The Great 2026 Re-Rating: Why Wall Street is Trading Code for Concrete

The hottest trade on Wall Street isn't a new AI model or a flashy SaaS startup — it's concrete, copper, and kilowatts. Welcome to the Great Re-Rating of 2026, where investors are discovering that the biggest winners in the AI revolution might be the companies you'd least expect: the ones that build, power, and transport the physical world.

What Is the HALO Trade?

HALO stands for Heavy Assets, Low Obsolescence — a term coined by Josh Brown (CEO of Ritholtz Wealth Management) to describe companies whose physical infrastructure becomes more valuable as AI scales, not less. Think about it: every new data center needs power. Every GPU needs rare earth minerals. Every AI model needs cooling infrastructure. These aren't disruption victims — they're disruption beneficiaries.

The thesis is elegant in its simplicity: while AI might automate code, it cannot automate a copper mine, a railroad, or a nuclear power plant.

The AI Supply Chain's Hidden Dependencies

Our Early Signal AI ecosystem maps 10 layers of the AI supply chain, from raw materials to AI applications. What we've discovered is that not all layers face the same disruption risk:

Core HALO Layers (Beneficiaries)

• Layer 0: Raw Materials — Rare earths, copper, silicon. AI needs more of these, not fewer.
• Layer 1: Semiconductor Manufacturing — ASML's lithography monopoly, TSMC's foundries. Irreplaceable physical infrastructure.
• Layer 3: Infrastructure — Data centers (Equinix, Digital Realty), power delivery (Vertiv, Eaton). AI scaling = infrastructure scaling.
• Layer 4: Cloud & Compute — Hyperscalers need physical footprints. Server farms aren't virtual.
• Layer 8: AI Transformation — Enterprises spending on physical upgrades (Honeywell, Emerson Electric).
• Layer 9: Secondary Beneficiaries — Power utilities (Vistra, Constellation Energy, NRG) seeing demand surge from AI data centers.

Anti-HALO Layer (Vulnerable)

• Layer 7: AI Applications — SaaS companies (Salesforce, ServiceNow, Workday) face the highest disruption risk. If AI can write code, it can replace software.

Why Now? The 2026 Inflection Point

Several forces are converging in 2026:

1. Power Demand Crisis. AI data centers are projected to consume 4–6% of US electricity by 2028, up from under 2% in 2023. Utility stocks aren't boring anymore — they're critical infrastructure.

2. CapEx Super-Cycle. Microsoft, Google, Amazon, and Meta have committed over $200 billion in combined data center spending through 2027. That money flows directly to construction firms, electrical equipment makers, and materials companies.

3. The Software Devaluation. As AI coding tools mature, the moat around enterprise software narrows. Why pay $300/seat/month for a CRM when an AI agent can do the same thing? The market is pricing this in.

4. Supply Chain Bottlenecks. Transformer manufacturing has a 3-year backlog. High-voltage switchgear is sold out through 2027. These physical constraints create pricing power that pure-software companies can only dream of.

The HALO Rotation Signal

At Early Signal, we've built a HALO rotation signal that measures the momentum spread between Core HALO layers and Anti-HALO layers within our AI supply chain universe. When the spread is positive (HALO layers outperforming), it suggests the market is pricing in physical-asset appreciation. When negative, software and application layers are leading.

This isn't about abandoning AI — it's about understanding which parts of the AI ecosystem have durable competitive advantages. A railroad's right-of-way is a 150-year-old moat. A SaaS company's codebase might be a 15-month advantage.

Building a HALO Portfolio

Our HALO universe includes stocks across seven sub-sectors:

• Industrials & Machinery: Caterpillar (CAT), John Deere (DE) — building the data centers
• Railroads & Transport: Union Pacific (UNP), CSX, Norfolk Southern (NSC) — moving the materials
• Energy Majors: Exxon Mobil (XOM), Chevron (CVX), ConocoPhillips (COP) — powering the grid
• Utilities: NextEra Energy (NEE), Southern Company (SO), Duke Energy (DUK) — delivering electrons
• Defense & Aerospace: Lockheed Martin (LMT), RTX, General Dynamics (GD) — AI for national security
• Materials & Mining: Nucor (NUE), Freeport-McMoRan (FCX), Linde (LIN) — raw inputs
• Retail & Logistics: Walmart (WMT), Home Depot (HD), Lowe's (LOW) — physical distribution networks

These companies share common characteristics: tangible assets that appreciate with AI demand, high barriers to entry, and business models that AI enhances rather than threatens.

Risks and Considerations

The HALO trade isn't risk-free:

• Valuation risk: If the AI CapEx cycle slows, infrastructure stocks could give back gains.
• Rate sensitivity: Many heavy-asset companies carry significant debt. Rising rates pressure margins.
• Commodity cycles: Energy and materials stocks are cyclical. A recession would hurt regardless of AI trends.
• Execution risk: Not every "old economy" company will successfully integrate AI. The winners will be those that use AI to optimize their irreplaceable physical assets.

Code Fades, Concrete Endures

The irony of the AI revolution is that its biggest beneficiaries might not be AI companies at all. While the market obsesses over which language model is slightly better, the real wealth is being created by the companies that build the physical substrate on which all AI runs.

The HALO trade isn't a bet against AI — it's a bet on the full stack. And at the bottom of that stack, beneath all the algorithms and APIs, you'll find concrete, copper, and kilowatts.

Explore our AI HALO Universe to see the full supply chain analysis and real-time HALO rotation signals.