The impact of artificial intelligence is reshaping the valuation logic of global stock markets. Given the structural particularities of the Asia-Pacific market, Bank of America Securities suggests the crowded "HALO" strategy—focusing on heavy assets and low obsolescence—has clear limitations. Instead, the REAL framework, which emphasizes regulatory barriers, enduring cycles, asset-heavy attributes, and local services, may offer a more suitable approach. This is especially relevant in markets with ample production capacity where heavy industries lack scarcity-driven moats, as the protective effect of the HALO strategy may be overestimated.
According to market sources, a team of analysts led by Winnie Wu at Bank of America noted in a recent report that the U.S. software sector has lost over $2 trillion in market value in the past five months, while India's IT sector has declined more than 40% from its peak in December 2024. Media, e-commerce, and fintech stocks across the Asia-Pacific region have also faced large-scale sell-offs. Concurrently, capital has rapidly rotated into heavy-asset sectors such as semiconductors, capital goods, energy, and utilities, leading to a surge in popularity for the HALO trade.
However, Bank of America warns that the HALO strategy has limitations in markets with abundant industrial capacity. In heavy industries lacking regulatory or scarcity-based moats, heavy assets can turn into liabilities, as AI shortens R&D cycles and enables alternative technological pathways, thereby exacerbating overcapacity and competition.
As a result, the REAL framework has been introduced to reassess corporate survival risks. This framework screens companies across four dimensions: regulatory barriers, enduring cycles, asset intensity, and local service density. Research from Bank of America indicates that even within sectors most impacted by AI—such as software and consumer internet—leading companies with REAL characteristics have demonstrated superior resilience to downturns and long-term investment value.
**Limitations of the HALO Strategy: Heavy Assets Are Not a Universal Moat** The core rationale behind the market's pursuit of HALO sectors lies in the lengthy construction cycles of tangible assets, which are difficult for AI to quickly replace. While this logic holds in areas where asset scarcity is assured, its limitations become evident in markets abundant in engineering talent and industrial capacity.
Bank of America points out that in heavy industries lacking regulatory constraints or scarcity moats—such as automobiles, solar energy, steel, and cement—supply can easily outstrip demand, triggering intense price competition. In these sectors, heavy assets not only fail to offer protection but may even become burdens, particularly as AI further shortens R&D cycles and opens alternative technological pathways. Conversely, some light-asset but labor-intensive service industries, such as healthcare and dining, may exhibit greater resilience to AI-driven substitution.
Bank of America also clarifies that the REAL framework does not imply immunity to AI disruption. AI can expand addressable markets, lower operating costs, compress innovation cycles, and reduce industry barriers. The key insight is that leading companies in high-REAL sectors face significantly lower existential threats compared to peers in low-REAL industries, especially in an environment of rapid AI-driven disruption where valuation compression in low-REAL sectors may be more prolonged and severe.
**The Four REAL Moats: Redefining Defensiveness** Bank of America defines the four dimensions of the REAL framework as follows:
- **Regulatory Barriers**: Systemically important banks, telecom operators, power and energy suppliers, and defense-related industries. These sectors involve social stability and national security, with strict regulations on licensing, foreign ownership, and critical infrastructure operations. The introduction of AI often brings higher monitoring requirements and human oversight burdens, increasing compliance costs rather than lowering barriers.
- **Enduring Cycles**: Semiconductors, capital goods, aerospace and shipbuilding, pharmaceuticals and biotech, and gaming IP. Barriers in these industries stem from time-intensive accumulation in the real world—advanced chip processes depend on multi-generational expertise and EUV equipment, new wafer fabs take years to build, aircraft require complex airworthiness certifications, drugs undergo multi-stage clinical and regulatory reviews, and game copyrights often span 50 to 70 years, enabling sustained IP monetization. AI can optimize certain processes but cannot bypass these multi-year certification and validation requirements.
- **Asset Heavy**: Natural resources and commodities, power grids and utilities, railways and ports, and livestock farming. Scarcity in these sectors arises from physical constraints—finite mineral reserves, lengthy permitting processes, high infrastructure costs—making it economically unviable for new entrants to replicate existing assets.
- **Local Services**: Hotels, restaurants, property management, childcare, eldercare, pet care, and on-site IT deployment and maintenance. These roles involve non-standardized, on-site environments requiring high human adaptability and low tolerance for error. Current AI and robotic systems are not yet economically viable substitutes.
**REAL Distribution Across Markets: ASEAN's Short-Term Edge and Long-Term Risks** An analysis of the sector composition of the MSCI Asia-Pacific Index reveals significant differences in AI risk exposure across markets. Southeast Asian markets are heavily skewed toward high-REAL sectors: approximately 79% in Singapore, 87% in Malaysia, and 94% in Indonesia, with banking as the core weighting. This structure provides relative resilience in a market environment dominated by AI narratives. Year-to-date, Thailand has gained 14.6% and Malaysia 5.1%, both outperforming India, which declined 10.6%.
However, Bank of America also warns that short-term resilience in Southeast Asian markets could turn into long-term vulnerability. If AI-driven automation lowers manufacturing costs in developed economies and makes near-shoring more attractive, demand for offshoring labor-intensive production could decline. Vietnam, Malaysia, and Thailand all exhibit high ratios of foreign direct investment to GDP and export dependency. Without adequate digital infrastructure and AI talent, these economies face structural challenges such as widening technology gaps and hindered integration into global supply chains.
**AI Disruption Meets Aging Demographics: A Dual-Axis Analysis** Bank of America analyzes AI disruption alongside Asia's low birth rates and accelerating population aging, creating a dual-axis matrix to assess the medium- to long-term structural positioning of various sectors.
Four sectors—healthcare, semiconductors, capital goods, and insurance—occupy the most favorable position. They possess strong near-term moats against AI disruption while benefiting from aging-driven automation demand and increased spending from a wealthier elderly population, making them "structural opportunity" sectors with both defensive attributes and long-term growth potential.
Real estate, utilities, and banking, while possessing high-REAL moats that provide short-term buffers against AI-induced valuation shocks, face long-term pressure from aging populations on housing formation and discretionary consumption. Bank of America judges these three sectors as offering "near-term downside protection against AI volatility but limited upside for long-term valuation re-rating."
Consumer durables, media and entertainment, retail distribution, and automobiles face the most significant pressure. Low AI moats combined with aging-related drags on discretionary spending create a dual disadvantage, making this group the most risk-concentrated under current conditions.
Comments