Fund Managers' AI Portfolio Dilemma: Balancing Performance Rankings and Long-Term Value

In Q1 2026, China's A-share market navigated turbulent waters amid the dual forces of AI hype and geopolitical disruptions. As mutual fund quarterly reports were released en masse, an industry-wide scrutiny of fund managers' "consistency between words and actions" quietly unfolded — numerous fund managers proclaimed in their reports that they "firmly believe in AI's long-term value," while their actual operations involved frequent swing trading and even substantial reductions in core AI holdings. This phenomenon has not only confused retail investors but also exposed the deep-seated anxiety within the mutual fund industry over reconciling short-term rankings with long-term positioning.

Q1 Report Analysis: Operational Divergence Behind 'Barbell' Portfolios

According to Economic Information Daily, the portfolio structures of top-performing funds in Q1 2026 displayed a classic "barbell" configuration: on one end, AI technology assets representing future directions — spanning frontier tracks such as large model computing power, AI Agent applications, and embodied intelligence; on the other end, HALO assets (high-barrier, low-substitutability core resource targets) characterized by heavy assets and low obsolescence risk, such as energy infrastructure and scarce mineral resources.

This allocation strategy is inherently sound — using AI technology to capture growth upside while anchoring safety margins with HALO assets, balancing offense and defense. However, when we shift from macro allocation to micro-level operations, unsettling signals emerge.

Multiple industry analysts found that some fund managers clearly expressed long-term optimism about the AI value chain in their Q1 investment outlooks, yet their actual portfolio adjustments stood in stark contrast to their written statements. Some fund managers wrote that "AI is the most certain industrial trend for the next decade," while their holding data showed massive reductions in leading AI computing stocks during Q1. Others claimed to "adhere to long-term principles, unmoved by short-term fluctuations," yet their trading records revealed turnover rates far exceeding industry averages.

The Deeper Logic of the 'Say-Do Gap': Rational Choices Under Ranking Anxiety

This "say-do gap" phenomenon is not isolated — it stems from complex institutional causes.

First, short-term ranking mechanisms impose rigid constraints. The current mutual fund industry's evaluation system remains centered on short-term performance rankings. Quarterly and semi-annual rankings directly affect fund managers' compensation, assets under management, and even career trajectories. Under such incentive structures, even fund managers who genuinely believe in AI's long-term value find it extremely difficult to truly "stand pat" when facing periodic sector pullbacks. A quarterly ranking decline can trigger a chain reaction of investor redemptions and shrinking management scale.

Second, the AI sector's inherent high volatility compounds operational difficulty. In Q1 2026, influenced by global geopolitical disruptions and expectations of tightening U.S. AI regulatory policies, the A-share AI sector experienced multiple rounds of sharp rallies and sell-offs. Some AI concept stocks saw monthly price swings exceeding 30%, making "holding steady" an act requiring extraordinary courage. Many fund managers opted for "reduce first, add back later" strategies amid the volatility, attempting to minimize net asset value drawdowns through swing trading — directly contradicting the "long-term holding" philosophy espoused in their quarterly reports.

Third, narrative strategies under information asymmetry. Some industry insiders candidly acknowledge that quarterly report investment outlooks serve a "marketing function" to some degree. In a market environment where AI concepts remain red-hot, expressing bullishness on AI helps attract incremental capital inflows. Meanwhile, flexible portfolio adjustments in actual operations represent fund managers' "real votes" based on their own judgment. This disconnect between written narratives and actual operations is essentially rational behavior under information asymmetry, but it erodes trust among retail investors.

Industry Reflection: The Dual Challenge of Institutional Reform and Information Transparency

The spread of the "say-do gap" phenomenon is prompting deep reflection from regulators and the industry alike.

From an institutional perspective, regulators have repeatedly emphasized guiding mutual funds toward long-term and value investing in recent years. Multiple industry reform measures released in 2025 — including extending fund manager evaluation cycles to three years or longer and reducing the weight of short-term rankings — have been piloted at some leading fund companies. However, judging from actual Q1 2026 performance, the transmission effects of these institutional changes still need time. The "implicit evaluation" based on short-term rankings remains deeply entrenched, and achieving a genuine long-term transformation of evaluation mechanisms remains a long road ahead.

From an information disclosure perspective, current quarterly report narratives lack binding constraints. Fund managers' investment outlooks are largely "soft expressions" — even when inconsistent with actual operations, they do not constitute regulatory violations. Some experts suggest exploring cross-referencing mechanisms between quarterly report narratives and holding changes, requiring fund managers to provide explanations for significant deviations, thereby enhancing the authenticity and effectiveness of information disclosure.

From an investor education perspective, retail investors also need to develop more rational cognitive frameworks. Fund managers' quarterly report views should be treated as "reference signals" rather than "binding commitments." Investors should focus more on medium-to-long-term performance, changes in portfolio concentration, turnover rates, and other quantitative indicators, rather than making subscription or redemption decisions based solely on written narratives.

Outlook: AI Investment Enters Deep Waters of 'De-Bubbling'

Standing at the beginning of Q2 2026, AI technology investment is approaching a critical watershed.

On one hand, AI industry fundamentals continue to improve. Domestic large model capabilities are iterating rapidly, AI Agent commercialization in vertical scenarios such as finance, healthcare, and education is accelerating, and the embodied intelligence sector has reached mass production milestones for several benchmark products. From an industrial trend perspective, the long-term investment thesis for AI remains intact.

On the other hand, following earlier valuation expansion, differentiation within the AI sector is intensifying. Companies with genuine technological moats and profitability will stand out, while purely concept-driven stocks face valuation compression. This means that the core competitive advantage in future AI investing will shift from "daring to go heavy" to "excelling at stock selection," requiring fund managers to demonstrate deeper industrial research capabilities and stronger conviction in their holdings.

For the mutual fund industry, resolving the "say-do gap" dilemma cannot happen overnight. It requires systematic restructuring of evaluation mechanisms, continuous improvement of information disclosure systems, and deep-seated transformation of the industry's investment culture. Only when long-term value investing ceases to be merely "pretty words" in quarterly reports and becomes truly internalized as fund managers' behavioral principles can mutual funds deliver genuinely sustainable returns for their investors amid the AI era's industrial wave.