CITIC Securities Li Chong: HALO is worth attention while the new AI direction is unclear

While the geopolitical shadow of the U.S.-Iran and Israel-Iran conflict has not yet fully blanketed global markets, a transformation aimed at revaluing AI value has already quietly begun. In this transformation, the concept of HALO (heavy assets, low attrition rates) has emerged out of nowhere, becoming a crucial link connecting AI Disruption (artificial intelligence disruption) with market defensive needs.

Recently, Li Chong, Chief Analyst for Overseas Macro at CITIC Securities’ Joint Research team, said in an interview with China Securities Journal that the underlying logic of the HALO narrative is a comprehensive reappraisal in which the market completes its shift from chasing “light assets with high growth” during the AI bubble phase to revaluing low-replacement-risk assets’ survival premium. The break in power supply bottlenecks crushes expectations of infinite capacity expansion, driving asset pricing to anchor on “immune assets” that are difficult to digitize and have tangible barriers. This logic has triggered differentiation across global markets: Europe’s heavy-asset segments have shown resilience, while risks are emerging in India’s labor outsourcing targets. A-shares and Hong Kong-listed shares fit the framework of real-asset scarcity, benefiting hard-core tangible assets. At the same time, investment boundaries still need to be clarified—balancing short-term defensive positioning while flexibly capturing long-term growth opportunities bound to the AI track.

Underlying logic of the narrative:

From the AI bubble to survival-premium revaluation

The core of HALO is that, against the backdrop of rapid iteration of AI technology, assets that are hard to replace through digitization and that have tangible barriers become “immune assets” for countering technological disruption. The market is shifting from a pricing logic of “scalable light-asset narratives” to one of “constructable, hard-to-replace real capacity and networks.”

“Unlike the earlier AI bubble phase, when the logic was simply to pursue high growth with light assets, the HALO narrative signals that the market has moved from a ‘systemic bubble-pricing’ stage to a ‘structural differentiation-pricing’ stage.” Li Chong believes that the trigger for this shift stems from the exposure of real bottlenecks within AI expansion logic, and that the key inflection point is management’s statement that “we are not short on compute power, we’re short on electricity.” In the first phase, the market assumes that compute expansion has a high degree of certainty, and the growth room for core assets is viewed almost as a linear extrapolation. But when power constraints become a focus of public discussion, the expansion logic shifts from infinite supply to constrained expansion; the growth path starts to show real boundaries. The market then has no choice but to reassess who can benefit from resource bottlenecks, who will face replacement shocks, and who has survival certainty.

The essence of HALO is a one-off survival-premium revaluation of low-replacement-risk assets—that is, the Survivors in the market. This logic is especially evident in the performance of country-level assets. Li Chong said that Europe has a relatively defensive profile under the narrative background of AI Disruption, even though its industrial structure—featuring higher weights of heavy assets such as energy, industry, finance, and consumer sectors and relatively lower technology-iteration risk—makes it harder to capture AI’s rapid growth. During the repricing process of global markets’ differentiation on AI themes earlier in the year, Europe’s major stock indexes showed stronger resilience, confirming the value of the Survivors.

“Relatively speaking, the Indian market, characterized by high labor-intensive service and outsourcing attributes, has recently lagged more.” Li Chong said candidly that as AI becomes widely used in areas such as code generation, automated testing, and advisory support, the market is increasingly concerned that India’s IT business model—centered on labor-intensive outsourcing services—may face a structural shrinkage in demand in the future. This differentiation clearly shows that the HALO narrative is not a simple rotation of asset categories, but rather a renewed understanding of the business essence in the AI era: in an environment where code expands and technology iterates faster, real assets and business models with low attrition rates become scarce resources instead.

Cross-market transmission of market conditions:

Differentiated pattern and clarification of investment boundaries

The structural changes in the HALO narrative are like a stone thrown into global capital markets, creating ripples upon ripples across different markets and sectors, forming a distinct pattern of differentiation. Li Chong said that in the U.S. stock market, before the Israel-Iran conflict broke out, AI infrastructure, energy and power, semiconductor equipment, and related physical resource assets strengthened at least on a phase-by-phase basis, becoming structural winners; meanwhile, some software-type assets with higher replacement risk were under relative pressure, turning into short-term Losers. This differentiation is reflected not only at the industry level, but also triggers the repricing of country-level assets across the globe.

When it comes to emerging markets such as A-shares and Hong Kong-listed stocks, the HALO narrative resonates deeply with the “code expansion—real-asset scarcity” framework proposed by the Strategy Group of CITIC Securities’ Research Department. The framework argues that the rapid global expansion in the amount of code brought by AI may clearly disrupt all kinds of businesses with low physical dependence and high market competition, while businesses with high physical dependence and high regulatory and emotional barriers may become even scarcer. Under this logic, regardless of whether AI Agents can immediately play a substitution role for these businesses in the short term, market investors still actively go long on fortress-like assets that are currently isolated from AI shocks, while temporarily avoiding businesses that are most easily overturned by AI-disrupted, innovation-driven destruction. Resource-related industries, traditional manufacturing, energy, services consumption, real-estate chain, and other related sectors fall precisely within the category of fortress-like assets temporarily isolated from generative AI shocks.

