the powerful driving force of the US tech cycle centered on semiconductors.
The combination of these two factors created an extremely concentrated market, with semiconductor and optical module sectors nearly doubling, while other sectors exhibited typical bear market patterns despite a bull market level of trading volume.
Two Major Variables Tearing Apart the Market
The pressure on capital is clearly visible in the changes of broad-based ETF holdings.
The Huatai-PB CSI 300 ETF once exceeded 100 billion shares in 2024, then began to shrink rapidly after entering 2026: in January, it quickly fell below 50 billion shares, and after the semiconductor rally in April, continued downward, now dropping below 20 billion shares.
Market estimates suggest that large institutions reduced their holdings by an aggregate of about 1 trillion RMB in the first half of 2026, persistently withdrawing and significantly suppressing the broad market.
The other mainline comes from across the ocean. The US Smart Semiconductor Index rose 120% in the first half, the Philadelphia Semiconductor Index rose 101%, and the Global Semiconductor Index rose 110%.
Explosive growth in AI computing power demand and a 2–3 year wafer fab expansion cycle have created a huge supply-demand gap, with the memory industry chain benefiting first. From April 2026, US and A-share semiconductor sectors resonated, and related stocks nearly doubled universally.
The strength of this resonance effect led to a concentration of funds from the entire market, causing what was once considered “defensive base” dividend indexes to experience a 12% drawdown in June alone—something rare in bull markets historically, directly illustrating the structural extremity of this round of the market.
The “Logic of Holding On” for Locked-in Funds
The high market concentration also reflects a structural predicament for existing funds.
Funds that bought consumer and new energy sectors at highs during the 2020–2021 bull market are still largely stuck. These funds have not been unlocked, cannot chase the semiconductor rally, nor bring fresh capital to the market, making 2026’s bull market participation relatively limited across society, with no obvious signs of money moving out of deposit accounts.
Investors staying in consumer sectors are not without logic. Kweichow Moutai’s PE is about 18x, Shanxi Fenjiu about 12x, Yili about 13x—valuations are at historically low levels against massive bull market trading volumes.
This group chooses to hold on, waiting for a style rotation. However, there has been no rotation nor switching among sectors all half year.
Losses in the first half triggered a significant shift in investor behavior.
More market participants are shifting their research focus from A-share fundamentals to US stock indices, using the Philadelphia Semiconductor Index as the core reference for operation in semiconductor and optical module stocks—if the index rises, they chase relevant stocks; if it falls, it signals the end of good news and they exit.
This shift has a practical reason. Targets such as optical modules, MLCC, PCB, long questioned as labor-intensive industries, have rallied alongside US tech narratives in this round, making fundamental analysis considerably less valuable.
Ningquan’s Allocation Logic: Internet Giants and Value Depressions
Ningquan Asset Management's semi-annual report details its holdings and decision logic in the first half.
In the tech sector, the company was not completely absent, but gradually bought leading internet firms at cheap valuations from 2022 to 2024. The core logic: Internet giants combine computing power, AI technical reserves, and application scenario advantages, plus stable cash flow to support long-term capital expenditure.
However, their timing was off. During the “DeepSeek Moment” in the first half of 2025, these stocks performed well but then kept weakening. Ningquan expressed confusion in the letter:
Chinese internet giants, with market caps far smaller than their American counterparts, nevertheless moved in completely opposite directions. That truly confounds us.
On the AI infrastructure theme, Ningquan admitted to “severely underestimating the enthusiasm for this theme,” and gave reasons for not following: related manufacturing firms have average business models, disputable long-term moats, and must keep spending capital for growth.
The company reflected that “just a surge in demand alone bringing so much prosperity, reaching such high valuations and market caps—we really didn’t see that coming.”
Currently, Ningquan’s holdings concentrate in two directions:
First are undervalued, high dividend, cash flow-positive companies, covering telecom operators, home appliances, power, finance, chemicals, property services, etc.;Second are some deeply depressed cyclical industries, including leading developers and related construction material leaders, and solar panel leaders.
It’s noteworthy that Ningquan currently manages over 50 billion RMB; its holding choices and public statements have an impact on the market. Whether its heavy allocation to undervalued sectors can see valuation restoration, and whether the AI infrastructure sector will experience sharp adjustments as Ningquan warned, will be the core variables testing this contrarian logic.
Risk Warning and DisclaimerThe market is risky; investment requires caution. This article does not constitute personal investment advice and does not take into account the specific investment objectives, financial situation, or needs of individual users. Users should consider whether any opinions, viewpoints, or conclusions in this article fit their own situations. Invest at your own risk. ```