Recently, volatility in AI-related sectors has increased notably, and global tech assets have become more correlated. Within computing-power hardware subsectors such as optical modules, price swings have intensified further due to factors including demand expectations, pricing trends, technological iteration, trading crowding, and valuation digestion. The market has diverged in its expectations for the AI industry trend, capex intensity, sustainability of price increases, changes in ARR growth, and the pace of earnings realization.
Against this backdrop, the Huabao ChiNext Artificial Intelligence ETF (159363), which focuses on leading optical module companies, has seen rising on-exchange attention, with stronger willingness to allocate on dips. Regarding these changes, fund manager Cao Xuchen of the Huabao ChiNext Artificial Intelligence ETF (159363) shared his latest views across dimensions including the overall market, the technology industry, A-share structure, and product direction.
Overall judgment: The market may recover in stages, along a winding path
The recent market correction has been sizable, with major global markets weakening in sync. Cao Xuchen believes this performance has, to some extent, eased concerns that A-shares are decoupled from global markets; if global markets recover later, A-shares may also benefit. However, the strength of the recovery may not be simply linear relative to the prior decline—larger declines do not necessarily correspond to stronger rebounds. The market may subsequently show a rather winding, staged recovery rather than a one-way rise. Stage characteristics may include: an initial oversold rebound, followed by a recovery in macro expectations, and later stages depending on easing price-increase pressure and validation of industry logic. The core support for a staged recovery is that the AI industry is still developing; sequential growth may slow, but year-over-year growth may remain relatively fast. If such stages occur, they are more likely to be driven by solid earnings realization than by valuation expansion.
Tech view: The AI industry trend remains, but the driving logic may shift from the second derivative to the first derivative
On the tech sector, Cao Xuchen points out that AI may be an industrial transformation that surpasses the internet and lasts more than a decade, so long-term year-over-year growth may still be favorable. He mainly analyzes three aspects: ARR, price increases, and future market conditions.
(1) The AI coding field has seen breakthroughs and has driven hardware-side price increases. Recent public information shows that Anthropic's ARR (annual recurring revenue) sequential growth has slowed. If this trend continues, it may mean the AI investment logic is transitioning from being driven by the second derivative to being driven by the first derivative. The second-derivative phase often features price increases, capacity expansion, and broad sector gains; the first-derivative phase emphasizes steady growth, new product volume ramp-ups, and low-valuation recovery.
(2) The AI coding boom once brought price increases; currently, AI coding sequential growth has slowed somewhat, but prices in areas such as storage remain elevated. Demand expectations are declining while cost expectations remain high, which may compress earnings expectations. Regarding whether AI demand can continue to support a price-increase cycle, two observations are worth noting.
First, under optimistic estimates, if U.S. GDP in 2027 is about $35 trillion and AI capex is about $1.7 trillion, then AI capex would account for about 5% of GDP. Historically, a similar high capex share can be compared to the 19th-century railway construction period. Executives at related companies have also expressed the view that capital capacity may be difficult to fully match the potential of AI demand, so the industry's internal structure may face adjustments.
Second, long-cycle price increases are mostly concentrated in special assets such as gold and certain crypto assets, or in high-end consumer goods; sustained explosive price increases in industrial goods are more difficult. More likely, industrial goods experience 1–3 years of price increases during the second-derivative upcycle, and then take 6 months to 1 year to reduce prices and return to the starting point during the first-derivative cycle.
(3) Before previously price-increased categories see clearly significant declines, it may be difficult for the AI sector to experience a major trend-level move. The reason is that the prior stage's price-increase narrative should not be simply extrapolated linearly to the next stage; the next stage may shift toward using lower-cost AI computing power to drive upward earnings revisions at large-model and AI application companies. Whether this logic can become the mainline for a later stage still needs observation. The skepticism that "if storage prices do not rise, the AI industry will be harmed" may not hold: some terminal and consumer electronics companies may have already reflected cost-decline expectations to some extent, and their recent market performance has been relatively strong.
