AI revolution, restructuring of geopolitical order, weakening of dollar credit -- why the "three major narratives" are gaining prominence

In the traditional market analysis framework, we usually focus on three core driving factors: fundamentals (real macroeconomic conditions, industry cycles, corporate earnings, etc.), risk appetite (policy changes, events, etc.), and liquidity (fund size, structure, and trading congestion). Among these, fundamentals hold a long-term core position, as they not only directly affect the corporate earnings expectations on the numerator side but also influence the denominator side; fundamentals impact market sentiment and risk appetite, and also affect the liquidity environment through monetary policy.

However, in the past two years, we have found that the market is increasingly dominated by narratives, exhibiting several obvious characteristics:

First, there are often discrepancies between the market and fundamentals, and market volatility may far exceed the extent of fundamental developments, even diverging in direction;

Second, global capital reallocation tends to amplify narrative power, and capital flows often exhibit convergence characteristics, making it easy to form crowded trades;

Third, volatility is significantly amplified, often exhibiting nonlinear characteristics, where small events can trigger severe fluctuations;

Fourth, the correlation between different asset classes has changed; traditionally, stocks, bonds, commodities, and currencies have different driving factors, often showing low or even negative correlations, but when narratives dominate the market, the low correlation between assets is broken, increasing the difficulty of asset allocation and risk hedging.

Why has narrative begun to dominate the market? We believe the key lies in the chain of transmission: “AI revolution, geopolitical order reconstruction → narrative development → AI dissemination & attention scarcity → capital inflow,” which forms the reflexivity of the AI era.

First, the current three major narratives globally—AI revolution, geopolitical order reconstruction, and weakening of US dollar credit

Narratives can be divided into small narratives and big narratives. Historically, small narratives have been numerous, such as the “Internet Plus” around 2015, which affected some tech stocks, and the “new energy” narrative in 2021, which boosted related sectors. The impact of small narratives is often relatively limited, affecting individual assets and specific targets, and their duration is usually short; once the narrative weakens, it quickly recedes. However, we are currently facing three super narratives simultaneously, whose grandeur and complexity are unprecedented.

(1) The AI revolution is a Kondratiev wave-level cycle, capable of overshadowing many smaller economic cycles, far exceeding the explanatory scope of traditional economic cycles.

First, from the current investment scale, the capital expenditure of US tech companies is projected to account for about 1.9% of GDP by 2025, and may continue to rise above 2% in 2026, with global increases in AI capital expenditure;

Second, in terms of growth stimulation, AI is not merely an industrial contribution; through the improvement of total factor productivity, according to our previous report “AI: A New Round of Technological Revolution Changing the World” (April 22, 2025), it is predicted that in the next decade, AI can contribute 0.5-1.5 percentage points to global potential GDP growth;

Third, from a macro paradigm perspective, AI is different from previous technological revolutions, transitioning from assisting labor to replacing labor, a depth and breadth of change that is unprecedented; Fourth, from the market performance perspective, the AI revolution has brought about drastic structural differentiation, and the performance of computing power chains and capital expenditure beneficiaries is far from what traditional macroeconomics can explain.

It is foreseeable that the impact of the AI revolution will be significant, but the specific path of AI transformation remains difficult to predict—we do not know where the boundaries of AI development lie, what new business models will look like, or what new scenarios and formats will emerge. The market is prone to pricing long-term and uncertain prospects into current asset prices.

(2) The reconstruction of the geopolitical order can only be deduced, making it difficult to generalize.

The impact of the ongoing reconstruction of the global geopolitical order is also substantial, but geopolitical order is the most unpredictable information. Our analysis of geopolitical events often needs to be based on hypothetical deductions, making it difficult to form general conclusions, which become sources of market volatility. Events such as Russia-Ukraine, U.S.-Israel-Iran, etc., continuously confirm the unpredictability of geopolitical events.

(3) The weakening of U.S. dollar credit requires global funds to seek reallocation directions.

