Recently, an interesting phenomenon has emerged: ByteDance's DouBao has surpassed 50 trillion tokens in daily average usage, with its new version gaining instant popularity, demonstrating remarkable user stickiness and ecosystem penetration. Meanwhile, Alibaba's AI health assistant "Ant Afu" soared to the top of app store charts immediately after its new version was released, rapidly reaching 15 million monthly active users.
This year, there has been a palpable sense that AI has quietly integrated into the workflows of various industries, much like essential utilities, and has become embedded in our daily lives—covering clothing, food, housing, transportation, education, and health. The development of AI is gradually transitioning from a phase focused on "computing power infrastructure" to a new era that tests genuine commercial profitability. Alongside this shift, the complexity of AI investing has also evolved.
Looking at institutional forecasts for 2026, AI remains a nearly unanimous choice on the growth trajectory. However, capital is no longer solely betting on AI narratives and computing infrastructure; it is also focusing on whether the technology can be practically implemented to create verifiable business models and profit growth.
For instance: Dongxing Securities analysis indicates that the artificial intelligence industry is entering a new phase of "three-dimensional resonance" involving policy, technology, and demand. Strategic support at the national level, the critical juncture of technology moving from the lab to industrialization, coupled with the release of genuine global demand—such as the surge in DouBao's usage—collectively form the core logic for AI investment in 2026. Huaan Securities analysis suggests that if 2023 was the inaugural year for large models, and 2024-2025 the period of accelerated computing infrastructure build-out, then 2026 is expected to be a crucial stage where AI transitions from infrastructure to scaled commercial applications and deep industrial penetration. The competition in AI will no longer be about who has the most parameters, but about whose technology can be deployed at scale to genuinely enhance efficiency across various sectors. Everbright Securities released a report stating that AI applications are expected to achieve breakthroughs in 2026, with computing costs projected to continue declining, paving the way for application prosperity.
In summary, AI investment has entered a new phase characterized by the resonance of "commercial implementation" and "infrastructure," making the landscape more complex. For ordinary investors facing such a intricate field, compounded by AI's rapid technological iterations and significant stock volatility, how can one more easily and effectively share in the dividends of AI development?
With stock selection difficulty and investment risk both rising, ordinary investors can better diversify risk and capture trends by allocating through index funds. For example, the ChiNext Artificial Intelligence ETF (159363) and the STAR Artificial Intelligence ETF (589520)—this AI "twin stars" pairing—serve as convenient and efficient tools for AI allocation.
The ChiNext Artificial Intelligence ETF (159363) is the market's first ETF product tracking the ChiNext Artificial Intelligence Index (970070.CNI). This index covers the entire industry chain from AI basic hardware and software algorithms to vertical applications, making it less likely to miss out on rallies in any sector. It places significant weight on the AI computing power sector, particularly with over 56% allocation to optical modules/CPO. In terms of constituent stocks, it leans towards technologically mature AI leaders that are already realizing or close to realizing earnings, such as the top three holdings "Yi Zhong Tian," which are closely linked to overseas computing power investments, offering high visibility and elasticity in orders and performance.
The STAR Artificial Intelligence ETF (589520) exhibits distinct characteristics of "autonomous control and heavy weighting in domestic chips." Its underlying index, the STAR Artificial Intelligence Index (950180.CSI), has a semiconductor allocation exceeding 50%, including components like Cambricon, leading domestic computing power chip companies, closely tied to the logic of domestic chip substitution.
Furthermore, these two indices are based on the ChiNext and STAR Market boards respectively, offering higher earnings elasticity during industry upswings. As of December 23, the ChiNext Artificial Intelligence ETF (159363) has achieved a cumulative net value increase of 93.48% over the past year, with a Sharpe Ratio of 2.07, significantly outperforming its peers. The newly established STAR Artificial Intelligence ETF (589520) has seen a cumulative net value increase of 36.73% over the past six months, with a net value growth rate of 50.27% in the third quarter alone.
As of the end of the third quarter this year, the sizes of the two products reached 4.352 billion yuan and 576 million yuan respectively, showing steady growth. Particularly, the ChiNext Artificial Intelligence ETF (159363) is the largest ETF tracking its respective index, providing investors with a solid liquidity foundation.
If you are more optimistic about global AI computing power demand and earnings certainty, consider focusing on the ChiNext Artificial Intelligence ETF (159363) or its off-exchange feeder funds (Class A 023407; Class C 023408). This choice targets the relatively certain growth opportunities arising from AI technology empowering various industries globally and driving hardware demand.
If you are more bullish on the substitution potential and high growth of domestic AI and are willing to tolerate higher volatility, consider the STAR Artificial Intelligence ETF (589520) or its off-exchange feeder funds (Class A 024560; Class C 024561). This option offers stronger explosive potential during industry upswings and is suitable for capturing opportunities from cutting-edge technological breakthroughs.
If you wish to adopt a balanced approach covering both "hardware and software" aspects of the AI industry, consider allocating to both funds simultaneously. This strategy allows you to capture frontier technological breakthroughs while also sharing in the commercial returns from technology implementation, providing a balanced approach for a more composed investment mindset.
Finally, it's important to remember that the characteristics of high growth and high volatility in the AI industry are unlikely to change, and the market may inevitably experience fluctuations due to short-term sentiments (such as concerns about capital expenditure or high valuations). However, as many institutional views suggest, with the three-dimensional resonance of policy, technology, and demand, the long-term industrial trend for AI is clear. For ordinary investors, it is advisable to participate using strategies like phased entry and long-term holding, ensuring you remain invested to share in the full-cycle dividends from the industry's early explosion to comprehensive maturity.
Risk Warning: The above content is solely an objective summary based on public information. Past performance is not indicative of future results. Before making investment decisions, investors should carefully read the fund's legal documents, fully understand the product's characteristics and risks, and make prudent decisions based on their own risk tolerance. Investing involves risks; proceed with caution.
Comments