A-shares: This correction is closer to negative feedback under a liquidity shock
Cao Xuchen says that from a historical comparison, in periods such as 2000, 2004, 2016, and 2021, when the market fell, large-cap stocks usually declined less than the index; however, in the recent correction, large-caps fell relatively more, while small-caps were relatively less affected. This characteristic makes the current correction closer to passive redemption pressure under the liquidity shock in early 2024, rather than being purely fundamentals-driven, reflected as negative feedback formed by institutional redemptions and stop-losses. Policy continues to release signals to maintain the stable development of the capital market, so the market may be near a staged bottom area, but confirmation of the bottom still needs observation. After a rebound, there may still be a second round of digestion, and it also takes time for investor sentiment to shift from one-way net redemptions to one-way net subscriptions.
At the product level, it is advisable to continue to monitor allocation opportunities such as the Huabao ChiNext Artificial Intelligence ETF (159363), the Huabao Hong Kong Stock Connect Information Technology ETF (159131), and the Huabao STAR Market Chip ETF (589190). Earlier stages may be more about repairing fundamentals and macro expectations, while later stages will require new industry logic to determine the strength of the market. Over the long term, continue to focus on the value-creation ability of high-quality companies and high-quality industries.
To allocate to optical module leaders plus AI applications, it is advisable to monitor the Huabao ChiNext Artificial Intelligence ETF (159363) and its off-exchange feeder funds (Class A 023407, Class C 023408), focusing on optical module CPO leaders while also covering AI applications. The underlying index's combined weight in Zhongji Innolight, Eoptolink, and T&S Communications exceeds 35%, making it a core standard-bearer for AI computing power.
Data source: Shanghai and Shenzhen stock exchanges, Wind, etc. Note: As of 2026.9.30, according to the China Securities Index, the top three constituents of the ChiNext Artificial Intelligence Index are Eoptolink (weight 12.98%), Zhongji Innolight (weight 12.09%), and T&S Communications (weight 10.8%). Reminder: Recent market volatility may be relatively large, and short-term gains or losses do not indicate future performance. Investors must invest rationally based on their own funding conditions and risk tolerance, and pay close attention to position and risk management. ETF fund-related fee explanation: When investors subscribe or redeem fund shares, subscription and redemption agents may charge a commission of no more than 0.5%. On-exchange trading fees are subject to the actual charges of the securities company, and no sales service fee is charged. Feeder fund-related fee explanation: Huabao ChiNext Artificial Intelligence ETF Initiating Feeder C does not charge subscription fees; the redemption fee is 1.5% within 7 days and 0% on or after 7 days; the sales service fee is 0.3%. Huabao ChiNext Artificial Intelligence ETF Initiating Feeder A subscription fee is 1% for amounts below 1 million yuan, 0.6% for 1 million yuan (inclusive) to 2 million yuan, and 1,000 yuan per transaction for 2 million yuan (inclusive) and above; the redemption fee is 1.5% within 7 days and 0% on or after 7 days; no sales service fee is charged. According to the fund manager's assessment, the risk levels of the Huabao ChiNext Artificial Intelligence ETF, Huabao Hong Kong Stock Connect Information Technology ETF, and Huabao STAR Market Chip ETF are R4-medium-high risk, suitable for aggressive (C4) and above investors. Please refer to the sales institution for suitability matching opinions. Risk disclosure: The Huabao ChiNext Artificial Intelligence ETF passively tracks the ChiNext Artificial Intelligence Index. The index base date is 2018.12.28 and the launch date is 2024.7.11. The index's annual returns for 2021–2025 were: 17.57%, -34.52%, 47.83%, 38.44%, and 106.35%, respectively. The corresponding annualized volatility over the same period was 23.73%, 27.34%, 38.02%, 45.42%, and 41.1%. The index constituent composition is adjusted from time to time according to the index compilation rules, and its backtested historical performance does not indicate the index's future performance. The index constituents mentioned in the text are for display only; descriptions of individual stocks do not constitute investment advice in any form, nor do they represent the holdings information or trading actions of any fund under the manager. Any information appearing in this article (including but not limited to individual stocks, comments, forecasts, charts, indicators, theories, and any form of expression) is for reference only, and investors must be responsible for any investment decisions they make independently. In addition, any views, analyses, and forecasts in this article do not constitute investment advice in any form to readers, nor shall they bear any responsibility for direct or indirect losses caused by the use of the content of this article. Fund investment involves risks; past performance of a fund does not represent its future performance, and the performance of other funds managed by the fund manager does not constitute a guarantee of the fund's performance. Fund investment must be prudent. MACD golden cross signals have formed, and these stocks have performed well.
Comments