Moreover, narratives themselves can reinforce each other, and even small narratives are strengthening the larger narrative. For example, industries such as new energy, chips, and rare earths have gained more strategic significance under the grand narrative of geopolitical competition, rising from a purely industrial cycle to a national strategic level, forming a more consistent macro narrative. This integration of narratives further strengthens market consistency.

Second, the production and dissemination methods of information in the AI transformation.

In the 2015 bull market, new media accelerated information transmission, boosting thematic trends. Investors made investment decisions based on various viewpoints and stories on new media, downplaying fundamental analysis.

Current AI technology further transforms the way information is disseminated, enhancing narrative capabilities, reducing generation costs, and accelerating transmission speed. First, AI lowers the cost of content creation; in the past, creating a high-quality analytical article required a significant amount of time and effort, but now AI can generate a seemingly professional analytical article in minutes, leading to an explosive growth of market content. Second, AI accurately pushes narratives based on algorithms and investor preferences, prioritizing the most engaging content for users, and combined with the use of new media, further accelerates the speed of information dissemination.

The cost of information generation approaches zero, and the transmission speed is exponential, resulting in grand narratives spreading faster and wider, consensus forming more quickly, and thus driving capital allocation, making it easier to influence the market.

Third, the scarcity of investor attention.

In the era of information explosion, investor attention has become the most scarce resource. Among thousands of stocks and countless macro paths, the one that stands out is inevitably the "most grand and consistent" narrative. This essentially forms an "information cocoon," where investors are guided by algorithms and their own preferences, focusing only on certain specific topics and assets. Undoubtedly, this scarcity and concentration of attention attract a large influx of capital.

The above factors reinforce each other, forming the reflexivity of the AI era. The traditional reflexivity of Soros is that there exists a bidirectional, cyclical relationship of mutual shaping between market participants' cognition and market reality, where prices not only reflect fundamentals but also change fundamentals in return, leading the market to remain in a state of dynamic imbalance over the long term The reflexivity of the AI era is further amplified, as narrative costs are lower, deduction speeds are faster, consensus formation is quicker, and capital inflows more easily create positive feedback loops. The stronger the narrative, the more capital flows in, leading to rising prices, which further reinforce consensus and narrative through valuation and capital.

One case is the self-reinforcement of AI narratives and the fundamentals of tech giants. The AI narrative drives capital into the tech sector, pushing up the valuations of tech giants, and high valuations lead to more capital expenditures and more favorable financing conditions. Increased capital expenditures improve short-term fundamentals, further strengthening the narrative, resulting in reinforced narratives and more capital inflows.

Another case is the prediction market Polymarket, where market transactions by some individuals form probability expectations, which then influence market expectations through dissemination, further attracting capital for related trades, thereby reinforcing market consistency.

The reflexivity of the AI era forms a powerful feedback loop mechanism, but it also makes deviations in market pricing potentially larger, bringing about inherent risks.

First, do not confuse narrative with reality. The essence of a narrative is still a pricing of the future, which leads to a certain divergence from the fundamental reality. When the pace of reality (such as energy shortages, investment return expectations) significantly lags behind the narrative imagination, the narrative may weaken temporarily, as seen this year when the market no longer "rewards" the capital expenditures of tech giants.

Second, do not simply extrapolate linearly. In a narrative-driven market, changes are often nonlinear; technology may have "singularity moments" and may also exhibit "diminishing returns"; geopolitics can only be deduced, not induced, all of which reduce the effectiveness of linear extrapolation.

Third, do not confuse short-term trading with long-term allocation. Some narratives may be correct in themselves, but that does not mean they will rise continuously. Reflexivity amplifies price deviations, and the market may have already priced in the best or worst extreme scenarios, with high valuations and crowded trades becoming the greatest vulnerabilities. When narratives show marginal changes, positive feedback loops can turn into negative feedback loops, potentially causing greater liquidity shocks. If one buys at the peak of a narrative's reinforced deviation, it does not guarantee profits and may even face significant short-term loss risks.

