The current capital market is undergoing unprecedented transformation and challenges, making the optimization of asset allocation with the aid of professional investment research a core concern for investors. Concurrently, China's public fund industry is experiencing a profound shift from scale expansion to high-quality development. The iterative renewal of fund manager teams and the deep reshaping of investment research systems are injecting new vitality into the market. Amidst this interplay of industry transformation and market volatility, Securities Times China launches the "Weekly Fund Manager Watch" column, aiming to identify industry trends with a professional perspective that "pierces through the fog." The column will systematically deconstruct the underlying logic of their investment frameworks and forward-looking market thinking through in-depth dialogues with excellent fund managers, building a professional bridge connecting industrial change and asset allocation, and providing investors with valuable references combining industry depth and market timeliness. In this edition, Tang Xiao斌, Fund Manager of GF Vision Select, stated that AI is a long-term industrial wave comparable to the internet and new energy. If 2023 to 2025 was a period of technological explosion, 2026 may usher in AI's "Darwinian Moment." Chen Ziyang, Fund Manager of Great Wall Cycle Preferred, believes the entire cyclical sector is entering a new pattern dominated by strong demand from "new quality productive forces," long-term supply constraints, and diversified driving factors. Pharmaceutical Fund Manager Fang Wei, Cyclical Fund Manager Jin Ye, New Energy Fund Manager Li Yifan, Media Fund Manager Lu Yiqiao, and Technology Fund Manager Gao Peng from Galaxy Fund unanimously agree that AI is becoming a significant background factor influencing industry efficiency improvements and the reshaping of valuation logic.
GF Fund's Tang Xiaobin: Computing Power Drives Storage into "Super Cycle," AI May Approach "Darwinian Moment" "AI is a long-term industrial wave comparable to the internet and new energy. If 2023 to 2025 was a period of technological explosion, 2026 may usher in AI's 'Darwinian Moment'," said Tang Xiaobin, Fund Manager of GF Vision Select. Against the backdrop of sustained high景气 in the AI industry chain this year, the fund he manages has achieved leading performance due to its forward-looking positioning in the storage and semiconductor equipment themes. Tang Xiaobin pointed out that identifying market hotspots is easy, but understanding the industry's essence and selecting companies with genuine Alpha requires building cognitive barriers through deep research and maintaining long-term commitment to the industry. Based on a comprehensive assessment of industry trends, profit certainty, and institutional holdings structure in the fourth quarter of 2025, he proactively shifted the allocation focus towards storage and related semiconductor sectors, capturing the market's upward momentum.
Differentiated Allocation and Forward-Looking Adjustments Lock in High-Growth Themes This year, the high景气 of the AI industry chain has continued, with storage and semiconductor equipment performing strongly. As of January 25, 2026, GF Vision Select achieved a year-to-date return of 35% through selective stock picking, significantly outperforming the storage industry index and ranking among the top three funds in the market. The fund's performance breakthrough primarily stems from Tang Xiaobin's differentiated portfolio layout and precise adjustment operations. Differing from peers who simply focused on storage modules, he constructed a core holding of "storage + semiconductor equipment," capturing超额 returns far exceeding the market. The storage portion mainly covers price-sensitive mainstream storage modules and niche storage like NOR Flash, fully capturing the benefits of the price increase cycle. In the semiconductor equipment segment, the fund covers core areas such as front-end, back-end, testing, and packaging & testing. Its profit logic is not only independent of storage chip price fluctuations but also directly benefits from the capacity expansion demands of domestic storage leaders. Consequently, this segment has recently continued to outperform storage modules, becoming a key incremental contributor to the portfolio's returns.
