Smartphone makers are rolling out flagship models in rapid succession, with on-device AI computing power emerging as the primary battleground and local inference capabilities becoming the new yardstick for user experience. Around the same time, the 2026 China Computing Power Conference concluded successfully, signaling that artificial intelligence is shifting from technology narratives to tangible value creation.

As the compute infrastructure foundation strengthens and edge applications accelerate, momentum is building in unison across the upstream, midstream, and downstream segments of the industry chain. Market attention on AI is gradually transitioning from "what technology can do" to "what value technology is creating." This subtle shift is fundamentally reshaping the investment logic for the AI sector, requiring a fresh evaluation of how AI companies should be assessed and how portfolios should be positioned. The release of industrial value is no longer tied to a single link, but rather depends on the coordinated evolution of the entire supply chain—a core distinction that sets this AI narrative apart from previous cycles.

Industrial Value: From Technological Breakthroughs to Commercial Deployment

From the investment perspective, the expansion potential of the industry chain remains well-supported. CITIC Securities notes in its research that during the 15th Five-Year Plan period, computing power networks are expected to become critical new infrastructure supporting the growth of new quality productive forces and absorbing expanded effective investment. Domestic computing power is poised to carry both industrial product and public utility attributes, likely becoming a vital driver of the next round of industrial upgrading and capital formation. This AI computing cycle features both "long-duration demand and slow supply," suggesting the boom period could extend longer than typical cycles.

ICBC Credit Suisse Asset Management also highlighted in recent market commentary that global compute capital expenditure still has further room to expand. Continued increases in capital spending by overseas internet giants will boost overall demand across the supply chain, with long-term growth prospects worth monitoring. The differences and linkages among various segments of the chain offer a key to understanding AI's industrial value. As a technology moves from breakthrough to widespread adoption, its value does not remain confined to any single link but cascades progressively along the supply chain. Upstream compute investment serves as the prerequisite, midstream model capabilities act as the engine, and downstream scenario deployment provides the outlet—together, their coordinated evolution paints a complete picture of an industry maturing. This means the transformation of AI value is closer to a gradual, point-to-surface, upstream-to-downstream process than an isolated burst in one segment. Industry dividends will rotate and transmit across different links over time.

A more rational approach to understanding the AI sector may not involve betting on a single company, but rather placing one's perspective across the full depth of the industry chain. An individual stock may stand out in one segment, yet it is difficult for it to fully capture the entire dividend of industrial evolution. In contrast, the overall upward movement of the supply chain is likely to be more sustainable than a single-point breakthrough. The holistic nature of the industry chain serves both as the defining feature that distinguishes this sector from general themes and as the foundation for assessing its value from a long-term perspective. As the narrative shifts from technology to implementation, the investment view should also shift from "point" to "surface." This does not negate the value of individual stocks, but rather adapts to the current stage of the industry. In a technology diffusion phase, breadth may offer greater certainty than depth.

Index Value: Complete Coverage of the Industry Chain from Upstream to Downstream

When industrial value spreads along the chain, the difficulty of picking individual stocks rises correspondingly, giving index-based tools a clear advantage. The CSI STAR & CHINEXT Artificial Intelligence Index (932456) selects 50 listed company securities from the STAR Market and ChiNext Board whose businesses involve AI foundational resources, AI technologies, and AI applications, spanning upstream compute chips, optical communications, and downstream AI application software. Single-stock weight and sector weight constraints ensure the index neither loses representativeness due to excessive concentration in any individual sample nor compromises structural stability at the sector level, thereby providing a relatively complete reflection of the industry's overall landscape.

The combination of the STAR Market and ChiNext Board endows this index with distinctive sample characteristics. These two boards bring together AI enterprises that combine hard-tech attributes with growth elasticity. The STAR Market component covers the domestic substitution chain focused on semiconductor chips, while the ChiNext component includes the overseas AI supply chain centered on communications equipment. The index encompasses both the hardware segments supporting the compute foundation and the software and scenarios that carry application deployment. This end-to-end coverage across upstream, midstream, and downstream makes the index itself a relatively complete map for observing and participating in the AI industry. Investors no longer need to repeatedly choose among numerous sub-sectors; through a basket of samples, they can share in the fruits of the entire supply chain's evolution, while conveniently accessing both the overseas AI leadership chain and the domestic substitution chain—a dual-engine approach to seizing AI-era investment opportunities. The index's sample composition and periodic adjustment mechanism also continuously calibrate to keep coverage aligned with the true trajectory of industrial evolution.

The deeper value of such a tool lies in returning the judgment to investors. Chasing a single hot spot demands precise timing and stock selection, while an index that spans the depth of the industry chain allows investors to focus primarily on their confidence in the sector's long-term direction. When industrial complexity exceeds the margin of error of individual stock research, the "simplification" offered by index-based tools itself becomes a scarce value. It reduces not only the barrier to entry but also the cognitive burden. Retail investors lacking extensive research resources and tolerance for error may benefit most from this approach.

Allocation Value: Delegating Judgment to Long-Term Industry Trends

ICBC Credit Suisse's STAR & CHINEXT AI ETF (588430), anchored to the STAR & CHINEXT AI Index, provides investors with a relatively direct participation path. From a valuation perspective, Wind data shows that as of September 21, 2026, the CSI STAR & CHINEXT Artificial Intelligence Index's price-to-earnings ratio (TTM) stood at 75.17 times, significantly narrowed from previous highs and positioned in the low 5.99% percentile range since the index's launch on May 14, 2025. The STAR & CHINEXT AI ETF (588430), which closely tracks this index, carries a management fee of 0.45% per annum and a custody fee of 0.10% per annum, offering a certain cost advantage that helps mitigate the hidden erosion of costs during long-term holding.

