AI HALO Investing: The Hidden Gem Everyone’s Chasing in 2024

There’s a term taking over investment circles right now, and it’s not what you’d expect in an AI-focused world.

HALO investing is quietly reshaping how the world’s biggest investment banks—Morgan Stanley (Mo Gen Shi Dan Li 摩根士丹利) and Goldman Sachs (Gao Sheng 高盛) are thinking about artificial intelligence opportunities.

While everyone’s obsessing over ChatGPT and software companies, a smarter class of investors is looking at something completely different: the heavy, physical infrastructure that powers the entire AI revolution.

Let’s break down what HALO investing actually is, why it matters for your portfolio, and where the real money’s flowing.

What Does HALO Stand For? (It’s Not What You Think)

HALO is an acronym that stands for “Heavy Assets, Low Obsolescence.”

That’s it.

But those two words contain a philosophy that’s reshaping trillions of dollars in investment decisions right now.

Here’s what each component actually means:

Heavy Assets:

• Business models built on massive physical capital foundations
• High barriers to entry and replication
• Requires significant infrastructure investment to compete
• Think: power plants, data centers, physical server farms

Low Obsolescence:

• Assets that maintain economic relevance through multiple technological cycles
• The infrastructure is a “must-have,” not something that gets replaced by innovation
• Enduring value regardless of whether new software or AI models emerge
• Think: energy infrastructure, data center facilities, manufacturing capacity

The traditional HALO playbook covers sectors like power grids, oil and gas pipelines, public utilities, transportation networks, critical equipment manufacturing, and industrial production capacity.

Morgan Stanley (Mo Gen Shi Dan Li 摩根士丹利) actually built a HALO basket (MSXXHALO) that breaks this down into seven structural pillars:

• Materials
• Utilities
• Railways
• Pipelines
• Waste management
• Defense
• Signal towers

In contrast, sectors like software, IT services, publishing, gaming, and logistics platforms are the exact opposite of HALO—they’re high-velocity, constantly disrupted, and face rapid obsolescence.

Enter AI HALO: The Real AI Investment Play

Here’s where it gets interesting for tech investors.

Most people assume HALO investing is just about “old economy” stuff—oil refineries, pipelines, and infrastructure companies stuck in the past.

Wrong.

The real insight is that the AI field has its own HALO assets—and they’re about to be the biggest beneficiaries of the AI boom.

According to research from Guosen Securities (Guo Xin Zheng Quan 国信证券), the AI HALO landscape breaks down into three primary layers across 16 sub-sectors.

Understanding these layers is critical because each one represents a different risk/reward profile and investment opportunity:

Layer 1: The Computing Power Service Layer

This is where companies rent computational resources to AI developers and enterprises.

• Computing power leasing – Companies providing on-demand GPU and TPU access
• Internet Data Centers (IDC) – Physical facilities where servers live and operate

Think of this as the “landlord” layer—they own the property where all the AI action happens.

Layer 2: The Computing Power Infrastructure Layer

This is the actual equipment and systems that make the data centers function.

• AI servers – The hardware that runs AI workloads
• Server power supplies – Critical for keeping everything running
• Liquid cooling systems – Essential infrastructure for preventing overheating
• Network switches – The connective tissue between systems

This layer is the backbone—without it, nothing runs.

Layer 3: Core Hardware and Components Layer

The deepest layer is where the actual silicon and components get made.

• Wafer foundries – Where chips are manufactured at scale
• AI chips – The processors that power everything
• Circuit boards and modules – Connecting components together
• PCB (Printed Circuit Boards) and substrates – The foundation for electronic assembly
• Memory packaging and testing – RAM and storage components
• Optical modules – High-speed data transmission between systems

This is where the highest margins typically exist, but also where capital requirements are most intense.

The Capital Expenditure Explosion: Where the Real Money Is

Here’s the number that should catch your attention:

Four major tech giants are about to spend nearly ¥9.42 trillion RMB ($1.31 trillion USD) on AI infrastructure between 2024 and 2026.

That’s Google (Gu Ge 谷歌), Microsoft (Wei Ruan 微软), Meta, and Amazon (Ya Ma Xun 亚马逊).

Let’s put that in perspective:

• 2024-2026 total spending: ¥9.42 trillion RMB ($1.31 trillion USD)
• 2026 alone: ¥4.74 trillion RMB ($660 billion USD)
• Comparison: This three-year total is triple the capital expenditure from 2021-2023

This isn’t gradual growth.

This is an explosion.

What are they spending this money on?

• AI chips
• AI servers
• Network switches
• Other critical hardware infrastructure

And here’s the crucial part: all of this hardware needs physical infrastructure to operate.

You need:

• Internet Data Centers to house it all
• Power plants and energy infrastructure to run it
• Cooling systems to manage the heat
• Manufacturing capacity to build the components
• Supply chain logistics to deliver everything

These are the AI HALO assets—and they’re about to capture massive value from this capital expenditure wave.

Why This Matters for Your Investment Strategy

The HALO investing thesis is simple but powerful:

When tech giants spend trillions on AI infrastructure, the real winners aren’t necessarily the software companies—they’re the companies that own and operate the heavy, physical assets that make AI possible.

This is a different way of thinking about AI exposure.

Instead of betting on the next OpenAI or a trendy AI software startup, you’re betting on the picks and shovels—the fundamental infrastructure that every AI company absolutely needs.

And unlike software, which can become obsolete overnight, computing infrastructure stays relevant through multiple tech cycles.

The data center you build today will be valuable in 2030, 2040, and beyond.

The AI chip manufacturer’s equipment will see upgrades but won’t disappear.

That’s the power of HALO investing—you’re not timing the next AI trend or betting on a specific technology.

You’re investing in the physical foundation that all of those trends require.

As the AI infrastructure boom accelerates through 2024-2026 and beyond, HALO assets will remain at the center of AI investment opportunities.

The question for investors isn’t whether AI will reshape the world—it’s whether you understand where the capital will actually flow.

—

References

• AI HALO Industry Chain Analysis – East Money Research Center (Dong Fang Cai Fu 东方财富)
• Industry Strategies and HALO Asset Frameworks – Guosen Securities (Guo Xin Zheng Quan 国信证券)
• The HALO Framework: Investing in Heavy Assets with Low Obsolescence – Morgan Stanley
• AI Infrastructure and the Next Wave of Investment – Goldman Sachs