This allocation originated from Tang Xiaobin's dual forward-looking assessment of market trading structure and industry trends at the end of the fourth quarter of 2025. At that time, domestic chip leaders had high overlap in the top ten holdings of public funds. Against the backdrop of steady fund规模 growth and limited new buying power, the possibility of further increased allocation by funds was low. Additionally, the market had concerns about their capacity and yield rate bottlenecks. Tang Xiaobin predicted that this sector might experience a period of consolidation and fluctuation within the next two quarters. Simultaneously, the storage sector was facing dual catalysts: leading companies preparing for listings and rising global memory prices, with institutional allocation ratios still relatively low. From both a trading structure and industry cycle perspective, it offered investment value for at least the next six months. Semiconductor equipment directly benefited from the capacity expansion of domestic storage leaders, presenting a clear growth path. Therefore, he decisively shifted holdings from domestic chips to storage and semiconductor equipment concept stocks. Tang Xiaobin stated that once he identifies a high-growth direction, he commits to it for at least 2 to 3 quarters, aiming to capture industry红利 by deeply exploring quality targets in sub-sectors and avoiding missing core opportunities due to frequent adjustments during short-term fluctuations.
Understanding Industries Relies on Depth, Holding Returns Relies on Conviction Behind the significant gains in the storage and semiconductor industry chain lies not only Tang Xiaobin's resolve to resist short-term market hype but also his adherence to his circle of competence and commitment to deep research. While market attention focused on themes like brain-computer interfaces, commercial spaceflight, and AI applications, he avoided a "scattergun" approach to分散配置, persistently concentrating on industry directions he understands well. He直言 that a portfolio attempting to cover all hot sectors难以深度把握 the core logic of any single area and cannot maintain conviction during industry volatility. "Identifying a hot industry is easy; just watching the market can achieve that. But truly understanding the industry's essence, selecting companies with stronger Alpha, and holding onto winning stocks require building firm conviction through deep research," Tang Xiaobin直言. No industry can rise in a straight line at a 45-degree angle全年. When an industry experiences significant volatility and sufficient turnover, an investor's ability to hold firm tests their depth of understanding of the industry.
To fortify the barrier of deep research, Tang Xiaobin has established an efficient information processing and decision-making mechanism. On routine trading days, he quickly reads twenty or thirty earnings call transcripts related to storage and semiconductors, uses AI tools to extract key points, and cross-references them with his own knowledge. Simultaneously, using meso-level sub-industry research as a starting point, after determining the semiconductor equipment direction, he further focuses on the storage equipment theme. Then, based on market conditions and marginal changes in the industry, he shifts towards overlooked sub-sectors with high potential, such as packaging and testing equipment, capturing differentiated超额收益 within his focused area of expertise.
Bullish on Storage "Super Cycle," Quality Small and Mid-Caps Offer Room for Growth Looking ahead from the beginning of 2026, Tang Xiaobin believes AI investment is an industry trend with long-term vitality, its importance comparable to the internet and new energy waves. "If 2023 to 2025 were the years of AI technology's 'big bang,' accompanied by some noise and喧嚣, then 2026 may enter AI's 'Darwinian Moment'," he said. From an industry trend investment perspective, he is optimistic about the storage "super cycle" and the opportunities from breakthroughs in semiconductor equipment localization. He judges that AI investment has entered its second stage: the initial phase led by large-cap leaders has concluded, and during the mid-term adjustment,个股 Alpha opportunities are becoming prominent. Stock-picking ability becomes key to obtaining超额收益, which also opens up growth space for quality small and mid-cap targets. Tang Xiaobin analyzes that the storage "super cycle" has arrived,核心 driven by the dual resonance of AI-driven demand surge and industry supply contraction. On the demand side, AI large models impose disruptive requirements on storage performance, HBM demand is exploding with leading manufacturers' capacity already booked, simultaneously squeezing supply for traditional electronics like phones and PCs, while enterprise SSD demand is positive. On the supply side, following the last景气 cycle, storage原厂 incurred huge losses, significantly收缩 capital expenditure, only making minor investments in new products like HBM. Current inventory destocking is sufficient,原厂 are halting quotes, DDR5 and HBM prices show significant elasticity, and domestic storage leaders are expected to accelerate their 2026 capacity expansion and capital expenditure pace. Comparing domestic and international chains, the domestic chain currently lacks advantage in performance, financials, and valuation compared to the international chain, but offers broader growth space. Estimates suggest the domestic chip market could reach 8 trillion yuan by 2027, with full localization potentially driving market capitalization towards 100 trillion yuan. The domestic chain's market performance lags the international chain by about a year. 2025 was primarily about thematic positioning, while 2026 is expected to see a Davis Double Click, with a growth slope likely steeper than that of the international chain.
