Pricing power emerges as Equinix accelerates AI capacity delivery
- Annualized gross bookings reached a record $533 million in the third quarter.
- Equinix demonstrated pricing power with new deal rates up 12% year over year.
- Management expects to deliver more than 13,000 cabinets in the fourth quarter to meet demand.
- Artificial intelligence drives roughly 60% of the largest deals for the company.
- Execution risk is the main focus as annual capital expenditures scale to between $5 billion and $7 billion.
A structural shift to AI
Equinix is a foundational landlord for the modern internet. Its data centers let companies place servers near clouds, networks, and business partners. This matters even more today as large enterprises transition artificial intelligence workloads from pilot testing into real, distributed applications.
The bull case was cemented in the third quarter of 2026. Following a massive guidance raise earlier in the year, demand continued to surge. Annualized gross bookings hit a record $533 million, up 35% year over year. More importantly, clear pricing power has emerged. Pricing on new deals was up 12% in the quarter, proving that customers will pay a premium for constrained power and connectivity.
To meet this demand, Equinix has heavily accelerated its delivery timeline. Management now expects to deliver over 13,000 cabinets in the fourth quarter of 2026. This is a massive pull forward from earlier plans, supported by pre purchased supply chains and deep utility partnerships.
The rapid growth changes the debate for investors. The bear case has completely rotated away from demand concerns. Now, the worry is purely about execution. Equinix must manage an accelerated build plan requiring $5 billion to $7 billion in annual capital expenditures. Any shakeout in new artificial intelligence cloud startups could also create localized churn, though management notes this group currently makes up less than 2% of recurring revenue.
Rent, connections, and density
Equinix makes most of its money from recurring contracts. Customers rent secure space and power in Equinix data centers, then pay for ways to connect to cloud providers, networks, partners, and other customers. More than 90% of revenue is recurring, and contracts usually run one to five years.
The core advantage of the model is the network effect. A data center with many networks, clouds, and enterprises becomes more useful to each new customer. That makes the platform sticky because leaving can mean giving up critical private connections. The company maintains extremely low customer churn of just 1.7%.
Artificial intelligence changes the physical math. By late 2025, about 60% of large deals were driven by artificial intelligence workloads, and those deals used 33% more power density than standard deployments. Higher density supports revenue growth, but it also raises the cost and complexity of cooling. Liquid cooling bookings alone jumped more than tenfold recently.
As a real estate investment trust, Equinix must distribute at least 90% of its taxable income to shareholders each year. This makes funding a massive build plan a complex balancing act of debt and equity.
What customers buy
Colocation
Customers rent secure data center space, power, and cooling. This includes rapidly scaling liquid cooled infrastructure for high density needs.
Interconnection and data exchange
Equinix sells direct private links among clouds and networks. New products like Fabric One create an automated, self composing network offering.
xScale data centers
These sites serve large footprint hyperscale customers through joint ventures. They add capacity for massive cloud demand but require significant capital.
Edge solutions
Services that place hardware closer to end users. This newly expands into artificial intelligence with the Equinix Inference Exchange for open models.
Support services
Equinix provides remote hands and professional help for data center deployment. These services deepen customer ties and trust.
A global revenue base
Segment mix is from the 2025 fiscal year. The customer base is highly diversified, with no single customer accounting for more than 2.5% of total revenue.
What could go wrong
Build-out execution
High impact · Medium oddsEquinix expects to deliver over 13,000 cabinets in a single quarter while managing $5 billion to $7 billion in annual capital spending. If projects arrive late, supply chains break, or costs spike, return yields will compress.
Power constraints
High impact · High oddsThe broader industry faces structural headwinds from power availability and restrictive legislation. While Equinix builds smaller facilities of 40 to 60 megawatts to navigate community pushback, macro power shortages remain a risk.
Credit risk from new cloud startups
Medium impact · Low oddsThe rapid proliferation of specialized artificial intelligence cloud providers creates potential credit risk. A shakeout in this space could create localized bad debt or customer churn.
Funding strain
Medium impact · Medium oddsEquinix must fund a massive capital expenditure pipeline while distributing at least 90% of taxable income to maintain its real estate investment trust status. The balance sheet will take on additional leverage to support the expansion.
In one breath
What does Equinix actually do?
Equinix runs data centers where companies place critical computing gear. It also sells private connections to clouds, networks, customers, and partners inside those sites.
Why is artificial intelligence important for Equinix?
These systems need dense computing power, strong cooling, and fast private connections. Roughly 60% of Equinix's largest deals are now tied to artificial intelligence workloads.
Is Equinix a REIT?
Yes. Equinix operates as a real estate investment trust for U.S. federal income tax purposes. This means it must distribute at least 90% of its taxable income to shareholders each year.
What is the biggest investor debate?
The debate is no longer about demand. It is entirely about whether Equinix can execute its massive, accelerated capacity expansion plan without supply chain delays or cost overruns.
Sources and research notes
This page combines Finn's company research with public filings and other cited materials. The thesis is reviewed when material company information changes; Finn Scores use the latest available scoring data.
- Thesis reviewed
- November 1, 2026
- Score data
- November 15, 2026
- Reviewed by
- Shivam Bharuka
Comparable REIT - Specialty companies
Companies near Equinix, Inc. in Finn's REIT - Specialty industry ranking.

