Focus on the “right and wrong” in the AI market: first-line private funds reshaping strategies for the second half of the year

Author: Wang Hui

The performance in China’s A-share market has been polarized in the early period. On one end, AI and the semiconductor industry chain shine brightly, becoming the main narrative that the market has most intensely played out; on the other end, more than half of industries have seen declines reaching double digits, and many individual stocks are facing a pronounced liquidity “suction effect.” This polarization saw a wave of correction at the mid-year point—since July, whether in overseas stock markets or in A shares, AI-related sectors have followed up with adjustments, while some non-AI sectors that had been under pressure earlier have rebounded rapidly.

Have the signals for a turning point in the shift of market style shown up? Standing at the threshold of the second half of the year, multiple top-tier private fund managers have begun to reconstruct their investment strategies.

**　　AI rally: growing disagreement**

Regarding the long-term outlook of the AI industry itself, almost no private fund interviewees are bearish. But on the core question of how to understand the nature of this round of AI rally, the divide is becoming increasingly clear.

In Zhongjing Asset Management partner and CEO Gao Yuncheng’s view, the deepest change in global capital markets over the past half year is that “intelligence is for the first time starting to be scaled and mass-produced by machines”—a genuine paradigm shift, not merely cyclical fluctuations. AI no longer exists only as a software tool. From this, he argues that the light-asset platform valuation framework built in the past around user numbers, traffic, and advertising revenue needs to be re-examined. As a core asset of the new era, those high-quality companies in the AI “five-layer cake” (energy and infrastructure, semiconductor design, manufacturing and core components, large models and operating systems, cloud platforms and entry points, industrial applications) will be redefined. Gao Yuncheng judges: “We may be standing at the starting point of intelligent industrialization. In such an era, the biggest risk is often not volatility, but still understanding the new world using the frameworks of the old world.”

At the same time, not all interviewees are willing to seamlessly connect today’s market enthusiasm with this grand vision.

Zhaoyang, fund manager at Rongyang Investment, believes that this round of AI rally is driven by a surge in real demand and industry companies’ earnings growth. This is fundamentally different from the high-valuation hype of many unprofitable companies during the 2000 internet bubble period. Instead, it is more like the new energy rally in 2021. “But if the trend reverses, most likely it will also be because growth on the demand side can’t sustain the current steep slope for the long term. After supply-demand balance is reached, further declines will appear in the supply-side landscape.” He said directly: “You can’t expect a tree to grow to the sky.”

Liang Li, general manager of Yunji Asset, warns of a disconnect between industry fundamentals and market pricing. He believes that “industry fundamentals” and “market pricing” should be viewed separately. “The fundamentals of the AI technology industry are still very strong,” there’s no doubt about that. But the problem lies precisely in pricing. The market has already priced in in advance the most optimistic scenario, while overlooking two things: first, the high profitability of many segments essentially comes from demand arriving too fast and supply failing to keep up; it can’t be maintained indefinitely; second, valuations of many companies are “too full of imagination.” In the past one to two years, the contribution of some companies’ related businesses to overall earnings is almost negligible, yet the market assigns them high valuations using very long-term, very optimistic outlooks.

For the outlook of the AI rally, it is not a debate between being bullish or bearish. Rather, it’s different understandings of the relationship between long-term vision and short-term pricing. Just as the broad-based advance in the first half of the AI sector has gradually transitioned into the sharp polarization in July, some private fund practitioners say that the stage where buying AI-related targets guarantees profits has already passed. The upcoming technology market will no longer be a “the strong always stay stronger” race across the board; it will be a much harder “gold-panning” contest.

**　　Market style: moving toward rebalancing**

If disagreement over the AI main theme forms the defensive thinking for some institutions in the second half, then the consensus on style rebalancing points to the direction for the next offensive phase.