Great Wall Fund's Chen Ziyang: "Nonferrous Metals Feast" Not Accidental, Chemical Industry May See Inflection Point Opportunity "Cycles are human nature." When first meeting Chen Ziyang, Fund Manager of Great Wall Cycle Preferred, he mentioned this phrase from the influential predecessor Zhou Jintao. In his view, investing requires counter-cyclical, even counter-human-nature thinking and operations to strive for more ideal investment returns amidst cyclical rotations. As a fund manager long focused on cyclical sectors, Chen Ziyang practices the investment philosophy of "investing in high-quality enterprises during景气 periods at reasonable or even low prices" in his investments. Early positioning has enabled the products he manages to achieve significant超额收益. Looking ahead to 2026, Chen Ziyang believes the entire cyclical sector is entering a new pattern dominated by strong demand from "new quality productive forces," long-term supply constraints, and diversified driving factors.
Four Factors Reshape and Drive Nonferrous Metals行情 回顾2025, the cyclical sector, represented by nonferrous metals, was undoubtedly the brightest主线 in the A-share market. Wind data shows that as of December 31, 2025, the Shenwan Nonferrous Metals index (801050.SI) rose 94.74% year-to-date, nearly doubling, leading among the 31 Shenwan primary industries. This "nonferrous metals feast" was not accidental. Chen Ziyang analyzes that it was driven by four factors: the US debt cycle, structural demand, supply chain security, and supply cycles, which systematically reshaped the pricing logic of nonferrous metals. First is the US debt cycle. Soaring US debt and deficits have raised global concerns about their sustainability and dollar credibility, prompting many central banks to reduce US Treasury holdings, increase gold reserves, and promote diversification of reserve systems, thereby supporting stronger precious metal prices. Second is structural demand驱动. The development of the AI industry and the acceleration of global energy transition have spawned incremental demand for industrial metals like copper and aluminum. Third is supply chain security. The current global supply chain格局 is shifting from emphasizing efficiency to emphasizing security – countries are increasing reserves of critical minerals, energy, and food to ensure industrial security, also driving demand for commodities. Finally, changes in the supply cycle. Capital expenditure for major nonferrous metal varieties basically peaked around 2011, followed by a long period of contraction, resulting in a significant industry output gap. Supply-side constraints persist, supporting industry prices.
Judging Industry Inflection Points via Capital Expenditure Cycles "From past experience, some industries can maintain high景气 or high ROE for a long time, while others cannot, which is greatly related to the length of the industry's capital expenditure cycle. We prefer to focus on an industry when both ROE and average PB are low, potentially achieving a unity of win rate and odds." Regarding his specific investment methodology, Chen Ziyang summarizes it as a capital expenditure cycle-based approach. Introducing his stock selection framework, Chen Ziyang stated that he comprehensively evaluates investment targets from three dimensions:景气, quality, and valuation, adhering to the principle of investing in景气 high-quality enterprises at reasonable or even low prices. In his decision-making framework, when these three conflict,景气 is given higher weight, while valuation requirements can be适度放宽. This inclination stems from his deep understanding and application of the "capital cycle" theory. Discussing stock selection in cyclical sectors, Chen Ziyang said the first focus is on the business model, followed by company competitiveness. However, the business models of upstream resources and midstream cyclical products differ significantly. He直言 preferring the business model of upstream resource products. "Upstream resource products have more controllable costs, with fluctuations mainly stemming from product price changes. For example, a copper mine has relatively fixed costs, with elasticity coming from copper price changes. Cyclical products like chemicals often follow a manufacturing-like business model, requiring raw material procurement and generating products through chemical reactions. Costs are closely tied to crude oil prices, product prices relate to downstream demand, and profits come from spreads, making both ends relatively difficult to control," he added.
