A message from our friends at Banyan Hill (sponsor)

Dear Reader,

We believe we've found the next Elon Musk – and so does Peter Thiel, who just made his largest investment ever: $1 billion into this man's company.

Unfortunately, the company is private. So unless you know Pete personally or have some deep connections... you can't get in.

Our research has uncovered a 4-letter ticker symbol that gives you early-stage exposure to this pre-IPO opportunity.

You won't see this ticker mentioned on CNBC or Yahoo Finance...

We're going to give you this pre-IPO ticker symbol completely free in this briefing.

(No strings attached – you will get all 4 letters of the ticker for free)

But you're going to want to move quickly.

Once it goes public, all the easy pre-IPO gains will be gone.

Don't miss the chance to ride this wave before the masses catch on.

Addison Wiggin
Founder, Grey Swan Investment Fraternity

FEATURED ARTICLE

AI Infrastructure Is the Real Trade — Data Centers, Power, and Cooling Are the New Battleground

The Rotation Beneath the Surface

The market’s first AI wave was software-driven — chatbots, copilots, model releases.

The second wave is physical.

Large language models require exponential increases in:

  • Compute density

  • Power availability

  • Cooling capacity

  • Network throughput

  • Real estate footprint

This is where capital is rotating now.

Active traders should recognize the structural shift:

AI is no longer just a software multiple story.
It is a capex and infrastructure cycle story.

And infrastructure cycles trade differently.

Macro Context: Why AI Infrastructure Is Now the Focus

1) Hyperscale Capex Is Expanding

Major hyperscalers have increased capital expenditures materially over the past two years, driven by AI training and inference workloads.

AI workloads require:

  • 3–5x power density vs traditional cloud racks

  • Advanced liquid cooling vs air cooling

  • Specialized grid connectivity

This is not incremental demand.

It is step-function demand.

2) Power Is the New Constraint

Data centers are power-intensive assets. AI training clusters can require:

  • 30–100 megawatts per facility

  • High-voltage grid interconnects

  • Redundant power and backup systems

Regions with constrained grid capacity are now experiencing multi-year development delays.

That supply constraint supports pricing power for:

  • Established data center landlords

  • Cooling system providers

  • Electrical infrastructure vendors

3) AI Compute Is Concentrated

Unlike consumer tech cycles, AI infrastructure spending is concentrated among:

  • Hyperscalers

  • Sovereign AI initiatives

  • Enterprise inference clusters

Concentration increases revenue visibility for suppliers — but also creates dependency risk.

Sector Breakdown

1) Data Center REITs

Equinix (EQIX)

Business model:

  • Retail colocation

  • Interconnection hubs

  • Dense metro-based infrastructure

Revenue profile:

  • Subscription-based

  • Recurring cash flows

  • High switching costs

Key AI exposure:

  • AI model training clusters colocated in high-connectivity hubs

  • Enterprise inference workloads

What matters for traders:

  • Bookings growth

  • Pricing spreads

  • Power availability commentary

Digital Realty (DLR)

Business model:

  • Wholesale and hyperscale leasing

  • Larger campus developments

  • Long-term contracts

AI exposure:

  • Hyperscale buildouts

  • Multi-megawatt deployments

DLR trades more like:

  • A capex expansion proxy

  • A hyperscale growth vehicle

The distinction:
EQIX = interconnection density
DLR = scale capacity

Traders should treat them differently.

2) Thermal & Power Infrastructure

Vertiv (VRT)

Vertiv is not a REIT.

It is an equipment and systems provider:

  • Liquid cooling systems

  • Power management systems

  • Thermal management solutions

  • Rack-level infrastructure

AI density increases cooling demand disproportionately.

Traditional air cooling is insufficient at high rack densities.

Liquid cooling adoption is accelerating.

Vertiv’s order backlog and margin expansion have reflected this shift.

Financial Profile Snapshot (Structural Themes)

While exact quarterly figures fluctuate, the structural patterns are clear:

EQIX

  • High EBITDA margins

  • Recurring revenue base

  • Steady FFO growth

  • Pricing power in constrained metros

DLR

  • Long-term lease duration

  • Exposure to hyperscale capex cycles

  • Higher sensitivity to new build timing

VRT

  • Revenue growth tied to hyperscale spending

  • Margin expansion as product mix shifts toward AI cooling

  • Operating leverage as scale increases

The key difference:

REITs trade off yield + FFO growth.
VRT trades off earnings growth + operating leverage.