The observation by Xia Jun, investment director at Shicheng Investment, is quite representative. He said that polarization in the market reached an all-time high level in the first half, but the firm’s view—“extreme polarization will converge rather than continue to expand”—was preliminarily validated in July. This view was echoed by multiple interviewees. Fang Lei, deputy general manager of Starstone Investment, said that the valuation divergence between technology and non-technology sectors is currently at a historical extreme, and on top of that, trading overcrowding in the technology sector is relatively high. The marginal flow of funds is likely to push the market style toward rebalancing. “In the second half of this year, market style has the chance to become more balanced. This balance will be more sustainable than the extreme divergence of the first half.”

Zhaoyang compares the two ends of the market to “standing in the light” and “old-un” assets, respectively. He believes that real opportunities lie in the middle ground between the two. In his view, when the spotlight on AI obscures everything else, a group of companies with genuine growth but that have not yet been pursued in this round of the rally are making the “impossible triangle” in stock investing (good industry, good company, good price) unusually possible. “This is the key direction that should be dug into in the next stage.” Zhaoyang further stated that pan-technology manufacturing is the direction with the most opportunities. China’s manufacturing strengths are not only reflected in the AI field, but also in many sub-sectors. Also worth noting is that the pharmaceutical industry is “back in the batter’s box.”

Liang Li looks toward an even broader universe. He revealed that his holdings are mainly concentrated in the internet, consumer, manufacturing, and cyclical industries. In these industries, there are many excellent companies whose earnings have reached new highs. Their business models are immune to AI; meanwhile, they can also reduce costs or increase business value through the application of AI technologies. However, “their stock prices have shown a trend completely mismatched with their performance.” Behind this is AI narrative achieving absolute monopoly over market attention. “When AI becomes the only focus of the market, there will be more opportunities outside the spotlight.” Liang Li said that as the market becomes relatively rational about AI, and returns to searching for other investment opportunities, the returns brought by the revaluation of this group of companies “may be very significant.”

There is also a view from private funds that in the first half, the market was accustomed to pricing grand narratives, while sidelining those tickers that quietly created new highs in earnings. This mismatch is likely to become an important source of the market’s excess returns in the second half.

**　　Future strategy: more focused on balance and earnings**

Consensus and disagreements are converging, and the strategy for the future is reshaped accordingly. The specific second-half positioning by first-tier private funds is coming into view.

“Building a diversified portfolio is a pragmatic choice to deal with the current market conditions.” Xia Jun said that the market performance since July has validated a fact: extreme overcrowding in a single track contains fragility. He disclosed that the subsequent operations of Shicheng Investment will select targets within the technology direction, focusing on companies that truly have technological barriers and the ability to deliver earnings.

Fang Lei distilled the key points of the second-half strategy into: offense and defense with balance, and balanced allocation. He suggested using industry investments in high-urgency trends as one end leaning toward offense, and traditional core assets with high cost-effectiveness as the other end leaning toward defense. “For the stock market investment in the second half of this year, the main line should be corporate earnings.” Fang Lei believes that technology-type assets may only have a possibility of another burst after earnings have absorbed the valuations. Non-technology assets, given their lower valuations and the potential for fundamentals to improve at the margin, have higher allocation value.

Looking ahead to the second half, a private fund interviewee said: whether standing “in the light” or “outside the light,” the ultimate anchor of market pricing must return to corporate earnings as the ballast.

The “intelligent industrialization” scenario drawn by Gao Yuncheng is inspiring, and his portfolio is also built around long-term industry trends such as semiconductors and AI infrastructure. He admitted that historically, every technological revolution has gone through hot capital expenditure spending, valuation bubbles, and periodic pullbacks, and the AI industry chain will not be an exception.

In addition, from Zhaoyang’s perspective, although there truly is upside potential in many current hot areas, the market seems to have “forgotten many uncertainties under the extremely optimistic assumptions.” From the standpoint of stock-level cost-effectiveness, Xia Jun further cautioned that in the second half, research and allocation to previously severely oversold sectors will be increased, especially directions whose valuations have already built in a significant margin of safety. When extreme divergence begins to converge, many private fund practitioners’ thinking points in the same direction: the market ultimately needs to weigh valuations using performance and fundamentals, not only price them based on the grand narrative of the technology wave.

(Editor: Xu Nannan)

Keywords: