Friday (October 30) marked the final trading day of October for A-shares, with major indices posting collective declines for the month amid shrinking trading volume. The ChiNext Index fell more than 23% during the month, recording its second-largest single-month drop in history. Among the hardest-hit sectors were optical module heavyweights, with Eoptolink plunging over 49% for the month, T&S Communications dropping more than 23%, and Zhongji Innolight declining over 31%. Among popular ETFs, the ChiNext Artificial Intelligence ETF (159363), which focuses on optical module leaders, saw its underlying index tumble 27.99% for the month, setting a new record for July declines and hitting a year-to-date low, though investor attention remained elevated on dips.
The ChiNext Artificial Intelligence ETF's (159363) underlying index performance this year warrants a closer review of what the market is actually worried about. The trigger for this sharp selloff centered on the late-October earnings window for the "EZY" trio: what the market traded was not year-over-year growth rates, but the expectation gap created by quarter-over-quarter data and the disconnect between that and cloud providers' still-accelerating Capex. For investors focused on optical modules, this kind of correction driven by "timing misalignment" rather than "demand reversal" actually offers an observation window to pay attention on dips.
Q3 Earnings: Sequential Divergence, Not Demand Reversal
For the first three quarters, the "EZY" trio maintained strong year-over-year performance, but Q3 revealed a critical sequential divergence: Zhongji Innolight (ASX: 03308) posted Q3 revenue of 10.216 billion yuan, up 25.89% quarter-over-quarter, with net profit attributable to shareholders of 3.137 billion yuan, up 30.04% sequentially, representing the smoothest delivery cadence. The real expectation gap emerged at Eoptolink and T&S Communications — Eoptolink's Q3 revenue came in at 6.068 billion yuan, down 4.97% quarter-over-quarter, breaking its previous streak of consecutive sequential growth; T&S Communications reported Q3 revenue of 1.463 billion yuan, down 3.23% sequentially. The selloff logic was not complicated: leading valuations had previously implied a linear extrapolation of continued sequential acceleration each quarter, so once sequential growth flattened or slightly declined — even if year-over-year still doubled — it was collectively priced as "peak growth has passed." However, both companies' sequential weakness was non-demand-related: Eoptolink's was due to a mismatch between customer pickup timing and revenue recognition, possibly an "occasional impact from individual customers," with Q4 already recovering and Q3 net margin of 39.3% actually hitting a new single-quarter high; T&S Communications was dragged by EML and other material shortages affecting deliveries, compounded by asset impairment surging 238.76% year-over-year. Zhongji Innolight's sequential acceleration precisely disproves the demand-reversal thesis: demand has not turned, and the difference in leading players' delivery capabilities was amplified in Q3. Notably, the Q3 sequential divergence did not ferment in isolation: in October, the U.S. AI supply chain weakened in tandem, the Philadelphia Semiconductor Index suffered multiple sharp drops, and overseas optical module restriction rumors added disturbance — cross-market sentiment amplified the decline in A-share optical modules. Overseas factors may have served as a decline amplifier.
Demand Anchor: Capex Support and "Delivery Assurance"
Cloud providers' Capex (capital expenditure) is the "demand anchor" for optical module prosperity, locking in the demand floor. The volume anchor: the four major U.S. tech giants have updated their 2026 capital expenditure expectations during the Q3 earnings window, with their combined upper limit potentially exceeding $740 billion; their actual total capital expenditure in 2025 was $455.8 billion — based on the upper end of guidance, that is nearly $300 billion more in a single year, an increase of over 60%. As long as total Capex does not contract, the optical module demand pool will not shrink. The structural anchor: in AI clusters, GPUs determine how strong computing power is, while optical modules determine whether computing power can be connected. The larger the cluster and the higher the speed, the more optical interconnect volume and unit prices rise in tandem: from 400G to 800G to 1.6T, the value per module is increasing, and the quantity per rack is also rising. Optical modules' share of Capex is structurally rising with scale expansion and speed upgrades — every incremental unit of Capex translates to more than one unit of incremental demand for optical modules. The timing anchor: optical modules follow an "order-to-schedule, delivery-to-revenue-recognition" model, with delivery cycles far shorter than heavy-asset segments like IDC and power, meaning Capex changes show up first in their financial statements. This also explains why Q3 sequential fluctuations triggered such intense pricing — optical modules are the most sensitive "thermometer" for Capex; and Capex still accelerating is the most direct rebuttal to "peak growth" concerns.
Paying Attention on Dips: High "Optical" ETF Tools
In summary, the aforementioned Q3 sequential divergence points more to delivery timing misalignment and material constraints rather than demand reversal; the cloud providers' Capex "demand anchor" has not loosened, and industrial progress in 1.6T/silicon photonics remains unchanged. After the sharp decline, an expectation gap has formed between valuation clearing and unchanged demand, which conversely offers an observation window to pay attention on dips. To position in optical module leaders plus AI applications, consider the ChiNext Artificial Intelligence ETF (159363) and its off-exchange feeder funds (Class A 023407, Class C 023408), which focus on optical module CPO leaders while also covering AI applications. The underlying index's combined weight in "Zhongji Innolight + Eoptolink + T&S Communications" exceeds 35%, making it a core flagship for AI computing power. Data source: Shanghai and Shenzhen Stock Exchanges, Wind, etc. Note: As of September 30, 2026, according to the CNI 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 significant, and short-term gains or losses do not predict future performance. Investors should invest rationally based on their own financial situation and risk tolerance, paying close attention to position sizing and risk management. ETF fee disclosure: When investors subscribe to or redeem fund shares, the subscription/redemption agent may charge a commission of no more than 0.5%. On-exchange trading fees are subject to actual charges by the securities firm, with no sales service fee. Feeder fund fee disclosure: Class C of the feeder fund charges no subscription fee; redemption fee is 1.5% within 7 days, 0% for 7 days (inclusive) and above; sales service fee is 0.3%. Class A of the feeder fund charges a subscription fee of 1% for amounts under 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; redemption fee is 1.5% within 7 days, 0% for 7 days (inclusive) and above; no sales service fee. According to the fund manager's assessment, the ChiNext Artificial Intelligence ETF, Hong Kong Stock Connect Information Technology ETF, and STAR Chip ETF are classified as R4-medium-high risk, suitable for aggressive (C4) and above investors; please refer to the sales institution for suitability matching opinions. Risk disclosure: The ChiNext Artificial Intelligence ETF passively tracks the ChiNext Artificial Intelligence Index, whose base date is December 28, 2018, and launch date is July 11, 2024. The index's annual returns from 2021 to 2025 were: 17.57%, -34.52%, 47.83%, 38.44%, and 106.35%, with annualized volatility of 23.73%, 27.34%, 38.02%, 45.42%, and 41.1% over the same periods. Index constituent composition is adjusted according to the index compilation rules as appropriate; its backtested historical performance does not predict future index performance. Index constituents mentioned in this article are for illustrative purposes only; individual stock descriptions do not constitute investment advice of any form and do not represent the holdings or trading activities of any fund under the manager. Any information appearing in this article (including but not limited to individual stocks, commentary, forecasts, charts, indicators, theories, and any form of expression) is for reference only, and investors are responsible for any investment decisions they make independently. Additionally, any views, analyses, and forecasts in this article do not constitute investment advice of any form to readers, nor shall they bear any responsibility for direct or indirect losses arising from the use of this content. 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 fund performance. Fund investment requires caution. MACD golden cross signals have formed, and these stocks are performing well!
Comments