Technical Framework

Infrastructure names often trade in multi-month trends, not daily momentum spikes.

Active traders should focus on:

1) 50-Day and 200-Day Moving Averages

  • AI infrastructure trends often respect 50-day support in strong uptrends.

  • Breaks below 200-day shift narrative from growth to cycle peak.

2) Relative Strength vs SPY

If:

  • EQIX and DLR outperform SPY on down market days → institutional accumulation.

If:

  • VRT outperforms semiconductors on pullbacks → cooling theme intact.

3) Volume Confirmation

Infrastructure breakouts require:

  • Above-average volume

  • Follow-through sessions

Low-volume pops are often headline-driven.

Scenario Modeling

Base Case: Sustained AI Capex Cycle

Catalyst:

  • Continued hyperscaler capex guidance increases

  • Power constraint commentary persists

  • Liquid cooling adoption accelerates

Impact:

  • EQIX and DLR grind higher on stable FFO growth

  • VRT outperforms on earnings revisions

Trader Approach:

  • Buy pullbacks to 50-day moving averages

  • Monitor backlog and bookings commentary

  • Favor VRT in momentum phases

Bull Case: Power Scarcity Premium

Catalyst:

  • Grid constraints worsen

  • Development timelines extend

  • Pricing power increases

Impact:

  • Data center REITs see multiple expansion

  • Infrastructure equipment names see margin expansion

Trader Approach:

  • Watch for breakouts above prior highs on strong volume

  • Monitor upward FFO revisions for REITs

  • Ride trend while respecting trailing stops

Bear Case: Capex Moderation

Catalyst:

  • Hyperscalers signal slowing AI spend

  • Model efficiency reduces compute requirements

  • Recession concerns reduce enterprise demand

Impact:

  • DLR corrects first (hyperscale sensitivity)

  • VRT sees earnings revision pressure

  • EQIX holds better due to recurring structure

Trader Approach:

  • Rotate into higher-quality balance sheets

  • Use relative weakness vs SPY as exit signal

  • Avoid catching falling knives below 200-day

Active Trader Playbook

If AI Headlines Drive Capex Higher

Watch:

  • VRT first for momentum

  • DLR for hyperscale confirmation

  • EQIX for stability

Execution:

  • Enter on pullbacks, not vertical spikes

  • Confirm with volume

  • Set invalidation below 50-day MA

If Yields Rise Sharply

REITs are rate-sensitive.

Watch:

  • EQIX and DLR relative strength vs utilities and other yield plays

If they hold despite rising yields:

  • Institutional accumulation likely.

If they break:

  • Macro pressure outweighing AI narrative.

If Market Rotates Out of Semis

Infrastructure often lags semiconductors by weeks.

If:

  • NVDA/AMD stall

  • Infrastructure begins outperforming

That signals rotation from compute to capacity.

This is a critical inter-market tell.

What Makes This Trade Different from the 2023 AI Rally

2023:

  • Multiple expansion

  • Narrative-driven

  • Software + chips

2024–2026:

  • Physical asset cycle

  • Power constraint economics

  • Margin expansion via pricing

Infrastructure trades on:

  • Build cycles

  • Supply constraints

  • Backlog visibility

It is slower but more durable.

Risk Factors Active Traders Must Monitor

  1. Grid policy changes

  2. Hyperscale spending revisions

  3. Cooling technology disruption

  4. REIT financing costs

  5. AI model efficiency improvements

Infrastructure cycles can overshoot.

Late-stage momentum entries carry risk.

Final Take

AI’s next leg is not a chatbot story.

It is a concrete-and-copper story.

The real bottleneck is:

  • Power

  • Cooling

  • Physical footprint

EQIX and DLR represent the real estate layer.
VRT represents the thermal and power layer.

The trade is not about guessing AI demand.

It is about identifying:

  • Capex persistence

  • Pricing power

  • Relative strength

Preparation beats prediction.

Build a watchlist around:

  • 50-day pullbacks

  • Volume-confirmed breakouts

  • Relative strength vs SPY

  • Yield sensitivity reactions

Then let institutional flows confirm which layer of AI infrastructure is leading.

Editorial Disclaimer

This commentary is for informational and educational purposes only and does not constitute investment advice. All market strategies involve risk, and past performance is not indicative of future results. Readers should conduct their own analysis or consult a licensed financial professional before making investment decisions.

Recommended for you

View all
caret-right