Bullish on Gold, Minor Metals, and the Chemical Industry For 2026, Chen Ziyang believes cyclical stocks will exhibit more differentiated characteristics, requiring精细布局 based on changes in supply and demand structure. Regarding favored directions, Chen Ziyang直言 the nonferrous metals行情 is likely to continue. The current high-profitability state of the nonferrous metals industry may persist for a relatively long cycle. Driven by sustained new demand, the nonferrous metals sector is gradually acquiring growth attributes and deserves valuation re-rating. Furthermore, compared to international peers, the valuation of domestic nonferrous metal enterprises is significantly lower, yet their growth and core competitiveness are not inferior. Meanwhile, domestic enterprises' continuous exploration and breakthroughs in core technologies like exploration, mining, and smelting have also made outstanding contributions to global mining development. He believes minor metals are in a cycle of strategic reserve enhancement, often facing supply constraints due to resource scarcity and limited new capacity, making their investment value particularly noteworthy. Regarding precious metals, Chen Ziyang remains optimistic. He believes the two major trends of central bank gold buying and household asset allocation demand show no signs of reversal, and the long-term allocation logic for gold remains solid. For silver, although the year-to-date gain is substantial and the gold-silver ratio is at a historical low, considering persistently low global silver inventories and a relatively宽松 liquidity environment, short-term price support factors still exist. For industrial metals like copper and aluminum, Chen Ziyang holds a阶段性 cautious attitude. However, from a medium-to-long-term perspective, based on the unchanged demand logic from energy transition and AI development, he remains optimistic about the long-term prospects of copper, aluminum, and other品种.
Beyond nonferrous metals, within the cyclical sector, Chen Ziyang特别强调 that the chemical industry may迎来 valuation re-rating, and inflection point opportunities deserve attention. Especially since the valuation percentile of the chemical sector is at a historically low level, while profit expectations are gradually repairing. "First, industry capital expenditure is declining, new capacity投放 is nearing its end, the industry will move from oversupply towards balance or even shortage, and profitability will修复; second, against the backdrop of 'anti-involution,' the barriers to entry for chemical projects will become increasingly high, and capacity may become a scarce resource or license; third, from a global perspective, no other country or region possesses a chemical industry sector as comprehensive, efficient, and cost-advantageous as China's. Once profits recover, chemical assets will be revalued," he analyzed.
Galaxy Fund: Moving Beyond "Concept Speculation," AI Quietly Reshapes Valuation Logic for Pharma, Media, and Energy Recently, five fund managers from Galaxy Fund covering pharmaceuticals, cyclicals, new energy, media, and technology collectively voiced their views, sharing latest perspectives on their respective investment focuses. Despite operating in different sectors, they reached a高度一致的 view on the key variable of AI: Pharmaceutical Fund Manager Fang Wei, Cyclical Fund Manager Jin Ye, New Energy Fund Manager Li Yifan, Media Fund Manager Lu Yiqiao, and Technology Fund Manager Gao Peng all agree that AI is becoming a significant background factor influencing industry efficiency improvements and the reshaping of valuation logic. From the commercialization progress at the application end, to demand spillover at the industry chain level, to the re-examination of corporate profit models and investment节奏, AI is no longer just a thematic concept for a single sector. Instead, it is gradually embedding itself into the investment decision-making frameworks of multiple industries in a more fundamental way.