“However, the HALO narrative is not a universal key; its investment boundaries are equally clear. HALO is not the same as structural winners. Relying solely on defensive positioning based on survival certainty cannot form a long-term main line.” Li Chong believes that this judgment rests on the fact that HALO assets more often reflect advantages in earnings stability and low replacement risk, and that their valuation increases come from a decline in risk premiums rather than accelerated earnings. Assets that truly have potential for sustained excess returns should be located at the key nodes along the AI expansion path, forming deep binding with resource bottlenecks or directions of technological upgrades, and possessing characteristics of earnings elasticity. Survival certainty can bring a phase-specific premium, but it does not automatically translate into sustained growth premiums—this is exactly the phase-specific nature of the HALO rally.

Such phase-specific attributes can easily cause investors to fall into two major cognitive misconceptions. Li Chong explains: first, equating HALO assets with long-term investment value targets and ignoring their defensive nature; second, overconcentrating in low-risk assets and missing the true winners along the AI expansion path. In fact, HALO trading is more a reflection of phase-by-phase style rotation rather than a new long-term growth paradigm. When AI has no clear new direction, HALO is worth paying attention to; once the AI expansion direction becomes clear, capital may once again embrace high growth. This process of dynamic balance forms the core characteristic of the HALO rally.

It is worth noting that Li Chong said that distinguishing Winner, Survivor, and Loser is the key to capturing the HALO rally. These three have dynamic attributes: different stages’ expansion bottlenecks, capital expenditure directions, and changes in the industrial chain’s center of gravity will all lead to a rebalancing of win-loss relationships within the structure. The current phase-specific winners are concentrated in AI infrastructure and resource bottleneck-related assets. At the same time, some HALO assets that the market regards as safe may also lose their “safety” attributes as the industrial logic evolves. Investors need to build a dynamic assessment framework rather than statically classifying certain industries as Survivors—this is the complexity and appeal of the HALO rally.

Market outlook:

HALO rotation and logic for multi-track mainline positioning

At the current point in time, the question of the sustainability of the HALO rally has become a key focus for the market. Fundamentally, HALO trading is more a manifestation of phase-based style rotation than a new long-term growth paradigm.

Li Chong believes that the emergence of HALO confirms the market’s maturation in risk pricing, but defensive positioning based solely on survival certainty cannot form a long-term main line. The real long-term main line still needs to be deeply tied to the AI expansion path, and be able to obtain sustained earnings elasticity through resource bottlenecks or technological upgrades. Therefore, the sustainability of the HALO rally will depend to a large extent on the pace of breakthroughs in AI technology and changes in the macro environment.

“Under the current environment of ongoing geopolitical conflicts disturbing risk appetite and liquidity, investment strategies need to find a balance between Survivors and Winners. For investors who are positioning in overseas markets, it is a relatively prudent choice to focus in phases on the two types of assets: Winners and Survivors.” Li Chong said that Winner assets are mainly concentrated at key nodes along the AI expansion path—areas deeply tied to resource bottlenecks such as energy and power, semiconductor equipment, AI infrastructure, and so on—these assets can obtain deterministic growth tailwinds during AI Disruption. Survivor assets, by contrast, are characterized by heavy assets and low attrition rates, such as Europe’s traditional manufacturing, energy facilities, and chain stores. These assets can maintain relatively stable performance through AI technology iterations and provide defensive returns.

In addition to the HALO main line, overseas markets in 2026 also have multiple investment leads worth watching; these main lines interweave with the HALO narrative and form a complex rotation relationship. Li Chong said that in the short term, the U.S.-Israel/Iran conflict remains the core variable affecting the market. Until the fighting clearly eases, overseas markets are advised to focus on safe-haven and inflation-related main lines. Safe-haven assets partially overlap with resource-type assets within HALO and can provide dual protection in an uncertainty environment. If the fighting clearly eases, oil prices fall, and risk appetite improves, and the global economy can return to the pre-war macro environment, then one can focus on resource and pro-cyclical sectors under a moderate recovery cycle—these sectors will benefit from the dual tailwinds of economic recovery and AI expansion.

For emerging markets such as A-shares and Hong Kong-listed stocks, the transmission impact of the HALO narrative will show localization characteristics. CITIC Securities’ Research Department Strategy Group believes that unlike overseas HALO trades that focus on screening “immune assets” under AI shocks, China’s “real-asset scarcity” trading places more emphasis on the logic of “AI exposure + supply constraints = a price-hike expectation.” Sectors such as resources, traditional manufacturing, and energy are not only Survivors under AI shocks, but may also gain additional growth momentum from global supply-chain restructuring and domestic economic recovery. Meanwhile, industries with high physical dependence and high regulatory and emotional barriers—such as high-end baijiu, top luxury goods, and services consumption—may also receive a valuation reappraisal within the framework of “code expansion and real-asset scarcity.”

From a longer-term perspective, Li Chong believes that the HALO narrative and AI Disruption are not opposing relationships, but rather two aspects that complement each other and maintain dynamic balance. Investors need to abandon either-or binary thinking and adopt a portfolio strategy of “short-term defense + long-term growth.” When AI technology has not yet made clear its new direction, allocating to HALO assets to capture survival premiums is appropriate. When the AI expansion path becomes clear, adjusting positions in time to embrace Winner assets is key. This flexible investment strategy is a rational choice to deal with the current complex market environment—it can both avoid the uncertainty risks brought by AI technology iteration and capture the long-term growth opportunities brought by AI development.

(Source: China Securities Journal)