In the face of a narrative-driven market, how should investors respond? Here are a few thoughts:

First, establish a trading framework for the AI era. Focus not only on fundamental data ("wind shifts") but also on "market movements" and "sentiment shifts," which can sometimes be even more critical. A simple framework is: (1) when fundamentals are weak and narratives are strong, pay more attention to trading opportunities; (2) when fundamentals are strong, narratives are weak, and crowding is low, adopt a slow bull strategy; (3) when both fundamentals and narratives are strong, and crowding is high, identify whether pricing has deviated and how significant that deviation is; (4) when both fundamentals and narratives are weak, look for other opportunities.

Second, know when to return or even "reverse." Even if a grand narrative is correct in the long term, when the narrative is highly consistent and capital is extremely crowded, one must be wary of reflexivity peaks. The important thing is not to predict where the peak is, but to set clear exit rules. At this time, trading indicators such as valuation percentiles, capital flows, and trading structures become particularly important When trading structures are extremely crowded and narratives can no longer attract incremental attention, even if the fundamentals remain positive, one must be more cautious; when valuations are digested, if the long-term narrative remains intact, adjustments present opportunities for re-entry.

Third, tracking and debugging. The key lies in marginal changes. When marginal changes in narratives are detected, timely adjustments to positions are necessary. Key marginal changes may include: (1) changes in policy signals (marginal adjustments in regulation, industry, monetary policy, etc.); (2) significant industrial information (capital expenditure validation, physical bottleneck validation…); (3) evolution of competitive landscape (new entrants, technological breakthroughs, disruptive applications, changes in market share…). In terms of tracking systems, in addition to market sentiment indicators such as valuation, volatility, and trading structure, one can also pay attention to changes in participation (whether individual investors are flooding in, media attention (whether all media are discussing the same topic), etc.

Fourth, narratives ultimately require fundamental verification, which remains the long-term weighing machine and "anchor." Narratives are often not a one-time game, but rather a multi-round game process of "narrative - verification/falsification - narrative reinforcement/narrative weakening - re-verification." Regular fundamental verification is key to narrative correction and requires continuous tracking of fundamental data. At the same time, beyond narratives related to AI and geopolitics, it is essential to retain a portion of the portfolio for verifiable profits and cash flow allocation, focusing on non-narrative sectors with robust fundamentals, stable cash flow, and reasonable valuations as a hedge.

In summary, we are currently in the overlap of four major super narratives: the AI revolution, the restructuring of geopolitical order, the weakening of US dollar credit, and the transition of China's old and new driving forces. This has become the core of our cognitive alpha exploration in recent years. Narratives influence market sentiment and capital flows, amplify asset price performance, and even affect fundamentals through reflexivity. If in the past few years, based on the judgment of "transition of old and new driving forces," we concluded that "bonds are better than stocks"; now, in the face of the wave of the "AI revolution," we are optimistic about investment opportunities in technology growth. At the same time, the reshaping of the geopolitical order implies increased defense spending and strategic reserves, which benefits resource products and consumables, and also promotes global central banks to diversify foreign exchange reserves, benefiting precious metals.

In this environment,

the Merrill Lynch clock framework based on growth + inflation -> monetary policy -> asset prices has failed

and has been replaced by,

grand narratives (the aforementioned major narratives) + paradigm shifts (efficiency yielding to security, monetary policy yielding to fiscal policy, etc.) + communication models in the AI era -> new dumbbell portfolio (world-changing technology, non-renewable resource products and consumables) + multi-asset rotation + cognitive alpha + positive correlation.

Risk warning and disclaimer

The market has risks, and investment should be cautious. This article does not constitute personal investment advice and does not take into account the specific investment goals, financial situation, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article align with their specific circumstances. Investing based on this is at one's own risk