From Independent Concept to General Tool: AI's Role is Changing In 2026, the market's focus is shifting from whether a company has an AI concept to whether AI can genuinely change its profit model. This means AI's positioning in investment is gradually transitioning from sentiment-driven thematic speculation to a variable that can be incorporated into financial models and valuation reassessments. Fang Wei pointed out that AI is not a new赛道 independent of existing industries but更像 a productivity tool, whose core value lies in transforming corporate efficiency structures. In industries like pharmaceuticals, AI's impact is not reflected in short-term revenue explosion but through enhancing R&D efficiency, optimizing processes, and reducing expense ratios, thereby altering companies' profit margin中枢 and long-term valuation assumptions. Once AI can substantively affect key parameters like expense ratios and profit margins, its investment logic有望进入 a stage where it can be quantified and modeled. On the application front, Lu Yiqiao believes 2026有望成为 the year of systematic AI application落地, rather than a continuation of sporadic breakthroughs. On one hand, large model capabilities have significantly improved, with reduced hallucination rates and enhanced generalization能力, making application experiences usable. On the other hand, major tech companies加速下场, integrating ecosystems and流量入口, providing a realistic foundation for the commercial diffusion of AI applications. In this context, AI is no longer just a technology showcase but a practical test of "how to monetize." Gao鹏补充指出 from an investment节奏 perspective that AI remains a 3-to-5-year long-term主线, but its evolution path is not linear. As the market shifts from hardware expansion to commercial validation,阶段性波动 or even a "trough of disillusionment" is not unexpected. The key lies not in short-term sentiment swings but in which directions and which companies can ultimately率先跑通 business models and deliver on performance.
Reshaping Business Models and Valuation Logic Although the aforementioned fund managers focus on different areas, they present高度一致的判断 on AI: AI has not altered their original investment frameworks but has become an important underlying variable affecting industry efficiency, business models, and valuation logic. Lu Yiqiao observes that AI is substantively changing content production and distribution efficiency. As large model capabilities improve and major tech ecosystems integrate, AI is beginning to meet conditions for commercial落地. Advertising distribution and content generation are更容易成为 the first beneficiaries. In this process, AI is not an independent investment主线 but a key tool driving efficiency leaps in the media industry and opening new market spaces. In the pharmaceutical field, Fang Wei同样强调 that AI does not constitute a new赛道脱离 the original industrial logic.相反, its core value is reflected in reshaping the operational structures of existing healthcare enterprises. Whether through improved R&D efficiency or changes in expense ratios and profit margin中枢, the effects will ultimately be reflected in corporate financial models and long-term valuation assumptions. In his view, only when AI is truly embedded in business processes, can be quantified, and reflected in financial statements do related companies possess sustained investment value. Despite differing application scenarios and industry attributes, the fund managers unanimously believe that the investment significance of AI lies not in the expansion of conceptual boundaries but in whether it becomes a tool that can enhance efficiency, improve profit quality, and ultimately translate into performance. Precisely because of this, AI is跨越 industry boundaries to become an important background factor重构 the investment logic of multiple sectors.
Reverse Pull on Upstream Resource and Energy Demand As AI technology moves from innovative R&D to scaled application, its influence is deeply传导 from terminal demand upstream along the industry chain. The strong demand for computing power, storage, and network facilities driven by AI development is continuously拉动 investment in nonferrous metals, energy, and related infrastructure sectors, making these areas increasingly critical "foundations" supporting AI's long-term development. In the nonferrous metals领域, Jin Ye指出 that AI computing power investment itself will significantly increase demand for upstream resources. For instance, the usage of metals like copper, tin, and aluminum in computing infrastructure and related facilities is continuously rising, providing medium-to-long-term support for the demand logic of certain nonferrous varieties. Such changes are not driven by sentiment or themes but by real demand increments from industrial capital expenditure, hence becoming noteworthy structural clues within the cyclical sector. In the new energy direction, Li Yifan focuses more on the industry itself, paying attention to energy storage and photovoltaics' performance delivery after cyclical reversals. He suggests that rising computing power demand further highlights the importance of power supply and energy storage systems. Some远期 application scenarios有望 provide new imagination space for new energy technology paths, but this change仍需回到 industry fundamentals for verification. Regarding investment节奏, the aforementioned fund managers also provided relatively一致的判断. They believe AI remains a long-term主线 spanning many years, but its evolution process will inevitably involve阶段性波动. As market focus shifts from capital expenditure to commercial兑现, some directions may experience adjustments or even a "trough of disillusionment." However, such volatility does not signify a logic reversal but instead provides investors with a window期 to检验 business models and筛选 companies with delivery capabilities.
Comments