Keep a close eye on one "variation point" in the second half of the year

Since mid-to-late July, the A-share and Hong Kong stock markets have undergone a sharp "high-to-low" rotation and deleveraging adjustment.
The AI computing power, optical modules, and semiconductor sectors, which had seen massive gains earlier, faced concentrated profit-taking. Market sentiment shifted rapidly between extreme euphoria and sudden panic, leaving many investors confused and anxious about "fear of missing out during rallies and fear of deep losses during downturns."
At this critical juncture, Huaxia Fund, a leading institution, held its "Active Investment Research 5.0 Launch and Mid-Year Investment Strategy Conference" on July 17. As a semi-annual statement from a top buy-side institution, its analysis often reflects the consensus among mainstream institutions. Kanjian Finance has sorted out the core logic of this strategy conference to help you capture market signals worth watching.

Investment Research Trends: From "Relying on Stars" to "Competing on Systems"

The current market environment has fundamentally changed: AI is restructuring industrial landscapes, and globalization is reshaping asset pricing logic.
Huaxia Fund's upgrade to its investment research system reveals a key shift in the public fund industry: using organizational capability to counter the uncertainty of individual investing.
Three core keywords deserve close attention:
Internationalization: The team sends personnel to root themselves in overseas industrial frontlines monthly to conduct cross-market cross-validation research;
Platformization: Seven investment research groups collaborate horizontally, abandoning the "myth of the all-powerful single fund manager";
Ecologicalization: Strictly constraining investment style drift, extending the evaluation cycle to three years, and avoiding interference from short-term market noise.
The core purpose of these upgrades is singular: In an increasingly complex market, a platform-based investment research system is far more stable and enduring than a single "genius player."

Macro Outlook: Focusing on the Market's "Variation Points" in the Second Half of the Year

Renowned economist Liu Yuhui's view is quite straightforward: The core theme of the current global capital market is solely AI.
The Philadelphia Semiconductor Index saw a maximum gain of up to 90% since April. Behind this extreme divergence is the two-way collision between the ebb of the US dollar tide and the super-cycle of the AI industry.
Zhang Jingsong of Huaxia Fund proposed the core investment proposition for the second half of the year: finding the market's "variation points." Before the variation point arrives, tech growth remains the main market theme; after the variation point lands, market styles may switch to defensive modes.
To precisely capture the variation point, focus on observing four core signals:
1. The pace at which expectations for Federal Reserve rate cuts are realized
2. The degree of valuation bubble in the AI sector
3. Whether the commercialization growth rate of the AI industry can continue
4. Whether price hikes in the industry chain suppress corporate capital expenditures
In one sentence: The AI track has not yet entered a comprehensive bubble stage, but caution is needed regarding the high probability of a market style switch in the fourth quarter.

Allocation Strategy

Main Line 1: AI Hardware, Sustained "Price and Volume Surge"
Fund Manager Li Xiangjie disclosed a set of heavy-hitting industry data: Global AI market total expenditure is expected to reach $3.49 trillion by 2027.
Industry prosperity continues to explode, with AI Token call volumes surging 60 times in half a year.
The current investment focus remains on the core hardware tracks of the AI 下半场 (second half):
Chip Layer: GPUs, High Bandwidth Memory (HBM), semiconductor equipment;
Infrastructure: Optical communication, PCBs, power supporting facilities.
From a valuation perspective, leading cloud vendors correspond to a PE of about 20 times, while the overall profit growth rate of the semiconductor industry remains at 25%-30%. Current valuations can be effectively digested by earnings growth.
The industry characteristics are very clear: The AI application layer is still in the exploration phase and has not yet produced an absolute leader. The certainty of performance and robustness of logic in the hardware layer are far ahead.
Main Line 2: Semiconductors, Moving from "Passive Substitution" to "Independent Innovation"
Fund Manager Gao Xiang reminded the market: Do not just stare at short-term price hike trends. Cyclical price hikes are only 阶段性 (phased) trends; full-stack independent innovation is the true long-term core theme of semiconductors.
The domestic semiconductor industry is currently undergoing a qualitative change: moving from past single-point passive import substitution to "chip-model synergy," which is an industrial ecosystem closed loop of domestic chips + domestic large models developing in tandem.
Industry space continues to be upgraded. The global semiconductor market size, previously predicted to be $1 trillion by 2030, has been revised up to $1.5 trillion by 2026.
Core investment logic: Enterprises that first achieve efficient training of domestic large models adapted to domestic chips will lock in the core valuation anchor for the next round of semiconductor rallies.
Main Line 3: Storage Super-Cycle, Breaking Historical Cycle Ceilings
Fund Manager Lv Jiawei pointed out that this AI-driven storage cycle is completely different from previous ones:
Producing 1 unit of HBM requires consuming 3-4 times the capacity of ordinary DRAM. HBM demand, accounting for only 10% of the industry, actually occupies 30% of industry capacity.
Coupled with the long construction cycles of wafer fabs and rigid upstream equipment supply, this AI manufacturing upcycle is likely longer than any historical cycle.
Besides the storage track，细分 segments such as electronic cloth and MLCC, which also face capacity expansion bottlenecks, possess long-term allocation value.
Main Line 4: Innovative Drugs, Welcoming the "DeepSeek Moment"
Fund Manager Wang Zeshi believes the innovative drug industry is facing dual benefits internally and externally:
Internally: Industry revenue growth has maintained high growth above 30% for consecutive years;
Externally: In the first half of this year, innovative drug BD (overseas licensing) transaction amounts broke through $100 billion, occupying half of the global market.
Relying on the abundant engineer dividend in China, the global market share of Chinese innovative drugs is expected to continue rising to 35%. This is a long-term growth opportunity for a trillion-scale track advancing to a six-trillion market cap.

Institutional Consensus: Tech is the Spear, Low-Valuation Leaders are the Shield

In the roundtable discussion, leading institutions reached several core consensuses:
1. The global attractiveness of domestic AI assets continues to rise. The proportion of global calls for domestic large models has climbed from 30% to 70%, and current foreign holdings remain underweight;
2. AI industrialization implementation continues to accelerate. Except for lithography machines, the substitution process for domestic semiconductor equipment is speeding up. End-side AI demand is exploding, with AI glasses sales growing 130% year-on-year, establishing a long-term logic;
3. Pay attention to market "ballast stone" assets: Leaders in lithium batteries, CXO, construction machinery, and internet sectors show steady growth and low valuations. Some leaders' valuations for 2027 are less than 15 times, offering extremely high safety margins.
Overall, the A-share market in the second half of the year is not a simple yes/no question of rising or falling, but a complex choice of rhythm.
For ordinary investors, accurately identifying the market's "variation points" is far more important than 纠结 (dwelling on) whether to be long or short. Tech growth is the spear for offense, while low-position, low-valuation core leaders are the shield protecting returns.
[Disclaimer] This article is compiled based on publicly available strategy conference content and is for viewpoint sharing only. It does not constitute any investment advice. The market involves risks, and investment requires caution.