To meet the allocation needs of off-exchange investors, the ICBC Credit Suisse STAR & CHINEXT AI Feeder Fund (Class A: 029101, Class C: 029102) will begin public offering on October 8. The product uses the STAR & CHINEXT AI ETF as its target ETF, investing no less than 90% of fund assets in the target ETF to closely track the underlying index. With an annual management fee of only 0.45%, off-exchange investors can also share the low-cost advantage. Supporting the day-to-day operations of these products is ICBC Credit Suisse's long-standing expertise in index investing. Since managing ETFs in 2009, the company's Index and Quantitative Investment Department has built a professional and stable investment research team, with index products extending across broad-based, sector-thematic, Hong Kong-listed, multi-asset allocation, and index-enhanced directions, covering different asset classes from onshore to offshore. In terms of management, the team places tracking error control at the core, relying on a well-developed ETF investment management process and risk control system to achieve end-to-end management covering pre-trade, during-trade, and post-trade stages, supported by a market-making system comprising multiple securities firms to ensure product liquidity and stable operations.

As the global semiconductor industry enters an expansion cycle, substantial increases in memory and advanced process capacity will translate into massive demand for semiconductor equipment, potentially bringing a multi-year order boom to the equipment supply chain. Meanwhile, global semiconductor equipment supply remains tight, with delivery times significantly extended. Domestic equipment and upstream component companies are expected to see accelerated order growth in the second half of the year. The evolution of artificial intelligence continues at full throttle, yet the realization of industrial value rarely happens overnight, making periodic market volatility inevitable. For investors seeking exposure to the depth of the AI industry chain while managing individual stock risk, index-based tools offer a relatively rational and sustainable participation path. Participating in batches over the long term through indexation, delegating judgment to the industry's long-term trends, may suit the characteristics of this track better than chasing short-term hot spots. Rather than agonizing over short-term fluctuations, it is more meaningful to anchor confidence in industrial value toward the long-term direction of technological evolution and commercialization progress.

Note: The STAR & CHINEXT AI Index recorded annual returns of 1.95%, -35.33%, 29.17%, 36.76%, and 103.09% from 2021 to 2025, according to Wind data. Historical index performance does not predict future results, does not represent fund performance, and does not constitute a guarantee of fund returns. Fee details: 1. For the STAR & CHINEXT AI ETF, on-exchange trading fees are subject to actual charges by securities firms. When subscribing or redeeming fund shares, the subscription/redemption agent broker may charge a commission of up to 0.5% of the amount, including relevant fees charged by the stock exchange and registration institution. This fund carries an annual management fee of 0.45% and an annual custody fee of 0.10%. 2. For the ICBC Credit Suisse STAR & CHINEXT AI Feeder Fund: Class A subscription and purchase fees (front-end): no subscription or purchase fee is charged for Class A shares subscribed through direct sales channels; for subscriptions or purchases through non-direct sales channels, if the subscription or purchase amount M is less than 5 million yuan, the fee rate is 0.30%; if M is 5 million yuan or above, the fee is 1,000 yuan per transaction. Class A redemption fees: for individual investors, if the holding period Y is less than 7 days, the fee rate is 1.50%; if Y is 7 days or more, the fee rate is 0.00%. For other investors: Y less than 7 days, 1.50%; 7 days ≤ Y < 30 days, 1.0%; 30 days ≤ Y < 180 days, 0.5%; Y ≥ 180 days, 0.00%. Class C redemption fees: for individual investors, Y less than 7 days, 1.50%; Y ≥ 7 days, 0.00%. For other investors: Y less than 7 days, 1.50%; 7 days ≤ Y < 30 days, 1.0%; 30 days ≤ Y < 180 days, 0.5%; Y ≥ 180 days, 0.00%. This fund has an annual management fee of 0.45%, an annual custody fee of 0.10%, and a Class C sales service fee of 0.20% per annum. Class A shares do not incur a sales service fee, and Class C shares do not incur subscription or purchase fees.

Risk disclosure: The views expressed are for reference only and are time-sensitive. They do not constitute investment advice or return commitments, nor do they represent the fund's specific future allocation direction. The fund manager operates and manages fund assets in accordance with the principles of diligence, honesty, credibility, and prudence, but does not guarantee that the fund will be profitable or that minimum returns will be achieved. The STAR & CHINEXT AI ETF (588430) and the ICBC Credit Suisse STAR & CHINEXT AI Feeder Fund are equity funds, carrying higher risks and returns than hybrid funds, bond funds, and money market funds. The STAR & CHINEXT AI ETF is an index fund that primarily employs a full replication strategy to track the performance of the underlying index, exhibiting risk-return characteristics similar to those of the underlying index and the stock market it represents. Investing in the ETF involves specific risks including underlying index volatility, deviation of fund portfolio returns from index returns, and other specific risks. The ICBC Credit Suisse STAR & CHINEXT AI Feeder Fund is an ETF feeder fund that tracks the underlying index's performance by investing in the target ETF, exhibiting risk-return characteristics similar to those of the underlying index and the securities market it represents. Investing in the ETF feeder fund involves specific risks including underlying index volatility, deviation of fund portfolio returns from index returns, and performance differences from the target ETF. Investing in equity funds involves significant return volatility risk. Funds carry risk. Before investing, investors should carefully read the Fund Contract, Prospectus, Fund Product Information Summary, and updated legal documents. On the basis of fully understanding the product's features, fee structure, charging standards of various distribution channels, and the suitability advice of sales institutions, investors should select investment products appropriate to their own risk tolerance. Fund investment requires caution. MACD golden cross signals are forming; these stocks are trending upward!